diff --git a/cv/semantic_segmentation/att_unet/pytorch/.gitignore b/cv/semantic_segmentation/att_unet/pytorch/.gitignore new file mode 100644 index 0000000000000000000000000000000000000000..787d13ec677987a9b2fdce6b96c15047e4a44b27 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/.gitignore @@ -0,0 +1,120 @@ +# Byte-compiled / optimized / DLL files +__pycache__/ +*.py[cod] +*$py.class + +# C extensions +*.so + +# Distribution / packaging +.Python +build/ +develop-eggs/ +dist/ +downloads/ +eggs/ +.eggs/ +lib/ +lib64/ +parts/ +sdist/ +var/ +wheels/ +*.egg-info/ +.installed.cfg +*.egg +MANIFEST + +# PyInstaller +# Usually these files are written by a python script from a template +# before PyInstaller builds the exe, so as to inject date/other infos into it. +*.manifest +*.spec + +# Installer logs +pip-log.txt +pip-delete-this-directory.txt + +# Unit test / coverage reports +htmlcov/ +.tox/ +.coverage +.coverage.* +.cache +nosetests.xml +coverage.xml +*.cover +.hypothesis/ +.pytest_cache/ + +# Translations +*.mo +*.pot + +# Django stuff: +*.log +local_settings.py +db.sqlite3 + +# Flask stuff: +instance/ +.webassets-cache + +# Scrapy stuff: +.scrapy + +# Sphinx documentation +docs/en/_build/ +docs/zh_cn/_build/ + +# PyBuilder +target/ + +# Jupyter Notebook +.ipynb_checkpoints + +# pyenv +.python-version + +# celery beat schedule file +celerybeat-schedule + +# SageMath parsed files +*.sage.py + +# Environments +.env +.venv +env/ +venv/ +ENV/ +env.bak/ +venv.bak/ +.DS_Store + +# Spyder project settings +.spyderproject +.spyproject + +# Rope project settings +.ropeproject + +# mkdocs documentation +/site + +# mypy +.mypy_cache/ + +data +.vscode +.idea + +# custom +*.pkl +*.pkl.json +*.log.json +work_dirs/ +mmseg/.mim + +# Pytorch +*.pth diff --git a/cv/semantic_segmentation/att_unet/pytorch/CITATION.cff b/cv/semantic_segmentation/att_unet/pytorch/CITATION.cff new file mode 100644 index 0000000000000000000000000000000000000000..cfd7cab05ddbbff59663dfd3559f4cf33c88c433 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/CITATION.cff @@ -0,0 +1,8 @@ +cff-version: 1.2.0 +message: "If you use this software, please cite it as below." +authors: + - name: "MMSegmentation Contributors" +title: "OpenMMLab Semantic Segmentation Toolbox and Benchmark" +date-released: 2020-07-10 +url: "https://github.com/open-mmlab/mmsegmentation" +license: Apache-2.0 diff --git a/cv/semantic_segmentation/att_unet/pytorch/LICENSE b/cv/semantic_segmentation/att_unet/pytorch/LICENSE new file mode 100644 index 0000000000000000000000000000000000000000..38e625bf59f9ba58f16f758bebaad865030c1e3e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/LICENSE @@ -0,0 +1,203 @@ +Copyright 2020 The MMSegmentation Authors. All rights reserved. + + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright 2020 The MMSegmentation Authors. + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/cv/semantic_segmentation/att_unet/pytorch/README.md b/cv/semantic_segmentation/att_unet/pytorch/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e6b236746374038dd8c33ddada48c9f1283e95b1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/README.md @@ -0,0 +1,134 @@ +# Attention U-Net: Learning Where to Look for the Pancreas + +## Model descripstion + +We propose a novel attention gate (AG) model for medical imaging that automatically learns to focus on target structures of varying shapes and sizes. Models trained with AGs implicitly learn to suppress irrelevant regions in an input image while highlighting salient features useful for a specific task. This enables us to eliminate the necessity of using explicit external tissue/organ localisation modules of cascaded convolutional neural networks (CNNs). AGs can be easily integrated into standard CNN architectures such as the U-Net model with minimal computational overhead while increasing the model sensitivity and prediction accuracy. + +## Step 1: Installing + +### Install packages + +```shell +yum install mesa-libGL +pip3 install -r requirements.txt +wget http://www.zlib.net/fossils/zlib-1.2.9.tar.gz +tar xvf zlib-1.2.9.tar.gz +cd zlib-1.2.9/ +./configure && make install +``` + +### Build Extension + +```shell +python3 setup.py build && cp build/lib.linux*/mmcv/_ext.cpython* mmcv +``` + +## Step 2: Prepare Datasets + +Download cityscapes from file server or official website [Cityscapes](https://www.cityscapes-dataset.com) + +```shell +mkdir -p data/ +ln -s ${CITYSCAPES_DATASET_PATH} data/ +``` + +## Step 3: Training + +**The available configs are as follows:** + +```shell + +# CityScapes +attunet_res34_512x1024_160k_cityscapes + + +### Training on mutil-cards +```shell +bash train_dist.sh [training args] # config file can be found in the configs directory +``` + +### Example + +```shell +bash train_dist.sh configs/attunet/attunet_res34_512x1024_160k_cityscapes.py 8 +``` + +### Training arguments + +```python +# the dir to save logs and models +work-dir: str = None + +# the checkpoint file to load weights from +load-from: str = None + +# the checkpoint file to resume from +resume-from: str = None + +# whether not to evaluate the checkpoint during training +no-validate: bool = False + +# (Deprecated, please use --gpu-id) number of gpus to +# use (only applicable to non-distributed training) +gpus: int = None + +# (Deprecated, please use --gpu-id) ids of gpus to use +# (only applicable to non-distributed training) +gpu-ids: int = None + +# id of gpu to use (only applicable to non-distributed training) +gpu-id: int = 0 + +# random seed +seed: int = None + +# Whether or not set different seeds for different ranks +diff_seed: bool = False + +# whether to set deterministic options for CUDNN backend. +deterministic: bool = False + +# --options is deprecated in favor of --cfg_options' and it +# will not be supported in version v0.22.0. Override some +# settings in the used config, the key-value pair in xxx=yyy +# format will be merged into config file. If the value to be +# overwritten is a list, it should be like key="[a,b]" or key=a,b +# It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" +# Note that the quotation marks are necessary and that no white space +# is allowed. +options: str = None + +# override some settings in the used config, the key-value pair +# in xxx=yyy format will be merged into config file. If the value +# to be overwritten is a list, it should be like key="[a,b]" or key=a,b +# It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" +# Note that the quotation marks are necessary and that no white +# space is allowed. +cfg-options: str = None + +# job launcher +launcher: str = "none" + +# local rank +local_rank: int = 0 + +# distributed backend +dist_backend: str = None + +# resume from the latest checkpoint automatically. +auto-resume: bool = False +``` + +## Results + +### Cityscapes + +#### Accuracy + +| Method | Crop Size | Lr schd | FPS (BI x 8) | mIoU (BI x 8) | +| ------ | --------- | ------: | -------- |--------------:| +| ATTUNet | 512x1024 | 160000 | 54.5180 | 69.39 | + +## Reference +-Ref: https://mmsegmentation.readthedocs.io/en/latest/dataset_prepare.html#cityscapes +-Ref: https://github.com/open-mmlab/mmsegmentation diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/ade20k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/ade20k.py new file mode 100644 index 0000000000000000000000000000000000000000..efc8b4bb20c981f3db6df7eb52b3dc0744c94cc0 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/ade20k.py @@ -0,0 +1,54 @@ +# dataset settings +dataset_type = 'ADE20KDataset' +data_root = 'data/ade/ADEChallengeData2016' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (512, 512) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations', reduce_zero_label=True), + dict(type='Resize', img_scale=(2048, 512), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2048, 512), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=4, + workers_per_gpu=4, + train=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/training', + ann_dir='annotations/training', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/validation', + ann_dir='annotations/validation', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/validation', + ann_dir='annotations/validation', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/ade20k_640x640.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/ade20k_640x640.py new file mode 100644 index 0000000000000000000000000000000000000000..14a4bb092f1db24eb871cd33fd98555cd814ef9c --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/ade20k_640x640.py @@ -0,0 +1,54 @@ +# dataset settings +dataset_type = 'ADE20KDataset' +data_root = 'data/ade/ADEChallengeData2016' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (640, 640) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations', reduce_zero_label=True), + dict(type='Resize', img_scale=(2560, 640), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2560, 640), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=4, + workers_per_gpu=4, + train=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/training', + ann_dir='annotations/training', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/validation', + ann_dir='annotations/validation', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/validation', + ann_dir='annotations/validation', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes.py new file mode 100644 index 0000000000000000000000000000000000000000..f21867c63e1835f6fceb61f066e802fd8fd2a735 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes.py @@ -0,0 +1,54 @@ +# dataset settings +dataset_type = 'CityscapesDataset' +data_root = 'data/cityscapes/' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (512, 1024) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2048, 1024), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=2, + workers_per_gpu=2, + train=dict( + type=dataset_type, + data_root=data_root, + img_dir='leftImg8bit/train', + ann_dir='gtFine/train', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + img_dir='leftImg8bit/val', + ann_dir='gtFine/val', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + img_dir='leftImg8bit/val', + ann_dir='gtFine/val', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_1024x1024.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_1024x1024.py new file mode 100644 index 0000000000000000000000000000000000000000..f98d929723b4539323ba6c9db867dfa4b01ffb22 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_1024x1024.py @@ -0,0 +1,35 @@ +_base_ = './cityscapes.py' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (1024, 1024) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2048, 1024), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + train=dict(pipeline=train_pipeline), + val=dict(pipeline=test_pipeline), + test=dict(pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_768x768.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_768x768.py new file mode 100644 index 0000000000000000000000000000000000000000..fde9d7c7d8076dabff081fce0989eec6a6f5ff07 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_768x768.py @@ -0,0 +1,35 @@ +_base_ = './cityscapes.py' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (768, 768) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=(2049, 1025), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2049, 1025), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + train=dict(pipeline=train_pipeline), + val=dict(pipeline=test_pipeline), + test=dict(pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_769x769.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_769x769.py new file mode 100644 index 0000000000000000000000000000000000000000..336c7b254fe392b4703039fec86a83acdbd2e1a5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_769x769.py @@ -0,0 +1,35 @@ +_base_ = './cityscapes.py' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (769, 769) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=(2049, 1025), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2049, 1025), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + train=dict(pipeline=train_pipeline), + val=dict(pipeline=test_pipeline), + test=dict(pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_832x832.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_832x832.py new file mode 100644 index 0000000000000000000000000000000000000000..b9325cc0087dffb0d3f51fe8a1ebcc45eb76b30d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/cityscapes_832x832.py @@ -0,0 +1,35 @@ +_base_ = './cityscapes.py' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (832, 832) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=(2048, 1024), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2048, 1024), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + train=dict(pipeline=train_pipeline), + val=dict(pipeline=test_pipeline), + test=dict(pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/coco-stuff10k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/coco-stuff10k.py new file mode 100644 index 0000000000000000000000000000000000000000..ec0496928b9464406a4013023c553cf3e7da526b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/coco-stuff10k.py @@ -0,0 +1,57 @@ +# dataset settings +dataset_type = 'COCOStuffDataset' +data_root = 'data/coco_stuff10k' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (512, 512) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations', reduce_zero_label=True), + dict(type='Resize', img_scale=(2048, 512), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2048, 512), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=4, + workers_per_gpu=4, + train=dict( + type=dataset_type, + data_root=data_root, + reduce_zero_label=True, + img_dir='images/train2014', + ann_dir='annotations/train2014', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + reduce_zero_label=True, + img_dir='images/test2014', + ann_dir='annotations/test2014', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + reduce_zero_label=True, + img_dir='images/test2014', + ann_dir='annotations/test2014', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/coco-stuff164k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/coco-stuff164k.py new file mode 100644 index 0000000000000000000000000000000000000000..a6a38f2ac4d1a39d4a89bb101462d3ec805c3aff --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/coco-stuff164k.py @@ -0,0 +1,54 @@ +# dataset settings +dataset_type = 'COCOStuffDataset' +data_root = 'data/coco_stuff164k' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (512, 512) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=(2048, 512), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2048, 512), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=4, + workers_per_gpu=4, + train=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/train2017', + ann_dir='annotations/train2017', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/val2017', + ann_dir='annotations/val2017', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + img_dir='images/val2017', + ann_dir='annotations/val2017', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_context.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_context.py new file mode 100644 index 0000000000000000000000000000000000000000..ff65bad1b86d7e3a5980bb5b9fc55798dc8df5f4 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_context.py @@ -0,0 +1,60 @@ +# dataset settings +dataset_type = 'PascalContextDataset' +data_root = 'data/VOCdevkit/VOC2010/' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) + +img_scale = (520, 520) +crop_size = (480, 480) + +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=img_scale, ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=img_scale, + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=4, + workers_per_gpu=4, + train=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClassContext', + split='ImageSets/SegmentationContext/train.txt', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClassContext', + split='ImageSets/SegmentationContext/val.txt', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClassContext', + split='ImageSets/SegmentationContext/val.txt', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_context_59.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_context_59.py new file mode 100644 index 0000000000000000000000000000000000000000..37585abab89834b95cd5bdd993b994fca1db65f6 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_context_59.py @@ -0,0 +1,60 @@ +# dataset settings +dataset_type = 'PascalContextDataset59' +data_root = 'data/VOCdevkit/VOC2010/' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) + +img_scale = (520, 520) +crop_size = (480, 480) + +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations', reduce_zero_label=True), + dict(type='Resize', img_scale=img_scale, ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=img_scale, + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=4, + workers_per_gpu=4, + train=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClassContext', + split='ImageSets/SegmentationContext/train.txt', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClassContext', + split='ImageSets/SegmentationContext/val.txt', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClassContext', + split='ImageSets/SegmentationContext/val.txt', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_voc12.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_voc12.py new file mode 100644 index 0000000000000000000000000000000000000000..e5ff704ae06d25f13a03d3dbe83d578ac9c89bad --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_voc12.py @@ -0,0 +1,58 @@ +# dataset settings + +dataset_type = 'PascalVOCDataset' +data_root = 'data/VOCdevkit/VOC2012' +img_norm_cfg = dict( + mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) +crop_size = (512, 512) +train_pipeline = [ + dict(type='LoadImageFromFile'), + dict(type='LoadAnnotations'), + dict(type='Resize', img_scale=(2048, 512), ratio_range=(0.5, 2.0)), + dict(type='RandomCrop', crop_size=crop_size, cat_max_ratio=0.75), + dict(type='RandomFlip', prob=0.5), + dict(type='PhotoMetricDistortion'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size=crop_size, pad_val=0, seg_pad_val=255), + dict(type='DefaultFormatBundle'), + dict(type='Collect', keys=['img', 'gt_semantic_seg']), +] +test_pipeline = [ + dict(type='LoadImageFromFile'), + dict( + type='MultiScaleFlipAug', + img_scale=(2048, 512), + # img_ratios=[0.5, 0.75, 1.0, 1.25, 1.5, 1.75], + flip=False, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ]) +] +data = dict( + samples_per_gpu=4, + workers_per_gpu=4, + train=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClass', + split='ImageSets/Segmentation/train.txt', + pipeline=train_pipeline), + val=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClass', + split='ImageSets/Segmentation/val.txt', + pipeline=test_pipeline), + test=dict( + type=dataset_type, + data_root=data_root, + img_dir='JPEGImages', + ann_dir='SegmentationClass', + split='ImageSets/Segmentation/val.txt', + pipeline=test_pipeline)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_voc12_aug.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_voc12_aug.py new file mode 100644 index 0000000000000000000000000000000000000000..3f23b6717d53ad29f02dd15046802a2631a5076b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/datasets/pascal_voc12_aug.py @@ -0,0 +1,9 @@ +_base_ = './pascal_voc12.py' +# dataset settings +data = dict( + train=dict( + ann_dir=['SegmentationClass', 'SegmentationClassAug'], + split=[ + 'ImageSets/Segmentation/train.txt', + 'ImageSets/Segmentation/aug.txt' + ])) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/default_runtime.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/default_runtime.py new file mode 100644 index 0000000000000000000000000000000000000000..b564cc4e7e7d9a67dacaaddecb100e4d8f5c005b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/default_runtime.py @@ -0,0 +1,14 @@ +# yapf:disable +log_config = dict( + interval=50, + hooks=[ + dict(type='TextLoggerHook', by_epoch=False), + # dict(type='TensorboardLoggerHook') + ]) +# yapf:enable +dist_params = dict(backend='nccl') +log_level = 'INFO' +load_from = None +resume_from = None +workflow = [('train', 1)] +cudnn_benchmark = True diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/models/attunet_r34.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/models/attunet_r34.py new file mode 100644 index 0000000000000000000000000000000000000000..4870b1d54fb46d1e648bb386ab7f6c871830b4eb --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/models/attunet_r34.py @@ -0,0 +1,45 @@ +# Copyright (c) 2023, Shanghai Iluvatar CoreX Semiconductor Co., Ltd. +# All Rights Reserved. +# +# Licensed under the Apache License, Version 2.0 (the "License"); you may +# not use this file except in compliance with the License. You may obtain +# a copy of the License at +# +# http://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, WITHOUT +# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the +# License for the specific language governing permissions and limitations +# under the License. + +# model settings +norm_cfg = dict(type='SyncBN', requires_grad=True) +model = dict( + type='EncoderDecoder', + pretrained=None, + backbone=dict(type='ResNet', + depth=34), + decode_head=dict( + type='ATTUNetHead', + in_channels=[64, 128, 256, 512], + channels=64, + input_transform=None, + # dropout_ratio=None, + num_classes=19, + norm_cfg=norm_cfg, + align_corners=False, + loss_decode=[ + dict( + type='CrossEntropyLoss', + loss_name='loss_ce', + use_sigmoid=False, + loss_weight=1.0), + dict(type='DiceLoss', loss_name='loss_dice', loss_weight=1.0), + dict(type="FocalLoss") + ]), + + # model training and testing settings + train_cfg=dict(), + test_cfg=dict(mode='whole') +) \ No newline at end of file diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_160k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_160k.py new file mode 100644 index 0000000000000000000000000000000000000000..eae05127a644be559866075372c019a5c189121e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_160k.py @@ -0,0 +1,9 @@ +# optimizer +optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005) +optimizer_config = dict() +# learning policy +lr_config = dict(policy='poly', power=0.9, min_lr=1e-4, by_epoch=False) +# runtime settings +runner = dict(type='IterBasedRunner', max_iters=160000) +checkpoint_config = dict(by_epoch=False, interval=16000) +evaluation = dict(interval=4000, metric='mIoU', pre_eval=True) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_1k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_1k.py new file mode 100644 index 0000000000000000000000000000000000000000..04cf410304dcdad73d94ca36665004043f69ea32 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_1k.py @@ -0,0 +1,9 @@ +# optimizer +optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005) +optimizer_config = dict() +# learning policy +lr_config = dict(policy='poly', power=0.9, min_lr=1e-4, by_epoch=False) +# runtime settings +runner = dict(type='IterBasedRunner', max_iters=1000) +checkpoint_config = dict(by_epoch=False, interval=1000) +evaluation = dict(interval=1000, metric='mIoU', pre_eval=True) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_20k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_20k.py new file mode 100644 index 0000000000000000000000000000000000000000..73c7021972bb8e955440b5afc8eaf0a4853b98a7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_20k.py @@ -0,0 +1,9 @@ +# optimizer +optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005) +optimizer_config = dict() +# learning policy +lr_config = dict(policy='poly', power=0.9, min_lr=1e-4, by_epoch=False) +# runtime settings +runner = dict(type='IterBasedRunner', max_iters=20000) +checkpoint_config = dict(by_epoch=False, interval=2000) +evaluation = dict(interval=2000, metric='mIoU', pre_eval=True) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_320k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_320k.py new file mode 100644 index 0000000000000000000000000000000000000000..a0b230626f638fc2072dbc78cca834edce3fdc23 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_320k.py @@ -0,0 +1,9 @@ +# optimizer +optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005) +optimizer_config = dict() +# learning policy +lr_config = dict(policy='poly', power=0.9, min_lr=1e-4, by_epoch=False) +# runtime settings +runner = dict(type='IterBasedRunner', max_iters=320000) +checkpoint_config = dict(by_epoch=False, interval=32000) +evaluation = dict(interval=32000, metric='mIoU') diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_40k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_40k.py new file mode 100644 index 0000000000000000000000000000000000000000..d2c502325944a1c5aa894283f28c610d67ab4da8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_40k.py @@ -0,0 +1,9 @@ +# optimizer +optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005) +optimizer_config = dict() +# learning policy +lr_config = dict(policy='poly', power=0.9, min_lr=1e-4, by_epoch=False) +# runtime settings +runner = dict(type='IterBasedRunner', max_iters=40000) +checkpoint_config = dict(by_epoch=False, interval=4000) +evaluation = dict(interval=4000, metric='mIoU', pre_eval=True) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_80k.py b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_80k.py new file mode 100644 index 0000000000000000000000000000000000000000..8365a878e9e19fff1080b0268ee26405eea34e43 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/_base_/schedules/schedule_80k.py @@ -0,0 +1,9 @@ +# optimizer +optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0005) +optimizer_config = dict() +# learning policy +lr_config = dict(policy='poly', power=0.9, min_lr=1e-4, by_epoch=False) +# runtime settings +runner = dict(type='IterBasedRunner', max_iters=80000) +checkpoint_config = dict(by_epoch=False, interval=8000) +evaluation = dict(interval=8000, metric='mIoU', pre_eval=True) diff --git a/cv/semantic_segmentation/att_unet/pytorch/configs/att_unet/attunet_r34_512x1024_160k_cityscapes.py b/cv/semantic_segmentation/att_unet/pytorch/configs/att_unet/attunet_r34_512x1024_160k_cityscapes.py new file mode 100644 index 0000000000000000000000000000000000000000..f08096afe90c75a661e102c8449a077e4f033c9d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/configs/att_unet/attunet_r34_512x1024_160k_cityscapes.py @@ -0,0 +1,6 @@ +# Copyright (c) 2023, Shanghai Iluvatar CoreX Semiconductor Co., Ltd. + +_base_ = [ + '../_base_/models/attunet_r34.py', '../_base_/datasets/cityscapes.py', + '../_base_/default_runtime.py', '../_base_/schedules/schedule_160k.py' +] \ No newline at end of file diff --git a/cv/semantic_segmentation/att_unet/pytorch/docker/Dockerfile b/cv/semantic_segmentation/att_unet/pytorch/docker/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..64482b4725ed2c7a21cc280efb4f0f6559c82aed --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/docker/Dockerfile @@ -0,0 +1,32 @@ +ARG PYTORCH="1.11.0" +ARG CUDA="11.3" +ARG CUDNN="8" + +FROM pytorch/pytorch:${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel + +ENV TORCH_CUDA_ARCH_LIST="6.0 6.1 7.0+PTX" +ENV TORCH_NVCC_FLAGS="-Xfatbin -compress-all" +ENV CMAKE_PREFIX_PATH="$(dirname $(which conda))/../" + +# To fix GPG key error when running apt-get update +RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/3bf863cc.pub +RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1804/x86_64/7fa2af80.pub + +RUN apt-get update && apt-get install -y git ninja-build libglib2.0-0 libsm6 libxrender-dev libxext6 \ + && apt-get clean \ + && rm -rf /var/lib/apt/lists/* + +RUN conda clean --all + +# Install MMCV +ARG PYTORCH +ARG CUDA +ARG MMCV +RUN ["/bin/bash", "-c", "pip install --no-cache-dir mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu${CUDA//./}/torch${PYTORCH}/index.html"] + +# Install MMSegmentation +RUN git clone https://github.com/open-mmlab/mmsegmentation.git /mmsegmentation +WORKDIR /mmsegmentation +ENV FORCE_CUDA="1" +RUN pip install -r requirements.txt +RUN pip install --no-cache-dir -e . diff --git a/cv/semantic_segmentation/att_unet/pytorch/docker/serve/Dockerfile b/cv/semantic_segmentation/att_unet/pytorch/docker/serve/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..c1d154528ce7dc86d628012c765063edd93b7865 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/docker/serve/Dockerfile @@ -0,0 +1,49 @@ +ARG PYTORCH="1.11.0" +ARG CUDA="11.3" +ARG CUDNN="8" +FROM pytorch/pytorch:${PYTORCH}-cuda${CUDA}-cudnn${CUDNN}-devel + +ARG MMCV="1.4.8" +ARG MMSEG="0.24.1" + +ENV PYTHONUNBUFFERED TRUE + +RUN apt-get update && \ + DEBIAN_FRONTEND=noninteractive apt-get install --no-install-recommends -y \ + ca-certificates \ + g++ \ + openjdk-11-jre-headless \ + # MMDet Requirements + ffmpeg libsm6 libxext6 git ninja-build libglib2.0-0 libsm6 libxrender-dev libxext6 \ + && rm -rf /var/lib/apt/lists/* + +ENV PATH="/opt/conda/bin:$PATH" +RUN export FORCE_CUDA=1 + +# TORCHSEVER +RUN pip install torchserve torch-model-archiver + +# MMLAB +ARG PYTORCH +ARG CUDA +RUN ["/bin/bash", "-c", "pip install mmcv-full==${MMCV} -f https://download.openmmlab.com/mmcv/dist/cu${CUDA//./}/torch${PYTORCH}/index.html"] +RUN pip install mmsegmentation==${MMSEG} + +RUN useradd -m model-server \ + && mkdir -p /home/model-server/tmp + +COPY entrypoint.sh /usr/local/bin/entrypoint.sh + +RUN chmod +x /usr/local/bin/entrypoint.sh \ + && chown -R model-server /home/model-server + +COPY config.properties /home/model-server/config.properties +RUN mkdir /home/model-server/model-store && chown -R model-server /home/model-server/model-store + +EXPOSE 8080 8081 8082 + +USER model-server +WORKDIR /home/model-server +ENV TEMP=/home/model-server/tmp +ENTRYPOINT ["/usr/local/bin/entrypoint.sh"] +CMD ["serve"] diff --git a/cv/semantic_segmentation/att_unet/pytorch/docker/serve/config.properties b/cv/semantic_segmentation/att_unet/pytorch/docker/serve/config.properties new file mode 100644 index 0000000000000000000000000000000000000000..efb9c47e40ab550bac765611e6c6c6f2a7152f11 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/docker/serve/config.properties @@ -0,0 +1,5 @@ +inference_address=http://0.0.0.0:8080 +management_address=http://0.0.0.0:8081 +metrics_address=http://0.0.0.0:8082 +model_store=/home/model-server/model-store +load_models=all diff --git a/cv/semantic_segmentation/att_unet/pytorch/docker/serve/entrypoint.sh b/cv/semantic_segmentation/att_unet/pytorch/docker/serve/entrypoint.sh new file mode 100644 index 0000000000000000000000000000000000000000..41ba00b048aed84b45c5a8015a016ff148e97d86 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/docker/serve/entrypoint.sh @@ -0,0 +1,12 @@ +#!/bin/bash +set -e + +if [[ "$1" = "serve" ]]; then + shift 1 + torchserve --start --ts-config /home/model-server/config.properties +else + eval "$@" +fi + +# prevent docker exit +tail -f /dev/null diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..435429d4882755dce6ec3bea16b8c41358aea539 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/__init__.py @@ -0,0 +1,13 @@ +# Copyright (c) OpenMMLab. All rights reserved. +# flake8: noqa +from .fileio import * +from .image import * +from .utils import * +from .version import * + +# The following modules are not imported to this level, so mmcv may be used +# without PyTorch. +# - runner +# - parallel +# - op +# - device diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..3d5599d9a7c61d5ebfd69501637efed2b2fb5a00 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/__init__.py @@ -0,0 +1,21 @@ +# Copyright (c) OpenMMLab. All rights reserved. +# yapf: disable +from .bricks import (ACTIVATION_LAYERS, CONV_LAYERS, NORM_LAYERS, + PADDING_LAYERS, PLUGIN_LAYERS, UPSAMPLE_LAYERS, + ContextBlock, Conv2d, Conv3d, ConvAWS2d, ConvModule, + ConvTranspose2d, ConvTranspose3d, ConvWS2d, + DepthwiseSeparableConvModule, GeneralizedAttention, + HSigmoid, HSwish, Linear, MaxPool2d, MaxPool3d, + NonLocal1d, NonLocal2d, NonLocal3d, Scale, Swish, + build_activation_layer, build_conv_layer, + build_norm_layer, build_padding_layer, build_plugin_layer, + build_upsample_layer, conv_ws_2d, is_norm) +from .builder import MODELS, build_model_from_cfg +# yapf: enable +from .utils import (INITIALIZERS, Caffe2XavierInit, ConstantInit, KaimingInit, + NormalInit, PretrainedInit, TruncNormalInit, UniformInit, + XavierInit, bias_init_with_prob, caffe2_xavier_init, + constant_init, fuse_conv_bn, get_model_complexity_info, + initialize, kaiming_init, normal_init, trunc_normal_init, + uniform_init, xavier_init) + diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..0f33124ed23fc6f27119a37bcb5ab004d3572be0 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/__init__.py @@ -0,0 +1,35 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .activation import build_activation_layer +from .context_block import ContextBlock +from .conv import build_conv_layer +from .conv2d_adaptive_padding import Conv2dAdaptivePadding +from .conv_module import ConvModule +from .conv_ws import ConvAWS2d, ConvWS2d, conv_ws_2d +from .depthwise_separable_conv_module import DepthwiseSeparableConvModule +from .drop import Dropout, DropPath +from .generalized_attention import GeneralizedAttention +from .hsigmoid import HSigmoid +from .hswish import HSwish +from .non_local import NonLocal1d, NonLocal2d, NonLocal3d +from .norm import build_norm_layer, is_norm +from .padding import build_padding_layer +from .plugin import build_plugin_layer +from .registry import (ACTIVATION_LAYERS, CONV_LAYERS, NORM_LAYERS, + PADDING_LAYERS, PLUGIN_LAYERS, UPSAMPLE_LAYERS) +from .scale import Scale +from .swish import Swish +from .upsample import build_upsample_layer +from .wrappers import (Conv2d, Conv3d, ConvTranspose2d, ConvTranspose3d, + Linear, MaxPool2d, MaxPool3d) + +__all__ = [ + 'ConvModule', 'build_activation_layer', 'build_conv_layer', + 'build_norm_layer', 'build_padding_layer', 'build_upsample_layer', + 'build_plugin_layer', 'is_norm', 'HSigmoid', 'HSwish', 'NonLocal1d', + 'NonLocal2d', 'NonLocal3d', 'ContextBlock', 'GeneralizedAttention', + 'ACTIVATION_LAYERS', 'CONV_LAYERS', 'NORM_LAYERS', 'PADDING_LAYERS', + 'UPSAMPLE_LAYERS', 'PLUGIN_LAYERS', 'Scale', 'ConvAWS2d', 'ConvWS2d', + 'conv_ws_2d', 'DepthwiseSeparableConvModule', 'Swish', 'Linear', + 'Conv2dAdaptivePadding', 'Conv2d', 'ConvTranspose2d', 'MaxPool2d', + 'ConvTranspose3d', 'MaxPool3d', 'Conv3d', 'Dropout', 'DropPath' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/activation.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/activation.py new file mode 100644 index 0000000000000000000000000000000000000000..26be59581ca21b3d9542b0e77d2cfcaeef747940 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/activation.py @@ -0,0 +1,93 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +import torch.nn.functional as F + +from mmcv.utils import TORCH_VERSION, build_from_cfg, digit_version +from .registry import ACTIVATION_LAYERS + +for module in [ + nn.ReLU, nn.LeakyReLU, nn.PReLU, nn.RReLU, nn.ReLU6, nn.ELU, + nn.Sigmoid, nn.Tanh +]: + ACTIVATION_LAYERS.register_module(module=module) + + +@ACTIVATION_LAYERS.register_module(name='Clip') +@ACTIVATION_LAYERS.register_module() +class Clamp(nn.Module): + """Clamp activation layer. + + This activation function is to clamp the feature map value within + :math:`[min, max]`. More details can be found in ``torch.clamp()``. + + Args: + min (Number | optional): Lower-bound of the range to be clamped to. + Default to -1. + max (Number | optional): Upper-bound of the range to be clamped to. + Default to 1. + """ + + def __init__(self, min=-1., max=1.): + super(Clamp, self).__init__() + self.min = min + self.max = max + + def forward(self, x): + """Forward function. + + Args: + x (torch.Tensor): The input tensor. + + Returns: + torch.Tensor: Clamped tensor. + """ + return torch.clamp(x, min=self.min, max=self.max) + + +class GELU(nn.Module): + r"""Applies the Gaussian Error Linear Units function: + + .. math:: + \text{GELU}(x) = x * \Phi(x) + where :math:`\Phi(x)` is the Cumulative Distribution Function for + Gaussian Distribution. + + Shape: + - Input: :math:`(N, *)` where `*` means, any number of additional + dimensions + - Output: :math:`(N, *)`, same shape as the input + + .. image:: scripts/activation_images/GELU.png + + Examples:: + + >>> m = nn.GELU() + >>> input = torch.randn(2) + >>> output = m(input) + """ + + def forward(self, input): + return F.gelu(input) + + +if (TORCH_VERSION == 'parrots' + or digit_version(TORCH_VERSION) < digit_version('1.4')): + ACTIVATION_LAYERS.register_module(module=GELU) +else: + ACTIVATION_LAYERS.register_module(module=nn.GELU) + + +def build_activation_layer(cfg): + """Build activation layer. + + Args: + cfg (dict): The activation layer config, which should contain: + + - type (str): Layer type. + - layer args: Args needed to instantiate an activation layer. + + Returns: + nn.Module: Created activation layer. + """ + return build_from_cfg(cfg, ACTIVATION_LAYERS) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/context_block.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/context_block.py new file mode 100644 index 0000000000000000000000000000000000000000..d60fdb904c749ce3b251510dff3cc63cea70d42e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/context_block.py @@ -0,0 +1,125 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +from torch import nn + +from ..utils import constant_init, kaiming_init +from .registry import PLUGIN_LAYERS + + +def last_zero_init(m): + if isinstance(m, nn.Sequential): + constant_init(m[-1], val=0) + else: + constant_init(m, val=0) + + +@PLUGIN_LAYERS.register_module() +class ContextBlock(nn.Module): + """ContextBlock module in GCNet. + + See 'GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond' + (https://arxiv.org/abs/1904.11492) for details. + + Args: + in_channels (int): Channels of the input feature map. + ratio (float): Ratio of channels of transform bottleneck + pooling_type (str): Pooling method for context modeling. + Options are 'att' and 'avg', stand for attention pooling and + average pooling respectively. Default: 'att'. + fusion_types (Sequence[str]): Fusion method for feature fusion, + Options are 'channels_add', 'channel_mul', stand for channelwise + addition and multiplication respectively. Default: ('channel_add',) + """ + + _abbr_ = 'context_block' + + def __init__(self, + in_channels, + ratio, + pooling_type='att', + fusion_types=('channel_add', )): + super(ContextBlock, self).__init__() + assert pooling_type in ['avg', 'att'] + assert isinstance(fusion_types, (list, tuple)) + valid_fusion_types = ['channel_add', 'channel_mul'] + assert all([f in valid_fusion_types for f in fusion_types]) + assert len(fusion_types) > 0, 'at least one fusion should be used' + self.in_channels = in_channels + self.ratio = ratio + self.planes = int(in_channels * ratio) + self.pooling_type = pooling_type + self.fusion_types = fusion_types + if pooling_type == 'att': + self.conv_mask = nn.Conv2d(in_channels, 1, kernel_size=1) + self.softmax = nn.Softmax(dim=2) + else: + self.avg_pool = nn.AdaptiveAvgPool2d(1) + if 'channel_add' in fusion_types: + self.channel_add_conv = nn.Sequential( + nn.Conv2d(self.in_channels, self.planes, kernel_size=1), + nn.LayerNorm([self.planes, 1, 1]), + nn.ReLU(inplace=True), # yapf: disable + nn.Conv2d(self.planes, self.in_channels, kernel_size=1)) + else: + self.channel_add_conv = None + if 'channel_mul' in fusion_types: + self.channel_mul_conv = nn.Sequential( + nn.Conv2d(self.in_channels, self.planes, kernel_size=1), + nn.LayerNorm([self.planes, 1, 1]), + nn.ReLU(inplace=True), # yapf: disable + nn.Conv2d(self.planes, self.in_channels, kernel_size=1)) + else: + self.channel_mul_conv = None + self.reset_parameters() + + def reset_parameters(self): + if self.pooling_type == 'att': + kaiming_init(self.conv_mask, mode='fan_in') + self.conv_mask.inited = True + + if self.channel_add_conv is not None: + last_zero_init(self.channel_add_conv) + if self.channel_mul_conv is not None: + last_zero_init(self.channel_mul_conv) + + def spatial_pool(self, x): + batch, channel, height, width = x.size() + if self.pooling_type == 'att': + input_x = x + # [N, C, H * W] + input_x = input_x.view(batch, channel, height * width) + # [N, 1, C, H * W] + input_x = input_x.unsqueeze(1) + # [N, 1, H, W] + context_mask = self.conv_mask(x) + # [N, 1, H * W] + context_mask = context_mask.view(batch, 1, height * width) + # [N, 1, H * W] + context_mask = self.softmax(context_mask) + # [N, 1, H * W, 1] + context_mask = context_mask.unsqueeze(-1) + # [N, 1, C, 1] + context = torch.matmul(input_x, context_mask) + # [N, C, 1, 1] + context = context.view(batch, channel, 1, 1) + else: + # [N, C, 1, 1] + context = self.avg_pool(x) + + return context + + def forward(self, x): + # [N, C, 1, 1] + context = self.spatial_pool(x) + + out = x + if self.channel_mul_conv is not None: + # [N, C, 1, 1] + channel_mul_term = torch.sigmoid(self.channel_mul_conv(context)) + out = out * channel_mul_term + if self.channel_add_conv is not None: + # [N, C, 1, 1] + channel_add_term = self.channel_add_conv(context) + out = out + channel_add_term + + return out diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv.py new file mode 100644 index 0000000000000000000000000000000000000000..f6c35fd70b3c6ae0812432e363adbc5711e71b18 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv.py @@ -0,0 +1,44 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from torch import nn + +from .registry import CONV_LAYERS + +CONV_LAYERS.register_module('Conv1d', module=nn.Conv1d) +CONV_LAYERS.register_module('Conv2d', module=nn.Conv2d) +CONV_LAYERS.register_module('Conv3d', module=nn.Conv3d) +CONV_LAYERS.register_module('Conv', module=nn.Conv2d) + + +def build_conv_layer(cfg, *args, **kwargs): + """Build convolution layer. + + Args: + cfg (None or dict): The conv layer config, which should contain: + - type (str): Layer type. + - layer args: Args needed to instantiate an conv layer. + args (argument list): Arguments passed to the `__init__` + method of the corresponding conv layer. + kwargs (keyword arguments): Keyword arguments passed to the `__init__` + method of the corresponding conv layer. + + Returns: + nn.Module: Created conv layer. + """ + if cfg is None: + cfg_ = dict(type='Conv2d') + else: + if not isinstance(cfg, dict): + raise TypeError('cfg must be a dict') + if 'type' not in cfg: + raise KeyError('the cfg dict must contain the key "type"') + cfg_ = cfg.copy() + + layer_type = cfg_.pop('type') + if layer_type not in CONV_LAYERS: + raise KeyError(f'Unrecognized layer type {layer_type}') + else: + conv_layer = CONV_LAYERS.get(layer_type) + + layer = conv_layer(*args, **kwargs, **cfg_) + + return layer diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv2d_adaptive_padding.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv2d_adaptive_padding.py new file mode 100644 index 0000000000000000000000000000000000000000..b45e758ac6cf8dfb0382d072fe09125bc7e9b888 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv2d_adaptive_padding.py @@ -0,0 +1,62 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import math + +from torch import nn +from torch.nn import functional as F + +from .registry import CONV_LAYERS + + +@CONV_LAYERS.register_module() +class Conv2dAdaptivePadding(nn.Conv2d): + """Implementation of 2D convolution in tensorflow with `padding` as "same", + which applies padding to input (if needed) so that input image gets fully + covered by filter and stride you specified. For stride 1, this will ensure + that output image size is same as input. For stride of 2, output dimensions + will be half, for example. + + Args: + in_channels (int): Number of channels in the input image + out_channels (int): Number of channels produced by the convolution + kernel_size (int or tuple): Size of the convolving kernel + stride (int or tuple, optional): Stride of the convolution. Default: 1 + padding (int or tuple, optional): Zero-padding added to both sides of + the input. Default: 0 + dilation (int or tuple, optional): Spacing between kernel elements. + Default: 1 + groups (int, optional): Number of blocked connections from input + channels to output channels. Default: 1 + bias (bool, optional): If ``True``, adds a learnable bias to the + output. Default: ``True`` + """ + + def __init__(self, + in_channels, + out_channels, + kernel_size, + stride=1, + padding=0, + dilation=1, + groups=1, + bias=True): + super().__init__(in_channels, out_channels, kernel_size, stride, 0, + dilation, groups, bias) + + def forward(self, x): + img_h, img_w = x.size()[-2:] + kernel_h, kernel_w = self.weight.size()[-2:] + stride_h, stride_w = self.stride + output_h = math.ceil(img_h / stride_h) + output_w = math.ceil(img_w / stride_w) + pad_h = ( + max((output_h - 1) * self.stride[0] + + (kernel_h - 1) * self.dilation[0] + 1 - img_h, 0)) + pad_w = ( + max((output_w - 1) * self.stride[1] + + (kernel_w - 1) * self.dilation[1] + 1 - img_w, 0)) + if pad_h > 0 or pad_w > 0: + x = F.pad(x, [ + pad_w // 2, pad_w - pad_w // 2, pad_h // 2, pad_h - pad_h // 2 + ]) + return F.conv2d(x, self.weight, self.bias, self.stride, self.padding, + self.dilation, self.groups) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv_module.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv_module.py new file mode 100644 index 0000000000000000000000000000000000000000..0078647a1f2aac61918f17fabb1aca2a50a7a08c --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv_module.py @@ -0,0 +1,206 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch.nn as nn + +from mmcv.utils import _BatchNorm, _InstanceNorm +from ..utils import constant_init, kaiming_init +from .activation import build_activation_layer +from .conv import build_conv_layer +from .norm import build_norm_layer +from .padding import build_padding_layer +from .registry import PLUGIN_LAYERS + + +@PLUGIN_LAYERS.register_module() +class ConvModule(nn.Module): + """A conv block that bundles conv/norm/activation layers. + + This block simplifies the usage of convolution layers, which are commonly + used with a norm layer (e.g., BatchNorm) and activation layer (e.g., ReLU). + It is based upon three build methods: `build_conv_layer()`, + `build_norm_layer()` and `build_activation_layer()`. + + Besides, we add some additional features in this module. + 1. Automatically set `bias` of the conv layer. + 2. Spectral norm is supported. + 3. More padding modes are supported. Before PyTorch 1.5, nn.Conv2d only + supports zero and circular padding, and we add "reflect" padding mode. + + Args: + in_channels (int): Number of channels in the input feature map. + Same as that in ``nn._ConvNd``. + out_channels (int): Number of channels produced by the convolution. + Same as that in ``nn._ConvNd``. + kernel_size (int | tuple[int]): Size of the convolving kernel. + Same as that in ``nn._ConvNd``. + stride (int | tuple[int]): Stride of the convolution. + Same as that in ``nn._ConvNd``. + padding (int | tuple[int]): Zero-padding added to both sides of + the input. Same as that in ``nn._ConvNd``. + dilation (int | tuple[int]): Spacing between kernel elements. + Same as that in ``nn._ConvNd``. + groups (int): Number of blocked connections from input channels to + output channels. Same as that in ``nn._ConvNd``. + bias (bool | str): If specified as `auto`, it will be decided by the + norm_cfg. Bias will be set as True if `norm_cfg` is None, otherwise + False. Default: "auto". + conv_cfg (dict): Config dict for convolution layer. Default: None, + which means using conv2d. + norm_cfg (dict): Config dict for normalization layer. Default: None. + act_cfg (dict): Config dict for activation layer. + Default: dict(type='ReLU'). + inplace (bool): Whether to use inplace mode for activation. + Default: True. + with_spectral_norm (bool): Whether use spectral norm in conv module. + Default: False. + padding_mode (str): If the `padding_mode` has not been supported by + current `Conv2d` in PyTorch, we will use our own padding layer + instead. Currently, we support ['zeros', 'circular'] with official + implementation and ['reflect'] with our own implementation. + Default: 'zeros'. + order (tuple[str]): The order of conv/norm/activation layers. It is a + sequence of "conv", "norm" and "act". Common examples are + ("conv", "norm", "act") and ("act", "conv", "norm"). + Default: ('conv', 'norm', 'act'). + """ + + _abbr_ = 'conv_block' + + def __init__(self, + in_channels, + out_channels, + kernel_size, + stride=1, + padding=0, + dilation=1, + groups=1, + bias='auto', + conv_cfg=None, + norm_cfg=None, + act_cfg=dict(type='ReLU'), + inplace=True, + with_spectral_norm=False, + padding_mode='zeros', + order=('conv', 'norm', 'act')): + super(ConvModule, self).__init__() + assert conv_cfg is None or isinstance(conv_cfg, dict) + assert norm_cfg is None or isinstance(norm_cfg, dict) + assert act_cfg is None or isinstance(act_cfg, dict) + official_padding_mode = ['zeros', 'circular'] + self.conv_cfg = conv_cfg + self.norm_cfg = norm_cfg + self.act_cfg = act_cfg + self.inplace = inplace + self.with_spectral_norm = with_spectral_norm + self.with_explicit_padding = padding_mode not in official_padding_mode + self.order = order + assert isinstance(self.order, tuple) and len(self.order) == 3 + assert set(order) == set(['conv', 'norm', 'act']) + + self.with_norm = norm_cfg is not None + self.with_activation = act_cfg is not None + # if the conv layer is before a norm layer, bias is unnecessary. + if bias == 'auto': + bias = not self.with_norm + self.with_bias = bias + + if self.with_explicit_padding: + pad_cfg = dict(type=padding_mode) + self.padding_layer = build_padding_layer(pad_cfg, padding) + + # reset padding to 0 for conv module + conv_padding = 0 if self.with_explicit_padding else padding + # build convolution layer + self.conv = build_conv_layer( + conv_cfg, + in_channels, + out_channels, + kernel_size, + stride=stride, + padding=conv_padding, + dilation=dilation, + groups=groups, + bias=bias) + # export the attributes of self.conv to a higher level for convenience + self.in_channels = self.conv.in_channels + self.out_channels = self.conv.out_channels + self.kernel_size = self.conv.kernel_size + self.stride = self.conv.stride + self.padding = padding + self.dilation = self.conv.dilation + self.transposed = self.conv.transposed + self.output_padding = self.conv.output_padding + self.groups = self.conv.groups + + if self.with_spectral_norm: + self.conv = nn.utils.spectral_norm(self.conv) + + # build normalization layers + if self.with_norm: + # norm layer is after conv layer + if order.index('norm') > order.index('conv'): + norm_channels = out_channels + else: + norm_channels = in_channels + self.norm_name, norm = build_norm_layer(norm_cfg, norm_channels) + self.add_module(self.norm_name, norm) + if self.with_bias: + if isinstance(norm, (_BatchNorm, _InstanceNorm)): + warnings.warn( + 'Unnecessary conv bias before batch/instance norm') + else: + self.norm_name = None + + # build activation layer + if self.with_activation: + act_cfg_ = act_cfg.copy() + # nn.Tanh has no 'inplace' argument + if act_cfg_['type'] not in [ + 'Tanh', 'PReLU', 'Sigmoid', 'HSigmoid', 'Swish', 'GELU' + ]: + act_cfg_.setdefault('inplace', inplace) + self.activate = build_activation_layer(act_cfg_) + + # Use msra init by default + self.init_weights() + + @property + def norm(self): + if self.norm_name: + return getattr(self, self.norm_name) + else: + return None + + def init_weights(self): + # 1. It is mainly for customized conv layers with their own + # initialization manners by calling their own ``init_weights()``, + # and we do not want ConvModule to override the initialization. + # 2. For customized conv layers without their own initialization + # manners (that is, they don't have their own ``init_weights()``) + # and PyTorch's conv layers, they will be initialized by + # this method with default ``kaiming_init``. + # Note: For PyTorch's conv layers, they will be overwritten by our + # initialization implementation using default ``kaiming_init``. + if not hasattr(self.conv, 'init_weights'): + if self.with_activation and self.act_cfg['type'] == 'LeakyReLU': + nonlinearity = 'leaky_relu' + a = self.act_cfg.get('negative_slope', 0.01) + else: + nonlinearity = 'relu' + a = 0 + kaiming_init(self.conv, a=a, nonlinearity=nonlinearity) + if self.with_norm: + constant_init(self.norm, 1, bias=0) + + def forward(self, x, activate=True, norm=True): + for layer in self.order: + if layer == 'conv': + if self.with_explicit_padding: + x = self.padding_layer(x) + x = self.conv(x) + elif layer == 'norm' and norm and self.with_norm: + x = self.norm(x) + elif layer == 'act' and activate and self.with_activation: + x = self.activate(x) + return x diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv_ws.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv_ws.py new file mode 100644 index 0000000000000000000000000000000000000000..a3941e27874993418b3b5708d5a7485f175ff9c8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/conv_ws.py @@ -0,0 +1,148 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +import torch.nn.functional as F + +from .registry import CONV_LAYERS + + +def conv_ws_2d(input, + weight, + bias=None, + stride=1, + padding=0, + dilation=1, + groups=1, + eps=1e-5): + c_in = weight.size(0) + weight_flat = weight.view(c_in, -1) + mean = weight_flat.mean(dim=1, keepdim=True).view(c_in, 1, 1, 1) + std = weight_flat.std(dim=1, keepdim=True).view(c_in, 1, 1, 1) + weight = (weight - mean) / (std + eps) + return F.conv2d(input, weight, bias, stride, padding, dilation, groups) + + +@CONV_LAYERS.register_module('ConvWS') +class ConvWS2d(nn.Conv2d): + + def __init__(self, + in_channels, + out_channels, + kernel_size, + stride=1, + padding=0, + dilation=1, + groups=1, + bias=True, + eps=1e-5): + super(ConvWS2d, self).__init__( + in_channels, + out_channels, + kernel_size, + stride=stride, + padding=padding, + dilation=dilation, + groups=groups, + bias=bias) + self.eps = eps + + def forward(self, x): + return conv_ws_2d(x, self.weight, self.bias, self.stride, self.padding, + self.dilation, self.groups, self.eps) + + +@CONV_LAYERS.register_module(name='ConvAWS') +class ConvAWS2d(nn.Conv2d): + """AWS (Adaptive Weight Standardization) + + This is a variant of Weight Standardization + (https://arxiv.org/pdf/1903.10520.pdf) + It is used in DetectoRS to avoid NaN + (https://arxiv.org/pdf/2006.02334.pdf) + + Args: + in_channels (int): Number of channels in the input image + out_channels (int): Number of channels produced by the convolution + kernel_size (int or tuple): Size of the conv kernel + stride (int or tuple, optional): Stride of the convolution. Default: 1 + padding (int or tuple, optional): Zero-padding added to both sides of + the input. Default: 0 + dilation (int or tuple, optional): Spacing between kernel elements. + Default: 1 + groups (int, optional): Number of blocked connections from input + channels to output channels. Default: 1 + bias (bool, optional): If set True, adds a learnable bias to the + output. Default: True + """ + + def __init__(self, + in_channels, + out_channels, + kernel_size, + stride=1, + padding=0, + dilation=1, + groups=1, + bias=True): + super().__init__( + in_channels, + out_channels, + kernel_size, + stride=stride, + padding=padding, + dilation=dilation, + groups=groups, + bias=bias) + self.register_buffer('weight_gamma', + torch.ones(self.out_channels, 1, 1, 1)) + self.register_buffer('weight_beta', + torch.zeros(self.out_channels, 1, 1, 1)) + + def _get_weight(self, weight): + weight_flat = weight.view(weight.size(0), -1) + mean = weight_flat.mean(dim=1).view(-1, 1, 1, 1) + std = torch.sqrt(weight_flat.var(dim=1) + 1e-5).view(-1, 1, 1, 1) + weight = (weight - mean) / std + weight = self.weight_gamma * weight + self.weight_beta + return weight + + def forward(self, x): + weight = self._get_weight(self.weight) + return F.conv2d(x, weight, self.bias, self.stride, self.padding, + self.dilation, self.groups) + + def _load_from_state_dict(self, state_dict, prefix, local_metadata, strict, + missing_keys, unexpected_keys, error_msgs): + """Override default load function. + + AWS overrides the function _load_from_state_dict to recover + weight_gamma and weight_beta if they are missing. If weight_gamma and + weight_beta are found in the checkpoint, this function will return + after super()._load_from_state_dict. Otherwise, it will compute the + mean and std of the pretrained weights and store them in weight_beta + and weight_gamma. + """ + + self.weight_gamma.data.fill_(-1) + local_missing_keys = [] + super()._load_from_state_dict(state_dict, prefix, local_metadata, + strict, local_missing_keys, + unexpected_keys, error_msgs) + if self.weight_gamma.data.mean() > 0: + for k in local_missing_keys: + missing_keys.append(k) + return + weight = self.weight.data + weight_flat = weight.view(weight.size(0), -1) + mean = weight_flat.mean(dim=1).view(-1, 1, 1, 1) + std = torch.sqrt(weight_flat.var(dim=1) + 1e-5).view(-1, 1, 1, 1) + self.weight_beta.data.copy_(mean) + self.weight_gamma.data.copy_(std) + missing_gamma_beta = [ + k for k in local_missing_keys + if k.endswith('weight_gamma') or k.endswith('weight_beta') + ] + for k in missing_gamma_beta: + local_missing_keys.remove(k) + for k in local_missing_keys: + missing_keys.append(k) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/depthwise_separable_conv_module.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/depthwise_separable_conv_module.py new file mode 100644 index 0000000000000000000000000000000000000000..722d5d8d71f75486e2db3008907c4eadfca41d63 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/depthwise_separable_conv_module.py @@ -0,0 +1,96 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn + +from .conv_module import ConvModule + + +class DepthwiseSeparableConvModule(nn.Module): + """Depthwise separable convolution module. + + See https://arxiv.org/pdf/1704.04861.pdf for details. + + This module can replace a ConvModule with the conv block replaced by two + conv block: depthwise conv block and pointwise conv block. The depthwise + conv block contains depthwise-conv/norm/activation layers. The pointwise + conv block contains pointwise-conv/norm/activation layers. It should be + noted that there will be norm/activation layer in the depthwise conv block + if `norm_cfg` and `act_cfg` are specified. + + Args: + in_channels (int): Number of channels in the input feature map. + Same as that in ``nn._ConvNd``. + out_channels (int): Number of channels produced by the convolution. + Same as that in ``nn._ConvNd``. + kernel_size (int | tuple[int]): Size of the convolving kernel. + Same as that in ``nn._ConvNd``. + stride (int | tuple[int]): Stride of the convolution. + Same as that in ``nn._ConvNd``. Default: 1. + padding (int | tuple[int]): Zero-padding added to both sides of + the input. Same as that in ``nn._ConvNd``. Default: 0. + dilation (int | tuple[int]): Spacing between kernel elements. + Same as that in ``nn._ConvNd``. Default: 1. + norm_cfg (dict): Default norm config for both depthwise ConvModule and + pointwise ConvModule. Default: None. + act_cfg (dict): Default activation config for both depthwise ConvModule + and pointwise ConvModule. Default: dict(type='ReLU'). + dw_norm_cfg (dict): Norm config of depthwise ConvModule. If it is + 'default', it will be the same as `norm_cfg`. Default: 'default'. + dw_act_cfg (dict): Activation config of depthwise ConvModule. If it is + 'default', it will be the same as `act_cfg`. Default: 'default'. + pw_norm_cfg (dict): Norm config of pointwise ConvModule. If it is + 'default', it will be the same as `norm_cfg`. Default: 'default'. + pw_act_cfg (dict): Activation config of pointwise ConvModule. If it is + 'default', it will be the same as `act_cfg`. Default: 'default'. + kwargs (optional): Other shared arguments for depthwise and pointwise + ConvModule. See ConvModule for ref. + """ + + def __init__(self, + in_channels, + out_channels, + kernel_size, + stride=1, + padding=0, + dilation=1, + norm_cfg=None, + act_cfg=dict(type='ReLU'), + dw_norm_cfg='default', + dw_act_cfg='default', + pw_norm_cfg='default', + pw_act_cfg='default', + **kwargs): + super(DepthwiseSeparableConvModule, self).__init__() + assert 'groups' not in kwargs, 'groups should not be specified' + + # if norm/activation config of depthwise/pointwise ConvModule is not + # specified, use default config. + dw_norm_cfg = dw_norm_cfg if dw_norm_cfg != 'default' else norm_cfg + dw_act_cfg = dw_act_cfg if dw_act_cfg != 'default' else act_cfg + pw_norm_cfg = pw_norm_cfg if pw_norm_cfg != 'default' else norm_cfg + pw_act_cfg = pw_act_cfg if pw_act_cfg != 'default' else act_cfg + + # depthwise convolution + self.depthwise_conv = ConvModule( + in_channels, + in_channels, + kernel_size, + stride=stride, + padding=padding, + dilation=dilation, + groups=in_channels, + norm_cfg=dw_norm_cfg, + act_cfg=dw_act_cfg, + **kwargs) + + self.pointwise_conv = ConvModule( + in_channels, + out_channels, + 1, + norm_cfg=pw_norm_cfg, + act_cfg=pw_act_cfg, + **kwargs) + + def forward(self, x): + x = self.depthwise_conv(x) + x = self.pointwise_conv(x) + return x diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/drop.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/drop.py new file mode 100644 index 0000000000000000000000000000000000000000..b0a026654ac2e3b994eb7a5248ca9faa277f8985 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/drop.py @@ -0,0 +1,65 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn + +from mmcv import build_from_cfg +from .registry import DROPOUT_LAYERS + + +def drop_path(x, drop_prob=0., training=False): + """Drop paths (Stochastic Depth) per sample (when applied in main path of + residual blocks). + + We follow the implementation + https://github.com/rwightman/pytorch-image-models/blob/a2727c1bf78ba0d7b5727f5f95e37fb7f8866b1f/timm/models/layers/drop.py # noqa: E501 + """ + if drop_prob == 0. or not training: + return x + keep_prob = 1 - drop_prob + # handle tensors with different dimensions, not just 4D tensors. + shape = (x.shape[0], ) + (1, ) * (x.ndim - 1) + random_tensor = keep_prob + torch.rand( + shape, dtype=x.dtype, device=x.device) + output = x.div(keep_prob) * random_tensor.floor() + return output + + +@DROPOUT_LAYERS.register_module() +class DropPath(nn.Module): + """Drop paths (Stochastic Depth) per sample (when applied in main path of + residual blocks). + + We follow the implementation + https://github.com/rwightman/pytorch-image-models/blob/a2727c1bf78ba0d7b5727f5f95e37fb7f8866b1f/timm/models/layers/drop.py # noqa: E501 + + Args: + drop_prob (float): Probability of the path to be zeroed. Default: 0.1 + """ + + def __init__(self, drop_prob=0.1): + super(DropPath, self).__init__() + self.drop_prob = drop_prob + + def forward(self, x): + return drop_path(x, self.drop_prob, self.training) + + +@DROPOUT_LAYERS.register_module() +class Dropout(nn.Dropout): + """A wrapper for ``torch.nn.Dropout``, We rename the ``p`` of + ``torch.nn.Dropout`` to ``drop_prob`` so as to be consistent with + ``DropPath`` + + Args: + drop_prob (float): Probability of the elements to be + zeroed. Default: 0.5. + inplace (bool): Do the operation inplace or not. Default: False. + """ + + def __init__(self, drop_prob=0.5, inplace=False): + super().__init__(p=drop_prob, inplace=inplace) + + +def build_dropout(cfg, default_args=None): + """Builder for drop out layers.""" + return build_from_cfg(cfg, DROPOUT_LAYERS, default_args) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/generalized_attention.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/generalized_attention.py new file mode 100644 index 0000000000000000000000000000000000000000..c8a74d2682f90af4d5665bd284838ebb1d068fb9 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/generalized_attention.py @@ -0,0 +1,412 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import math + +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F + +from ..utils import kaiming_init +from .registry import PLUGIN_LAYERS + + +@PLUGIN_LAYERS.register_module() +class GeneralizedAttention(nn.Module): + """GeneralizedAttention module. + + See 'An Empirical Study of Spatial Attention Mechanisms in Deep Networks' + (https://arxiv.org/abs/1711.07971) for details. + + Args: + in_channels (int): Channels of the input feature map. + spatial_range (int): The spatial range. -1 indicates no spatial range + constraint. Default: -1. + num_heads (int): The head number of empirical_attention module. + Default: 9. + position_embedding_dim (int): The position embedding dimension. + Default: -1. + position_magnitude (int): A multiplier acting on coord difference. + Default: 1. + kv_stride (int): The feature stride acting on key/value feature map. + Default: 2. + q_stride (int): The feature stride acting on query feature map. + Default: 1. + attention_type (str): A binary indicator string for indicating which + items in generalized empirical_attention module are used. + Default: '1111'. + + - '1000' indicates 'query and key content' (appr - appr) item, + - '0100' indicates 'query content and relative position' + (appr - position) item, + - '0010' indicates 'key content only' (bias - appr) item, + - '0001' indicates 'relative position only' (bias - position) item. + """ + + _abbr_ = 'gen_attention_block' + + def __init__(self, + in_channels, + spatial_range=-1, + num_heads=9, + position_embedding_dim=-1, + position_magnitude=1, + kv_stride=2, + q_stride=1, + attention_type='1111'): + + super(GeneralizedAttention, self).__init__() + + # hard range means local range for non-local operation + self.position_embedding_dim = ( + position_embedding_dim + if position_embedding_dim > 0 else in_channels) + + self.position_magnitude = position_magnitude + self.num_heads = num_heads + self.in_channels = in_channels + self.spatial_range = spatial_range + self.kv_stride = kv_stride + self.q_stride = q_stride + self.attention_type = [bool(int(_)) for _ in attention_type] + self.qk_embed_dim = in_channels // num_heads + out_c = self.qk_embed_dim * num_heads + + if self.attention_type[0] or self.attention_type[1]: + self.query_conv = nn.Conv2d( + in_channels=in_channels, + out_channels=out_c, + kernel_size=1, + bias=False) + self.query_conv.kaiming_init = True + + if self.attention_type[0] or self.attention_type[2]: + self.key_conv = nn.Conv2d( + in_channels=in_channels, + out_channels=out_c, + kernel_size=1, + bias=False) + self.key_conv.kaiming_init = True + + self.v_dim = in_channels // num_heads + self.value_conv = nn.Conv2d( + in_channels=in_channels, + out_channels=self.v_dim * num_heads, + kernel_size=1, + bias=False) + self.value_conv.kaiming_init = True + + if self.attention_type[1] or self.attention_type[3]: + self.appr_geom_fc_x = nn.Linear( + self.position_embedding_dim // 2, out_c, bias=False) + self.appr_geom_fc_x.kaiming_init = True + + self.appr_geom_fc_y = nn.Linear( + self.position_embedding_dim // 2, out_c, bias=False) + self.appr_geom_fc_y.kaiming_init = True + + if self.attention_type[2]: + stdv = 1.0 / math.sqrt(self.qk_embed_dim * 2) + appr_bias_value = -2 * stdv * torch.rand(out_c) + stdv + self.appr_bias = nn.Parameter(appr_bias_value) + + if self.attention_type[3]: + stdv = 1.0 / math.sqrt(self.qk_embed_dim * 2) + geom_bias_value = -2 * stdv * torch.rand(out_c) + stdv + self.geom_bias = nn.Parameter(geom_bias_value) + + self.proj_conv = nn.Conv2d( + in_channels=self.v_dim * num_heads, + out_channels=in_channels, + kernel_size=1, + bias=True) + self.proj_conv.kaiming_init = True + self.gamma = nn.Parameter(torch.zeros(1)) + + if self.spatial_range >= 0: + # only works when non local is after 3*3 conv + if in_channels == 256: + max_len = 84 + elif in_channels == 512: + max_len = 42 + + max_len_kv = int((max_len - 1.0) / self.kv_stride + 1) + local_constraint_map = np.ones( + (max_len, max_len, max_len_kv, max_len_kv), dtype=int) + for iy in range(max_len): + for ix in range(max_len): + local_constraint_map[ + iy, ix, + max((iy - self.spatial_range) // + self.kv_stride, 0):min((iy + self.spatial_range + + 1) // self.kv_stride + + 1, max_len), + max((ix - self.spatial_range) // + self.kv_stride, 0):min((ix + self.spatial_range + + 1) // self.kv_stride + + 1, max_len)] = 0 + + self.local_constraint_map = nn.Parameter( + torch.from_numpy(local_constraint_map).byte(), + requires_grad=False) + + if self.q_stride > 1: + self.q_downsample = nn.AvgPool2d( + kernel_size=1, stride=self.q_stride) + else: + self.q_downsample = None + + if self.kv_stride > 1: + self.kv_downsample = nn.AvgPool2d( + kernel_size=1, stride=self.kv_stride) + else: + self.kv_downsample = None + + self.init_weights() + + def get_position_embedding(self, + h, + w, + h_kv, + w_kv, + q_stride, + kv_stride, + device, + dtype, + feat_dim, + wave_length=1000): + # the default type of Tensor is float32, leading to type mismatch + # in fp16 mode. Cast it to support fp16 mode. + h_idxs = torch.linspace(0, h - 1, h).to(device=device, dtype=dtype) + h_idxs = h_idxs.view((h, 1)) * q_stride + + w_idxs = torch.linspace(0, w - 1, w).to(device=device, dtype=dtype) + w_idxs = w_idxs.view((w, 1)) * q_stride + + h_kv_idxs = torch.linspace(0, h_kv - 1, h_kv).to( + device=device, dtype=dtype) + h_kv_idxs = h_kv_idxs.view((h_kv, 1)) * kv_stride + + w_kv_idxs = torch.linspace(0, w_kv - 1, w_kv).to( + device=device, dtype=dtype) + w_kv_idxs = w_kv_idxs.view((w_kv, 1)) * kv_stride + + # (h, h_kv, 1) + h_diff = h_idxs.unsqueeze(1) - h_kv_idxs.unsqueeze(0) + h_diff *= self.position_magnitude + + # (w, w_kv, 1) + w_diff = w_idxs.unsqueeze(1) - w_kv_idxs.unsqueeze(0) + w_diff *= self.position_magnitude + + feat_range = torch.arange(0, feat_dim / 4).to( + device=device, dtype=dtype) + + dim_mat = torch.Tensor([wave_length]).to(device=device, dtype=dtype) + dim_mat = dim_mat**((4. / feat_dim) * feat_range) + dim_mat = dim_mat.view((1, 1, -1)) + + embedding_x = torch.cat( + ((w_diff / dim_mat).sin(), (w_diff / dim_mat).cos()), dim=2) + + embedding_y = torch.cat( + ((h_diff / dim_mat).sin(), (h_diff / dim_mat).cos()), dim=2) + + return embedding_x, embedding_y + + def forward(self, x_input): + num_heads = self.num_heads + + # use empirical_attention + if self.q_downsample is not None: + x_q = self.q_downsample(x_input) + else: + x_q = x_input + n, _, h, w = x_q.shape + + if self.kv_downsample is not None: + x_kv = self.kv_downsample(x_input) + else: + x_kv = x_input + _, _, h_kv, w_kv = x_kv.shape + + if self.attention_type[0] or self.attention_type[1]: + proj_query = self.query_conv(x_q).view( + (n, num_heads, self.qk_embed_dim, h * w)) + proj_query = proj_query.permute(0, 1, 3, 2) + + if self.attention_type[0] or self.attention_type[2]: + proj_key = self.key_conv(x_kv).view( + (n, num_heads, self.qk_embed_dim, h_kv * w_kv)) + + if self.attention_type[1] or self.attention_type[3]: + position_embed_x, position_embed_y = self.get_position_embedding( + h, w, h_kv, w_kv, self.q_stride, self.kv_stride, + x_input.device, x_input.dtype, self.position_embedding_dim) + # (n, num_heads, w, w_kv, dim) + position_feat_x = self.appr_geom_fc_x(position_embed_x).\ + view(1, w, w_kv, num_heads, self.qk_embed_dim).\ + permute(0, 3, 1, 2, 4).\ + repeat(n, 1, 1, 1, 1) + + # (n, num_heads, h, h_kv, dim) + position_feat_y = self.appr_geom_fc_y(position_embed_y).\ + view(1, h, h_kv, num_heads, self.qk_embed_dim).\ + permute(0, 3, 1, 2, 4).\ + repeat(n, 1, 1, 1, 1) + + position_feat_x /= math.sqrt(2) + position_feat_y /= math.sqrt(2) + + # accelerate for saliency only + if (np.sum(self.attention_type) == 1) and self.attention_type[2]: + appr_bias = self.appr_bias.\ + view(1, num_heads, 1, self.qk_embed_dim).\ + repeat(n, 1, 1, 1) + + energy = torch.matmul(appr_bias, proj_key).\ + view(n, num_heads, 1, h_kv * w_kv) + + h = 1 + w = 1 + else: + # (n, num_heads, h*w, h_kv*w_kv), query before key, 540mb for + if not self.attention_type[0]: + energy = torch.zeros( + n, + num_heads, + h, + w, + h_kv, + w_kv, + dtype=x_input.dtype, + device=x_input.device) + + # attention_type[0]: appr - appr + # attention_type[1]: appr - position + # attention_type[2]: bias - appr + # attention_type[3]: bias - position + if self.attention_type[0] or self.attention_type[2]: + if self.attention_type[0] and self.attention_type[2]: + appr_bias = self.appr_bias.\ + view(1, num_heads, 1, self.qk_embed_dim) + energy = torch.matmul(proj_query + appr_bias, proj_key).\ + view(n, num_heads, h, w, h_kv, w_kv) + + elif self.attention_type[0]: + energy = torch.matmul(proj_query, proj_key).\ + view(n, num_heads, h, w, h_kv, w_kv) + + elif self.attention_type[2]: + appr_bias = self.appr_bias.\ + view(1, num_heads, 1, self.qk_embed_dim).\ + repeat(n, 1, 1, 1) + + energy += torch.matmul(appr_bias, proj_key).\ + view(n, num_heads, 1, 1, h_kv, w_kv) + + if self.attention_type[1] or self.attention_type[3]: + if self.attention_type[1] and self.attention_type[3]: + geom_bias = self.geom_bias.\ + view(1, num_heads, 1, self.qk_embed_dim) + + proj_query_reshape = (proj_query + geom_bias).\ + view(n, num_heads, h, w, self.qk_embed_dim) + + energy_x = torch.matmul( + proj_query_reshape.permute(0, 1, 3, 2, 4), + position_feat_x.permute(0, 1, 2, 4, 3)) + energy_x = energy_x.\ + permute(0, 1, 3, 2, 4).unsqueeze(4) + + energy_y = torch.matmul( + proj_query_reshape, + position_feat_y.permute(0, 1, 2, 4, 3)) + energy_y = energy_y.unsqueeze(5) + + energy += energy_x + energy_y + + elif self.attention_type[1]: + proj_query_reshape = proj_query.\ + view(n, num_heads, h, w, self.qk_embed_dim) + proj_query_reshape = proj_query_reshape.\ + permute(0, 1, 3, 2, 4) + position_feat_x_reshape = position_feat_x.\ + permute(0, 1, 2, 4, 3) + position_feat_y_reshape = position_feat_y.\ + permute(0, 1, 2, 4, 3) + + energy_x = torch.matmul(proj_query_reshape, + position_feat_x_reshape) + energy_x = energy_x.permute(0, 1, 3, 2, 4).unsqueeze(4) + + energy_y = torch.matmul(proj_query_reshape, + position_feat_y_reshape) + energy_y = energy_y.unsqueeze(5) + + energy += energy_x + energy_y + + elif self.attention_type[3]: + geom_bias = self.geom_bias.\ + view(1, num_heads, self.qk_embed_dim, 1).\ + repeat(n, 1, 1, 1) + + position_feat_x_reshape = position_feat_x.\ + view(n, num_heads, w * w_kv, self.qk_embed_dim) + + position_feat_y_reshape = position_feat_y.\ + view(n, num_heads, h * h_kv, self.qk_embed_dim) + + energy_x = torch.matmul(position_feat_x_reshape, geom_bias) + energy_x = energy_x.view(n, num_heads, 1, w, 1, w_kv) + + energy_y = torch.matmul(position_feat_y_reshape, geom_bias) + energy_y = energy_y.view(n, num_heads, h, 1, h_kv, 1) + + energy += energy_x + energy_y + + energy = energy.view(n, num_heads, h * w, h_kv * w_kv) + + if self.spatial_range >= 0: + cur_local_constraint_map = \ + self.local_constraint_map[:h, :w, :h_kv, :w_kv].\ + contiguous().\ + view(1, 1, h*w, h_kv*w_kv) + + energy = energy.masked_fill_(cur_local_constraint_map, + float('-inf')) + + attention = F.softmax(energy, 3) + + proj_value = self.value_conv(x_kv) + proj_value_reshape = proj_value.\ + view((n, num_heads, self.v_dim, h_kv * w_kv)).\ + permute(0, 1, 3, 2) + + out = torch.matmul(attention, proj_value_reshape).\ + permute(0, 1, 3, 2).\ + contiguous().\ + view(n, self.v_dim * self.num_heads, h, w) + + out = self.proj_conv(out) + + # output is downsampled, upsample back to input size + if self.q_downsample is not None: + out = F.interpolate( + out, + size=x_input.shape[2:], + mode='bilinear', + align_corners=False) + + out = self.gamma * out + x_input + return out + + def init_weights(self): + for m in self.modules(): + if hasattr(m, 'kaiming_init') and m.kaiming_init: + kaiming_init( + m, + mode='fan_in', + nonlinearity='leaky_relu', + bias=0, + distribution='uniform', + a=1) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/hsigmoid.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/hsigmoid.py new file mode 100644 index 0000000000000000000000000000000000000000..e013d739e339ef972822b1b1d53ec6a1b26ac7cb --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/hsigmoid.py @@ -0,0 +1,46 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch.nn as nn + +from .registry import ACTIVATION_LAYERS + + +@ACTIVATION_LAYERS.register_module() +class HSigmoid(nn.Module): + """Hard Sigmoid Module. Apply the hard sigmoid function: + Hsigmoid(x) = min(max((x + bias) / divisor, min_value), max_value) + Default: Hsigmoid(x) = min(max((x + 3) / 6, 0), 1) + + Note: + In MMCV v1.4.4, we modified the default value of args to align with + PyTorch official. + + Args: + bias (float): Bias of the input feature map. Default: 3.0. + divisor (float): Divisor of the input feature map. Default: 6.0. + min_value (float): Lower bound value. Default: 0.0. + max_value (float): Upper bound value. Default: 1.0. + + Returns: + Tensor: The output tensor. + """ + + def __init__(self, bias=3.0, divisor=6.0, min_value=0.0, max_value=1.0): + super(HSigmoid, self).__init__() + warnings.warn( + 'In MMCV v1.4.4, we modified the default value of args to align ' + 'with PyTorch official. Previous Implementation: ' + 'Hsigmoid(x) = min(max((x + 1) / 2, 0), 1). ' + 'Current Implementation: ' + 'Hsigmoid(x) = min(max((x + 3) / 6, 0), 1).') + self.bias = bias + self.divisor = divisor + assert self.divisor != 0 + self.min_value = min_value + self.max_value = max_value + + def forward(self, x): + x = (x + self.bias) / self.divisor + + return x.clamp_(self.min_value, self.max_value) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/hswish.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/hswish.py new file mode 100644 index 0000000000000000000000000000000000000000..27096832f7eba4aa6f945af9b18b08f41cc2acfd --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/hswish.py @@ -0,0 +1,38 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn + +from mmcv.utils import TORCH_VERSION, digit_version +from .registry import ACTIVATION_LAYERS + + +class HSwish(nn.Module): + """Hard Swish Module. + + This module applies the hard swish function: + + .. math:: + Hswish(x) = x * ReLU6(x + 3) / 6 + + Args: + inplace (bool): can optionally do the operation in-place. + Default: False. + + Returns: + Tensor: The output tensor. + """ + + def __init__(self, inplace=False): + super(HSwish, self).__init__() + self.act = nn.ReLU6(inplace) + + def forward(self, x): + return x * self.act(x + 3) / 6 + + +if (TORCH_VERSION == 'parrots' + or digit_version(TORCH_VERSION) < digit_version('1.7')): + # Hardswish is not supported when PyTorch version < 1.6. + # And Hardswish in PyTorch 1.6 does not support inplace. + ACTIVATION_LAYERS.register_module(module=HSwish) +else: + ACTIVATION_LAYERS.register_module(module=nn.Hardswish, name='HSwish') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/non_local.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/non_local.py new file mode 100644 index 0000000000000000000000000000000000000000..92d00155ef275c1201ea66bba30470a1785cc5d7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/non_local.py @@ -0,0 +1,306 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from abc import ABCMeta + +import torch +import torch.nn as nn + +from ..utils import constant_init, normal_init +from .conv_module import ConvModule +from .registry import PLUGIN_LAYERS + + +class _NonLocalNd(nn.Module, metaclass=ABCMeta): + """Basic Non-local module. + + This module is proposed in + "Non-local Neural Networks" + Paper reference: https://arxiv.org/abs/1711.07971 + Code reference: https://github.com/AlexHex7/Non-local_pytorch + + Args: + in_channels (int): Channels of the input feature map. + reduction (int): Channel reduction ratio. Default: 2. + use_scale (bool): Whether to scale pairwise_weight by + `1/sqrt(inter_channels)` when the mode is `embedded_gaussian`. + Default: True. + conv_cfg (None | dict): The config dict for convolution layers. + If not specified, it will use `nn.Conv2d` for convolution layers. + Default: None. + norm_cfg (None | dict): The config dict for normalization layers. + Default: None. (This parameter is only applicable to conv_out.) + mode (str): Options are `gaussian`, `concatenation`, + `embedded_gaussian` and `dot_product`. Default: embedded_gaussian. + """ + + def __init__(self, + in_channels, + reduction=2, + use_scale=True, + conv_cfg=None, + norm_cfg=None, + mode='embedded_gaussian', + **kwargs): + super(_NonLocalNd, self).__init__() + self.in_channels = in_channels + self.reduction = reduction + self.use_scale = use_scale + self.inter_channels = max(in_channels // reduction, 1) + self.mode = mode + + if mode not in [ + 'gaussian', 'embedded_gaussian', 'dot_product', 'concatenation' + ]: + raise ValueError("Mode should be in 'gaussian', 'concatenation', " + f"'embedded_gaussian' or 'dot_product', but got " + f'{mode} instead.') + + # g, theta, phi are defaulted as `nn.ConvNd`. + # Here we use ConvModule for potential usage. + self.g = ConvModule( + self.in_channels, + self.inter_channels, + kernel_size=1, + conv_cfg=conv_cfg, + act_cfg=None) + self.conv_out = ConvModule( + self.inter_channels, + self.in_channels, + kernel_size=1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=None) + + if self.mode != 'gaussian': + self.theta = ConvModule( + self.in_channels, + self.inter_channels, + kernel_size=1, + conv_cfg=conv_cfg, + act_cfg=None) + self.phi = ConvModule( + self.in_channels, + self.inter_channels, + kernel_size=1, + conv_cfg=conv_cfg, + act_cfg=None) + + if self.mode == 'concatenation': + self.concat_project = ConvModule( + self.inter_channels * 2, + 1, + kernel_size=1, + stride=1, + padding=0, + bias=False, + act_cfg=dict(type='ReLU')) + + self.init_weights(**kwargs) + + def init_weights(self, std=0.01, zeros_init=True): + if self.mode != 'gaussian': + for m in [self.g, self.theta, self.phi]: + normal_init(m.conv, std=std) + else: + normal_init(self.g.conv, std=std) + if zeros_init: + if self.conv_out.norm_cfg is None: + constant_init(self.conv_out.conv, 0) + else: + constant_init(self.conv_out.norm, 0) + else: + if self.conv_out.norm_cfg is None: + normal_init(self.conv_out.conv, std=std) + else: + normal_init(self.conv_out.norm, std=std) + + def gaussian(self, theta_x, phi_x): + # NonLocal1d pairwise_weight: [N, H, H] + # NonLocal2d pairwise_weight: [N, HxW, HxW] + # NonLocal3d pairwise_weight: [N, TxHxW, TxHxW] + pairwise_weight = torch.matmul(theta_x, phi_x) + pairwise_weight = pairwise_weight.softmax(dim=-1) + return pairwise_weight + + def embedded_gaussian(self, theta_x, phi_x): + # NonLocal1d pairwise_weight: [N, H, H] + # NonLocal2d pairwise_weight: [N, HxW, HxW] + # NonLocal3d pairwise_weight: [N, TxHxW, TxHxW] + pairwise_weight = torch.matmul(theta_x, phi_x) + if self.use_scale: + # theta_x.shape[-1] is `self.inter_channels` + pairwise_weight /= theta_x.shape[-1]**0.5 + pairwise_weight = pairwise_weight.softmax(dim=-1) + return pairwise_weight + + def dot_product(self, theta_x, phi_x): + # NonLocal1d pairwise_weight: [N, H, H] + # NonLocal2d pairwise_weight: [N, HxW, HxW] + # NonLocal3d pairwise_weight: [N, TxHxW, TxHxW] + pairwise_weight = torch.matmul(theta_x, phi_x) + pairwise_weight /= pairwise_weight.shape[-1] + return pairwise_weight + + def concatenation(self, theta_x, phi_x): + # NonLocal1d pairwise_weight: [N, H, H] + # NonLocal2d pairwise_weight: [N, HxW, HxW] + # NonLocal3d pairwise_weight: [N, TxHxW, TxHxW] + h = theta_x.size(2) + w = phi_x.size(3) + theta_x = theta_x.repeat(1, 1, 1, w) + phi_x = phi_x.repeat(1, 1, h, 1) + + concat_feature = torch.cat([theta_x, phi_x], dim=1) + pairwise_weight = self.concat_project(concat_feature) + n, _, h, w = pairwise_weight.size() + pairwise_weight = pairwise_weight.view(n, h, w) + pairwise_weight /= pairwise_weight.shape[-1] + + return pairwise_weight + + def forward(self, x): + # Assume `reduction = 1`, then `inter_channels = C` + # or `inter_channels = C` when `mode="gaussian"` + + # NonLocal1d x: [N, C, H] + # NonLocal2d x: [N, C, H, W] + # NonLocal3d x: [N, C, T, H, W] + n = x.size(0) + + # NonLocal1d g_x: [N, H, C] + # NonLocal2d g_x: [N, HxW, C] + # NonLocal3d g_x: [N, TxHxW, C] + g_x = self.g(x).view(n, self.inter_channels, -1) + g_x = g_x.permute(0, 2, 1) + + # NonLocal1d theta_x: [N, H, C], phi_x: [N, C, H] + # NonLocal2d theta_x: [N, HxW, C], phi_x: [N, C, HxW] + # NonLocal3d theta_x: [N, TxHxW, C], phi_x: [N, C, TxHxW] + if self.mode == 'gaussian': + theta_x = x.view(n, self.in_channels, -1) + theta_x = theta_x.permute(0, 2, 1) + if self.sub_sample: + phi_x = self.phi(x).view(n, self.in_channels, -1) + else: + phi_x = x.view(n, self.in_channels, -1) + elif self.mode == 'concatenation': + theta_x = self.theta(x).view(n, self.inter_channels, -1, 1) + phi_x = self.phi(x).view(n, self.inter_channels, 1, -1) + else: + theta_x = self.theta(x).view(n, self.inter_channels, -1) + theta_x = theta_x.permute(0, 2, 1) + phi_x = self.phi(x).view(n, self.inter_channels, -1) + + pairwise_func = getattr(self, self.mode) + # NonLocal1d pairwise_weight: [N, H, H] + # NonLocal2d pairwise_weight: [N, HxW, HxW] + # NonLocal3d pairwise_weight: [N, TxHxW, TxHxW] + pairwise_weight = pairwise_func(theta_x, phi_x) + + # NonLocal1d y: [N, H, C] + # NonLocal2d y: [N, HxW, C] + # NonLocal3d y: [N, TxHxW, C] + y = torch.matmul(pairwise_weight, g_x) + # NonLocal1d y: [N, C, H] + # NonLocal2d y: [N, C, H, W] + # NonLocal3d y: [N, C, T, H, W] + y = y.permute(0, 2, 1).contiguous().reshape(n, self.inter_channels, + *x.size()[2:]) + + output = x + self.conv_out(y) + + return output + + +class NonLocal1d(_NonLocalNd): + """1D Non-local module. + + Args: + in_channels (int): Same as `NonLocalND`. + sub_sample (bool): Whether to apply max pooling after pairwise + function (Note that the `sub_sample` is applied on spatial only). + Default: False. + conv_cfg (None | dict): Same as `NonLocalND`. + Default: dict(type='Conv1d'). + """ + + def __init__(self, + in_channels, + sub_sample=False, + conv_cfg=dict(type='Conv1d'), + **kwargs): + super(NonLocal1d, self).__init__( + in_channels, conv_cfg=conv_cfg, **kwargs) + + self.sub_sample = sub_sample + + if sub_sample: + max_pool_layer = nn.MaxPool1d(kernel_size=2) + self.g = nn.Sequential(self.g, max_pool_layer) + if self.mode != 'gaussian': + self.phi = nn.Sequential(self.phi, max_pool_layer) + else: + self.phi = max_pool_layer + + +@PLUGIN_LAYERS.register_module() +class NonLocal2d(_NonLocalNd): + """2D Non-local module. + + Args: + in_channels (int): Same as `NonLocalND`. + sub_sample (bool): Whether to apply max pooling after pairwise + function (Note that the `sub_sample` is applied on spatial only). + Default: False. + conv_cfg (None | dict): Same as `NonLocalND`. + Default: dict(type='Conv2d'). + """ + + _abbr_ = 'nonlocal_block' + + def __init__(self, + in_channels, + sub_sample=False, + conv_cfg=dict(type='Conv2d'), + **kwargs): + super(NonLocal2d, self).__init__( + in_channels, conv_cfg=conv_cfg, **kwargs) + + self.sub_sample = sub_sample + + if sub_sample: + max_pool_layer = nn.MaxPool2d(kernel_size=(2, 2)) + self.g = nn.Sequential(self.g, max_pool_layer) + if self.mode != 'gaussian': + self.phi = nn.Sequential(self.phi, max_pool_layer) + else: + self.phi = max_pool_layer + + +class NonLocal3d(_NonLocalNd): + """3D Non-local module. + + Args: + in_channels (int): Same as `NonLocalND`. + sub_sample (bool): Whether to apply max pooling after pairwise + function (Note that the `sub_sample` is applied on spatial only). + Default: False. + conv_cfg (None | dict): Same as `NonLocalND`. + Default: dict(type='Conv3d'). + """ + + def __init__(self, + in_channels, + sub_sample=False, + conv_cfg=dict(type='Conv3d'), + **kwargs): + super(NonLocal3d, self).__init__( + in_channels, conv_cfg=conv_cfg, **kwargs) + self.sub_sample = sub_sample + + if sub_sample: + max_pool_layer = nn.MaxPool3d(kernel_size=(1, 2, 2)) + self.g = nn.Sequential(self.g, max_pool_layer) + if self.mode != 'gaussian': + self.phi = nn.Sequential(self.phi, max_pool_layer) + else: + self.phi = max_pool_layer diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/norm.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/norm.py new file mode 100644 index 0000000000000000000000000000000000000000..51efdc1844044ddb5791481397e09f5499711aa2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/norm.py @@ -0,0 +1,144 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import inspect + +import torch.nn as nn + +from mmcv.utils import is_tuple_of +from mmcv.utils.parrots_wrapper import SyncBatchNorm, _BatchNorm, _InstanceNorm +from .registry import NORM_LAYERS + +NORM_LAYERS.register_module('BN', module=nn.BatchNorm2d) +NORM_LAYERS.register_module('BN1d', module=nn.BatchNorm1d) +NORM_LAYERS.register_module('BN2d', module=nn.BatchNorm2d) +NORM_LAYERS.register_module('BN3d', module=nn.BatchNorm3d) +NORM_LAYERS.register_module('SyncBN', module=SyncBatchNorm) +NORM_LAYERS.register_module('GN', module=nn.GroupNorm) +NORM_LAYERS.register_module('LN', module=nn.LayerNorm) +NORM_LAYERS.register_module('IN', module=nn.InstanceNorm2d) +NORM_LAYERS.register_module('IN1d', module=nn.InstanceNorm1d) +NORM_LAYERS.register_module('IN2d', module=nn.InstanceNorm2d) +NORM_LAYERS.register_module('IN3d', module=nn.InstanceNorm3d) + + +def infer_abbr(class_type): + """Infer abbreviation from the class name. + + When we build a norm layer with `build_norm_layer()`, we want to preserve + the norm type in variable names, e.g, self.bn1, self.gn. This method will + infer the abbreviation to map class types to abbreviations. + + Rule 1: If the class has the property "_abbr_", return the property. + Rule 2: If the parent class is _BatchNorm, GroupNorm, LayerNorm or + InstanceNorm, the abbreviation of this layer will be "bn", "gn", "ln" and + "in" respectively. + Rule 3: If the class name contains "batch", "group", "layer" or "instance", + the abbreviation of this layer will be "bn", "gn", "ln" and "in" + respectively. + Rule 4: Otherwise, the abbreviation falls back to "norm". + + Args: + class_type (type): The norm layer type. + + Returns: + str: The inferred abbreviation. + """ + if not inspect.isclass(class_type): + raise TypeError( + f'class_type must be a type, but got {type(class_type)}') + if hasattr(class_type, '_abbr_'): + return class_type._abbr_ + if issubclass(class_type, _InstanceNorm): # IN is a subclass of BN + return 'in' + elif issubclass(class_type, _BatchNorm): + return 'bn' + elif issubclass(class_type, nn.GroupNorm): + return 'gn' + elif issubclass(class_type, nn.LayerNorm): + return 'ln' + else: + class_name = class_type.__name__.lower() + if 'batch' in class_name: + return 'bn' + elif 'group' in class_name: + return 'gn' + elif 'layer' in class_name: + return 'ln' + elif 'instance' in class_name: + return 'in' + else: + return 'norm_layer' + + +def build_norm_layer(cfg, num_features, postfix=''): + """Build normalization layer. + + Args: + cfg (dict): The norm layer config, which should contain: + + - type (str): Layer type. + - layer args: Args needed to instantiate a norm layer. + - requires_grad (bool, optional): Whether stop gradient updates. + num_features (int): Number of input channels. + postfix (int | str): The postfix to be appended into norm abbreviation + to create named layer. + + Returns: + tuple[str, nn.Module]: The first element is the layer name consisting + of abbreviation and postfix, e.g., bn1, gn. The second element is the + created norm layer. + """ + if not isinstance(cfg, dict): + raise TypeError('cfg must be a dict') + if 'type' not in cfg: + raise KeyError('the cfg dict must contain the key "type"') + cfg_ = cfg.copy() + + layer_type = cfg_.pop('type') + if layer_type not in NORM_LAYERS: + raise KeyError(f'Unrecognized norm type {layer_type}') + + norm_layer = NORM_LAYERS.get(layer_type) + abbr = infer_abbr(norm_layer) + + assert isinstance(postfix, (int, str)) + name = abbr + str(postfix) + + requires_grad = cfg_.pop('requires_grad', True) + cfg_.setdefault('eps', 1e-5) + if layer_type != 'GN': + layer = norm_layer(num_features, **cfg_) + if layer_type == 'SyncBN' and hasattr(layer, '_specify_ddp_gpu_num'): + layer._specify_ddp_gpu_num(1) + else: + assert 'num_groups' in cfg_ + layer = norm_layer(num_channels=num_features, **cfg_) + + for param in layer.parameters(): + param.requires_grad = requires_grad + + return name, layer + + +def is_norm(layer, exclude=None): + """Check if a layer is a normalization layer. + + Args: + layer (nn.Module): The layer to be checked. + exclude (type | tuple[type]): Types to be excluded. + + Returns: + bool: Whether the layer is a norm layer. + """ + if exclude is not None: + if not isinstance(exclude, tuple): + exclude = (exclude, ) + if not is_tuple_of(exclude, type): + raise TypeError( + f'"exclude" must be either None or type or a tuple of types, ' + f'but got {type(exclude)}: {exclude}') + + if exclude and isinstance(layer, exclude): + return False + + all_norm_bases = (_BatchNorm, _InstanceNorm, nn.GroupNorm, nn.LayerNorm) + return isinstance(layer, all_norm_bases) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/padding.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/padding.py new file mode 100644 index 0000000000000000000000000000000000000000..e4ac6b28a1789bd551c613a7d3e7b622433ac7ec --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/padding.py @@ -0,0 +1,36 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn + +from .registry import PADDING_LAYERS + +PADDING_LAYERS.register_module('zero', module=nn.ZeroPad2d) +PADDING_LAYERS.register_module('reflect', module=nn.ReflectionPad2d) +PADDING_LAYERS.register_module('replicate', module=nn.ReplicationPad2d) + + +def build_padding_layer(cfg, *args, **kwargs): + """Build padding layer. + + Args: + cfg (None or dict): The padding layer config, which should contain: + - type (str): Layer type. + - layer args: Args needed to instantiate a padding layer. + + Returns: + nn.Module: Created padding layer. + """ + if not isinstance(cfg, dict): + raise TypeError('cfg must be a dict') + if 'type' not in cfg: + raise KeyError('the cfg dict must contain the key "type"') + + cfg_ = cfg.copy() + padding_type = cfg_.pop('type') + if padding_type not in PADDING_LAYERS: + raise KeyError(f'Unrecognized padding type {padding_type}.') + else: + padding_layer = PADDING_LAYERS.get(padding_type) + + layer = padding_layer(*args, **kwargs, **cfg_) + + return layer diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/plugin.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/plugin.py new file mode 100644 index 0000000000000000000000000000000000000000..009f7529be8d9577ca8f4c0196e15a939bb2f21c --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/plugin.py @@ -0,0 +1,89 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import inspect +import platform + +from .registry import PLUGIN_LAYERS + +if platform.system() == 'Windows': + import regex as re +else: + import re + + +def infer_abbr(class_type): + """Infer abbreviation from the class name. + + This method will infer the abbreviation to map class types to + abbreviations. + + Rule 1: If the class has the property "abbr", return the property. + Rule 2: Otherwise, the abbreviation falls back to snake case of class + name, e.g. the abbreviation of ``FancyBlock`` will be ``fancy_block``. + + Args: + class_type (type): The norm layer type. + + Returns: + str: The inferred abbreviation. + """ + + def camel2snack(word): + """Convert camel case word into snack case. + + Modified from `inflection lib + `_. + + Example:: + + >>> camel2snack("FancyBlock") + 'fancy_block' + """ + + word = re.sub(r'([A-Z]+)([A-Z][a-z])', r'\1_\2', word) + word = re.sub(r'([a-z\d])([A-Z])', r'\1_\2', word) + word = word.replace('-', '_') + return word.lower() + + if not inspect.isclass(class_type): + raise TypeError( + f'class_type must be a type, but got {type(class_type)}') + if hasattr(class_type, '_abbr_'): + return class_type._abbr_ + else: + return camel2snack(class_type.__name__) + + +def build_plugin_layer(cfg, postfix='', **kwargs): + """Build plugin layer. + + Args: + cfg (None or dict): cfg should contain: + + - type (str): identify plugin layer type. + - layer args: args needed to instantiate a plugin layer. + postfix (int, str): appended into norm abbreviation to + create named layer. Default: ''. + + Returns: + tuple[str, nn.Module]: The first one is the concatenation of + abbreviation and postfix. The second is the created plugin layer. + """ + if not isinstance(cfg, dict): + raise TypeError('cfg must be a dict') + if 'type' not in cfg: + raise KeyError('the cfg dict must contain the key "type"') + cfg_ = cfg.copy() + + layer_type = cfg_.pop('type') + if layer_type not in PLUGIN_LAYERS: + raise KeyError(f'Unrecognized plugin type {layer_type}') + + plugin_layer = PLUGIN_LAYERS.get(layer_type) + abbr = infer_abbr(plugin_layer) + + assert isinstance(postfix, (int, str)) + name = abbr + str(postfix) + + layer = plugin_layer(**kwargs, **cfg_) + + return name, layer diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/registry.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/registry.py new file mode 100644 index 0000000000000000000000000000000000000000..c29279776dd523e706b6af8f9b9de700bed05ba7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/registry.py @@ -0,0 +1,16 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.utils import Registry + +CONV_LAYERS = Registry('conv layer') +NORM_LAYERS = Registry('norm layer') +ACTIVATION_LAYERS = Registry('activation layer') +PADDING_LAYERS = Registry('padding layer') +UPSAMPLE_LAYERS = Registry('upsample layer') +PLUGIN_LAYERS = Registry('plugin layer') + +DROPOUT_LAYERS = Registry('drop out layers') +POSITIONAL_ENCODING = Registry('position encoding') +ATTENTION = Registry('attention') +FEEDFORWARD_NETWORK = Registry('feed-forward Network') +TRANSFORMER_LAYER = Registry('transformerLayer') +TRANSFORMER_LAYER_SEQUENCE = Registry('transformer-layers sequence') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/scale.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/scale.py new file mode 100644 index 0000000000000000000000000000000000000000..c905fffcc8bf998d18d94f927591963c428025e2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/scale.py @@ -0,0 +1,21 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn + + +class Scale(nn.Module): + """A learnable scale parameter. + + This layer scales the input by a learnable factor. It multiplies a + learnable scale parameter of shape (1,) with input of any shape. + + Args: + scale (float): Initial value of scale factor. Default: 1.0 + """ + + def __init__(self, scale=1.0): + super(Scale, self).__init__() + self.scale = nn.Parameter(torch.tensor(scale, dtype=torch.float)) + + def forward(self, x): + return x * self.scale diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/swish.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/swish.py new file mode 100644 index 0000000000000000000000000000000000000000..e2ca8ed7b749413f011ae54aac0cab27e6f0b51f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/swish.py @@ -0,0 +1,25 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn + +from .registry import ACTIVATION_LAYERS + + +@ACTIVATION_LAYERS.register_module() +class Swish(nn.Module): + """Swish Module. + + This module applies the swish function: + + .. math:: + Swish(x) = x * Sigmoid(x) + + Returns: + Tensor: The output tensor. + """ + + def __init__(self): + super(Swish, self).__init__() + + def forward(self, x): + return x * torch.sigmoid(x) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/transformer.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/transformer.py new file mode 100644 index 0000000000000000000000000000000000000000..70c6623c74c7c358946d4ff80ba7e572f6919f6b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/transformer.py @@ -0,0 +1,944 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import math +import warnings +from typing import Sequence + +import torch +import torch.nn as nn +import torch.nn.functional as F + +from mmcv.cnn import (Linear, build_activation_layer, build_conv_layer, + build_norm_layer) +from mmcv.runner.base_module import BaseModule, ModuleList, Sequential +from mmcv.utils import (ConfigDict, build_from_cfg, deprecated_api_warning, + to_2tuple) +from .drop import build_dropout +from .registry import (ATTENTION, FEEDFORWARD_NETWORK, POSITIONAL_ENCODING, + TRANSFORMER_LAYER, TRANSFORMER_LAYER_SEQUENCE) + +# Avoid BC-breaking of importing MultiScaleDeformableAttention from this file +try: + from mmcv.ops.multi_scale_deform_attn import \ + MultiScaleDeformableAttention # noqa F401 + warnings.warn( + ImportWarning( + '``MultiScaleDeformableAttention`` has been moved to ' + '``mmcv.ops.multi_scale_deform_attn``, please change original path ' # noqa E501 + '``from mmcv.cnn.bricks.transformer import MultiScaleDeformableAttention`` ' # noqa E501 + 'to ``from mmcv.ops.multi_scale_deform_attn import MultiScaleDeformableAttention`` ' # noqa E501 + )) + +except ImportError: + warnings.warn('Fail to import ``MultiScaleDeformableAttention`` from ' + '``mmcv.ops.multi_scale_deform_attn``, ' + 'You should install ``mmcv-full`` if you need this module. ') + + +def build_positional_encoding(cfg, default_args=None): + """Builder for Position Encoding.""" + return build_from_cfg(cfg, POSITIONAL_ENCODING, default_args) + + +def build_attention(cfg, default_args=None): + """Builder for attention.""" + return build_from_cfg(cfg, ATTENTION, default_args) + + +def build_feedforward_network(cfg, default_args=None): + """Builder for feed-forward network (FFN).""" + return build_from_cfg(cfg, FEEDFORWARD_NETWORK, default_args) + + +def build_transformer_layer(cfg, default_args=None): + """Builder for transformer layer.""" + return build_from_cfg(cfg, TRANSFORMER_LAYER, default_args) + + +def build_transformer_layer_sequence(cfg, default_args=None): + """Builder for transformer encoder and transformer decoder.""" + return build_from_cfg(cfg, TRANSFORMER_LAYER_SEQUENCE, default_args) + + +class AdaptivePadding(nn.Module): + """Applies padding adaptively to the input. + + This module can make input get fully covered by filter + you specified. It support two modes "same" and "corner". The + "same" mode is same with "SAME" padding mode in TensorFlow, pad + zero around input. The "corner" mode would pad zero + to bottom right. + + Args: + kernel_size (int | tuple): Size of the kernel. Default: 1. + stride (int | tuple): Stride of the filter. Default: 1. + dilation (int | tuple): Spacing between kernel elements. + Default: 1. + padding (str): Support "same" and "corner", "corner" mode + would pad zero to bottom right, and "same" mode would + pad zero around input. Default: "corner". + + Example: + >>> kernel_size = 16 + >>> stride = 16 + >>> dilation = 1 + >>> input = torch.rand(1, 1, 15, 17) + >>> adap_pad = AdaptivePadding( + >>> kernel_size=kernel_size, + >>> stride=stride, + >>> dilation=dilation, + >>> padding="corner") + >>> out = adap_pad(input) + >>> assert (out.shape[2], out.shape[3]) == (16, 32) + >>> input = torch.rand(1, 1, 16, 17) + >>> out = adap_pad(input) + >>> assert (out.shape[2], out.shape[3]) == (16, 32) + """ + + def __init__(self, kernel_size=1, stride=1, dilation=1, padding='corner'): + super(AdaptivePadding, self).__init__() + assert padding in ('same', 'corner') + + kernel_size = to_2tuple(kernel_size) + stride = to_2tuple(stride) + dilation = to_2tuple(dilation) + + self.padding = padding + self.kernel_size = kernel_size + self.stride = stride + self.dilation = dilation + + def get_pad_shape(self, input_shape): + """Calculate the padding size of input. + + Args: + input_shape (:obj:`torch.Size`): arrange as (H, W). + + Returns: + Tuple[int]: The padding size along the + original H and W directions + """ + input_h, input_w = input_shape + kernel_h, kernel_w = self.kernel_size + stride_h, stride_w = self.stride + output_h = math.ceil(input_h / stride_h) + output_w = math.ceil(input_w / stride_w) + pad_h = max((output_h - 1) * stride_h + + (kernel_h - 1) * self.dilation[0] + 1 - input_h, 0) + pad_w = max((output_w - 1) * stride_w + + (kernel_w - 1) * self.dilation[1] + 1 - input_w, 0) + return pad_h, pad_w + + def forward(self, x): + """Add padding to `x` + + Args: + x (Tensor): Input tensor has shape (B, C, H, W). + + Returns: + Tensor: The tensor with adaptive padding + """ + pad_h, pad_w = self.get_pad_shape(x.size()[-2:]) + if pad_h > 0 or pad_w > 0: + if self.padding == 'corner': + x = F.pad(x, [0, pad_w, 0, pad_h]) + elif self.padding == 'same': + x = F.pad(x, [ + pad_w // 2, pad_w - pad_w // 2, pad_h // 2, + pad_h - pad_h // 2 + ]) + return x + + +class PatchEmbed(BaseModule): + """Image to Patch Embedding. + + We use a conv layer to implement PatchEmbed. + + Args: + in_channels (int): The num of input channels. Default: 3 + embed_dims (int): The dimensions of embedding. Default: 768 + conv_type (str): The type of convolution + to generate patch embedding. Default: "Conv2d". + kernel_size (int): The kernel_size of embedding conv. Default: 16. + stride (int): The slide stride of embedding conv. + Default: 16. + padding (int | tuple | string): The padding length of + embedding conv. When it is a string, it means the mode + of adaptive padding, support "same" and "corner" now. + Default: "corner". + dilation (int): The dilation rate of embedding conv. Default: 1. + bias (bool): Bias of embed conv. Default: True. + norm_cfg (dict, optional): Config dict for normalization layer. + Default: None. + input_size (int | tuple | None): The size of input, which will be + used to calculate the out size. Only works when `dynamic_size` + is False. Default: None. + init_cfg (`mmcv.ConfigDict`, optional): The Config for initialization. + Default: None. + """ + + def __init__(self, + in_channels=3, + embed_dims=768, + conv_type='Conv2d', + kernel_size=16, + stride=16, + padding='corner', + dilation=1, + bias=True, + norm_cfg=None, + input_size=None, + init_cfg=None): + super(PatchEmbed, self).__init__(init_cfg=init_cfg) + + self.embed_dims = embed_dims + if stride is None: + stride = kernel_size + + kernel_size = to_2tuple(kernel_size) + stride = to_2tuple(stride) + dilation = to_2tuple(dilation) + + if isinstance(padding, str): + self.adaptive_padding = AdaptivePadding( + kernel_size=kernel_size, + stride=stride, + dilation=dilation, + padding=padding) + # disable the padding of conv + padding = 0 + else: + self.adaptive_padding = None + padding = to_2tuple(padding) + + self.projection = build_conv_layer( + dict(type=conv_type), + in_channels=in_channels, + out_channels=embed_dims, + kernel_size=kernel_size, + stride=stride, + padding=padding, + dilation=dilation, + bias=bias) + + if norm_cfg is not None: + self.norm = build_norm_layer(norm_cfg, embed_dims)[1] + else: + self.norm = None + + if input_size: + input_size = to_2tuple(input_size) + # `init_out_size` would be used outside to + # calculate the num_patches + # e.g. when `use_abs_pos_embed` outside + self.init_input_size = input_size + if self.adaptive_padding: + pad_h, pad_w = self.adaptive_padding.get_pad_shape(input_size) + input_h, input_w = input_size + input_h = input_h + pad_h + input_w = input_w + pad_w + input_size = (input_h, input_w) + + # https://pytorch.org/docs/stable/generated/torch.nn.Conv2d.html + h_out = (input_size[0] + 2 * padding[0] - dilation[0] * + (kernel_size[0] - 1) - 1) // stride[0] + 1 + w_out = (input_size[1] + 2 * padding[1] - dilation[1] * + (kernel_size[1] - 1) - 1) // stride[1] + 1 + self.init_out_size = (h_out, w_out) + else: + self.init_input_size = None + self.init_out_size = None + + def forward(self, x): + """ + Args: + x (Tensor): Has shape (B, C, H, W). In most case, C is 3. + + Returns: + tuple: Contains merged results and its spatial shape. + + - x (Tensor): Has shape (B, out_h * out_w, embed_dims) + - out_size (tuple[int]): Spatial shape of x, arrange as + (out_h, out_w). + """ + + if self.adaptive_padding: + x = self.adaptive_padding(x) + + x = self.projection(x) + out_size = (x.shape[2], x.shape[3]) + x = x.flatten(2).transpose(1, 2) + if self.norm is not None: + x = self.norm(x) + return x, out_size + + +class PatchMerging(BaseModule): + """Merge patch feature map. + + This layer groups feature map by kernel_size, and applies norm and linear + layers to the grouped feature map ((used in Swin Transformer)). + Our implementation uses `nn.Unfold` to + merge patches, which is about 25% faster than the original + implementation. However, we need to modify pretrained + models for compatibility. + + Args: + in_channels (int): The num of input channels. + to gets fully covered by filter and stride you specified. + out_channels (int): The num of output channels. + kernel_size (int | tuple, optional): the kernel size in the unfold + layer. Defaults to 2. + stride (int | tuple, optional): the stride of the sliding blocks in the + unfold layer. Default: None. (Would be set as `kernel_size`) + padding (int | tuple | string ): The padding length of + embedding conv. When it is a string, it means the mode + of adaptive padding, support "same" and "corner" now. + Default: "corner". + dilation (int | tuple, optional): dilation parameter in the unfold + layer. Default: 1. + bias (bool, optional): Whether to add bias in linear layer or not. + Defaults: False. + norm_cfg (dict, optional): Config dict for normalization layer. + Default: dict(type='LN'). + init_cfg (dict, optional): The extra config for initialization. + Default: None. + """ + + def __init__(self, + in_channels, + out_channels, + kernel_size=2, + stride=None, + padding='corner', + dilation=1, + bias=False, + norm_cfg=dict(type='LN'), + init_cfg=None): + super().__init__(init_cfg=init_cfg) + self.in_channels = in_channels + self.out_channels = out_channels + if stride: + stride = stride + else: + stride = kernel_size + + kernel_size = to_2tuple(kernel_size) + stride = to_2tuple(stride) + dilation = to_2tuple(dilation) + + if isinstance(padding, str): + self.adaptive_padding = AdaptivePadding( + kernel_size=kernel_size, + stride=stride, + dilation=dilation, + padding=padding) + # disable the padding of unfold + padding = 0 + else: + self.adaptive_padding = None + + padding = to_2tuple(padding) + self.sampler = nn.Unfold( + kernel_size=kernel_size, + dilation=dilation, + padding=padding, + stride=stride) + + sample_dim = kernel_size[0] * kernel_size[1] * in_channels + + if norm_cfg is not None: + self.norm = build_norm_layer(norm_cfg, sample_dim)[1] + else: + self.norm = None + + self.reduction = nn.Linear(sample_dim, out_channels, bias=bias) + + def forward(self, x, input_size): + """ + Args: + x (Tensor): Has shape (B, H*W, C_in). + input_size (tuple[int]): The spatial shape of x, arrange as (H, W). + Default: None. + + Returns: + tuple: Contains merged results and its spatial shape. + + - x (Tensor): Has shape (B, Merged_H * Merged_W, C_out) + - out_size (tuple[int]): Spatial shape of x, arrange as + (Merged_H, Merged_W). + """ + B, L, C = x.shape + assert isinstance(input_size, Sequence), f'Expect ' \ + f'input_size is ' \ + f'`Sequence` ' \ + f'but get {input_size}' + + H, W = input_size + assert L == H * W, 'input feature has wrong size' + + x = x.view(B, H, W, C).permute([0, 3, 1, 2]) # B, C, H, W + + if self.adaptive_padding: + x = self.adaptive_padding(x) + H, W = x.shape[-2:] + + # Use nn.Unfold to merge patch. About 25% faster than original method, + # but need to modify pretrained model for compatibility + # if kernel_size=2 and stride=2, x should has shape (B, 4*C, H/2*W/2) + x = self.sampler(x) + + out_h = (H + 2 * self.sampler.padding[0] - self.sampler.dilation[0] * + (self.sampler.kernel_size[0] - 1) - + 1) // self.sampler.stride[0] + 1 + out_w = (W + 2 * self.sampler.padding[1] - self.sampler.dilation[1] * + (self.sampler.kernel_size[1] - 1) - + 1) // self.sampler.stride[1] + 1 + + output_size = (out_h, out_w) + x = x.transpose(1, 2) # B, H/2*W/2, 4*C + x = self.norm(x) if self.norm else x + x = self.reduction(x) + return x, output_size + + +@ATTENTION.register_module() +class MultiheadAttention(BaseModule): + """A wrapper for ``torch.nn.MultiheadAttention``. + + This module implements MultiheadAttention with identity connection, + and positional encoding is also passed as input. + + Args: + embed_dims (int): The embedding dimension. + num_heads (int): Parallel attention heads. + attn_drop (float): A Dropout layer on attn_output_weights. + Default: 0.0. + proj_drop (float): A Dropout layer after `nn.MultiheadAttention`. + Default: 0.0. + dropout_layer (obj:`ConfigDict`): The dropout_layer used + when adding the shortcut. + init_cfg (obj:`mmcv.ConfigDict`): The Config for initialization. + Default: None. + batch_first (bool): When it is True, Key, Query and Value are shape of + (batch, n, embed_dim), otherwise (n, batch, embed_dim). + Default to False. + """ + + def __init__(self, + embed_dims, + num_heads, + attn_drop=0., + proj_drop=0., + dropout_layer=dict(type='Dropout', drop_prob=0.), + init_cfg=None, + batch_first=False, + **kwargs): + super(MultiheadAttention, self).__init__(init_cfg) + if 'dropout' in kwargs: + warnings.warn( + 'The arguments `dropout` in MultiheadAttention ' + 'has been deprecated, now you can separately ' + 'set `attn_drop`(float), proj_drop(float), ' + 'and `dropout_layer`(dict) ', DeprecationWarning) + attn_drop = kwargs['dropout'] + dropout_layer['drop_prob'] = kwargs.pop('dropout') + + self.embed_dims = embed_dims + self.num_heads = num_heads + self.batch_first = batch_first + + self.attn = nn.MultiheadAttention(embed_dims, num_heads, attn_drop, + **kwargs) + + self.proj_drop = nn.Dropout(proj_drop) + self.dropout_layer = build_dropout( + dropout_layer) if dropout_layer else nn.Identity() + + @deprecated_api_warning({'residual': 'identity'}, + cls_name='MultiheadAttention') + def forward(self, + query, + key=None, + value=None, + identity=None, + query_pos=None, + key_pos=None, + attn_mask=None, + key_padding_mask=None, + **kwargs): + """Forward function for `MultiheadAttention`. + + **kwargs allow passing a more general data flow when combining + with other operations in `transformerlayer`. + + Args: + query (Tensor): The input query with shape [num_queries, bs, + embed_dims] if self.batch_first is False, else + [bs, num_queries embed_dims]. + key (Tensor): The key tensor with shape [num_keys, bs, + embed_dims] if self.batch_first is False, else + [bs, num_keys, embed_dims] . + If None, the ``query`` will be used. Defaults to None. + value (Tensor): The value tensor with same shape as `key`. + Same in `nn.MultiheadAttention.forward`. Defaults to None. + If None, the `key` will be used. + identity (Tensor): This tensor, with the same shape as x, + will be used for the identity link. + If None, `x` will be used. Defaults to None. + query_pos (Tensor): The positional encoding for query, with + the same shape as `x`. If not None, it will + be added to `x` before forward function. Defaults to None. + key_pos (Tensor): The positional encoding for `key`, with the + same shape as `key`. Defaults to None. If not None, it will + be added to `key` before forward function. If None, and + `query_pos` has the same shape as `key`, then `query_pos` + will be used for `key_pos`. Defaults to None. + attn_mask (Tensor): ByteTensor mask with shape [num_queries, + num_keys]. Same in `nn.MultiheadAttention.forward`. + Defaults to None. + key_padding_mask (Tensor): ByteTensor with shape [bs, num_keys]. + Defaults to None. + + Returns: + Tensor: forwarded results with shape + [num_queries, bs, embed_dims] + if self.batch_first is False, else + [bs, num_queries embed_dims]. + """ + + if key is None: + key = query + if value is None: + value = key + if identity is None: + identity = query + if key_pos is None: + if query_pos is not None: + # use query_pos if key_pos is not available + if query_pos.shape == key.shape: + key_pos = query_pos + else: + warnings.warn(f'position encoding of key is' + f'missing in {self.__class__.__name__}.') + if query_pos is not None: + query = query + query_pos + if key_pos is not None: + key = key + key_pos + + # Because the dataflow('key', 'query', 'value') of + # ``torch.nn.MultiheadAttention`` is (num_query, batch, + # embed_dims), We should adjust the shape of dataflow from + # batch_first (batch, num_query, embed_dims) to num_query_first + # (num_query ,batch, embed_dims), and recover ``attn_output`` + # from num_query_first to batch_first. + if self.batch_first: + query = query.transpose(0, 1) + key = key.transpose(0, 1) + value = value.transpose(0, 1) + + out = self.attn( + query=query, + key=key, + value=value, + attn_mask=attn_mask, + key_padding_mask=key_padding_mask)[0] + + if self.batch_first: + out = out.transpose(0, 1) + + return identity + self.dropout_layer(self.proj_drop(out)) + + +@FEEDFORWARD_NETWORK.register_module() +class FFN(BaseModule): + """Implements feed-forward networks (FFNs) with identity connection. + + Args: + embed_dims (int): The feature dimension. Same as + `MultiheadAttention`. Defaults: 256. + feedforward_channels (int): The hidden dimension of FFNs. + Defaults: 1024. + num_fcs (int, optional): The number of fully-connected layers in + FFNs. Default: 2. + act_cfg (dict, optional): The activation config for FFNs. + Default: dict(type='ReLU') + ffn_drop (float, optional): Probability of an element to be + zeroed in FFN. Default 0.0. + add_identity (bool, optional): Whether to add the + identity connection. Default: `True`. + dropout_layer (obj:`ConfigDict`): The dropout_layer used + when adding the shortcut. + init_cfg (obj:`mmcv.ConfigDict`): The Config for initialization. + Default: None. + """ + + @deprecated_api_warning( + { + 'dropout': 'ffn_drop', + 'add_residual': 'add_identity' + }, + cls_name='FFN') + def __init__(self, + embed_dims=256, + feedforward_channels=1024, + num_fcs=2, + act_cfg=dict(type='ReLU', inplace=True), + ffn_drop=0., + dropout_layer=None, + add_identity=True, + init_cfg=None, + **kwargs): + super(FFN, self).__init__(init_cfg) + assert num_fcs >= 2, 'num_fcs should be no less ' \ + f'than 2. got {num_fcs}.' + self.embed_dims = embed_dims + self.feedforward_channels = feedforward_channels + self.num_fcs = num_fcs + self.act_cfg = act_cfg + self.activate = build_activation_layer(act_cfg) + + layers = [] + in_channels = embed_dims + for _ in range(num_fcs - 1): + layers.append( + Sequential( + Linear(in_channels, feedforward_channels), self.activate, + nn.Dropout(ffn_drop))) + in_channels = feedforward_channels + layers.append(Linear(feedforward_channels, embed_dims)) + layers.append(nn.Dropout(ffn_drop)) + self.layers = Sequential(*layers) + self.dropout_layer = build_dropout( + dropout_layer) if dropout_layer else torch.nn.Identity() + self.add_identity = add_identity + + @deprecated_api_warning({'residual': 'identity'}, cls_name='FFN') + def forward(self, x, identity=None): + """Forward function for `FFN`. + + The function would add x to the output tensor if residue is None. + """ + out = self.layers(x) + if not self.add_identity: + return self.dropout_layer(out) + if identity is None: + identity = x + return identity + self.dropout_layer(out) + + +@TRANSFORMER_LAYER.register_module() +class BaseTransformerLayer(BaseModule): + """Base `TransformerLayer` for vision transformer. + + It can be built from `mmcv.ConfigDict` and support more flexible + customization, for example, using any number of `FFN or LN ` and + use different kinds of `attention` by specifying a list of `ConfigDict` + named `attn_cfgs`. It is worth mentioning that it supports `prenorm` + when you specifying `norm` as the first element of `operation_order`. + More details about the `prenorm`: `On Layer Normalization in the + Transformer Architecture `_ . + + Args: + attn_cfgs (list[`mmcv.ConfigDict`] | obj:`mmcv.ConfigDict` | None )): + Configs for `self_attention` or `cross_attention` modules, + The order of the configs in the list should be consistent with + corresponding attentions in operation_order. + If it is a dict, all of the attention modules in operation_order + will be built with this config. Default: None. + ffn_cfgs (list[`mmcv.ConfigDict`] | obj:`mmcv.ConfigDict` | None )): + Configs for FFN, The order of the configs in the list should be + consistent with corresponding ffn in operation_order. + If it is a dict, all of the attention modules in operation_order + will be built with this config. + operation_order (tuple[str]): The execution order of operation + in transformer. Such as ('self_attn', 'norm', 'ffn', 'norm'). + Support `prenorm` when you specifying first element as `norm`. + Default：None. + norm_cfg (dict): Config dict for normalization layer. + Default: dict(type='LN'). + init_cfg (obj:`mmcv.ConfigDict`): The Config for initialization. + Default: None. + batch_first (bool): Key, Query and Value are shape + of (batch, n, embed_dim) + or (n, batch, embed_dim). Default to False. + """ + + def __init__(self, + attn_cfgs=None, + ffn_cfgs=dict( + type='FFN', + embed_dims=256, + feedforward_channels=1024, + num_fcs=2, + ffn_drop=0., + act_cfg=dict(type='ReLU', inplace=True), + ), + operation_order=None, + norm_cfg=dict(type='LN'), + init_cfg=None, + batch_first=False, + **kwargs): + + deprecated_args = dict( + feedforward_channels='feedforward_channels', + ffn_dropout='ffn_drop', + ffn_num_fcs='num_fcs') + for ori_name, new_name in deprecated_args.items(): + if ori_name in kwargs: + warnings.warn( + f'The arguments `{ori_name}` in BaseTransformerLayer ' + f'has been deprecated, now you should set `{new_name}` ' + f'and other FFN related arguments ' + f'to a dict named `ffn_cfgs`. ', DeprecationWarning) + ffn_cfgs[new_name] = kwargs[ori_name] + + super(BaseTransformerLayer, self).__init__(init_cfg) + + self.batch_first = batch_first + + assert set(operation_order) & set( + ['self_attn', 'norm', 'ffn', 'cross_attn']) == \ + set(operation_order), f'The operation_order of' \ + f' {self.__class__.__name__} should ' \ + f'contains all four operation type ' \ + f"{['self_attn', 'norm', 'ffn', 'cross_attn']}" + + num_attn = operation_order.count('self_attn') + operation_order.count( + 'cross_attn') + if isinstance(attn_cfgs, dict): + attn_cfgs = [copy.deepcopy(attn_cfgs) for _ in range(num_attn)] + else: + assert num_attn == len(attn_cfgs), f'The length ' \ + f'of attn_cfg {num_attn} is ' \ + f'not consistent with the number of attention' \ + f'in operation_order {operation_order}.' + + self.num_attn = num_attn + self.operation_order = operation_order + self.norm_cfg = norm_cfg + self.pre_norm = operation_order[0] == 'norm' + self.attentions = ModuleList() + + index = 0 + for operation_name in operation_order: + if operation_name in ['self_attn', 'cross_attn']: + if 'batch_first' in attn_cfgs[index]: + assert self.batch_first == attn_cfgs[index]['batch_first'] + else: + attn_cfgs[index]['batch_first'] = self.batch_first + attention = build_attention(attn_cfgs[index]) + # Some custom attentions used as `self_attn` + # or `cross_attn` can have different behavior. + attention.operation_name = operation_name + self.attentions.append(attention) + index += 1 + + self.embed_dims = self.attentions[0].embed_dims + + self.ffns = ModuleList() + num_ffns = operation_order.count('ffn') + if isinstance(ffn_cfgs, dict): + ffn_cfgs = ConfigDict(ffn_cfgs) + if isinstance(ffn_cfgs, dict): + ffn_cfgs = [copy.deepcopy(ffn_cfgs) for _ in range(num_ffns)] + assert len(ffn_cfgs) == num_ffns + for ffn_index in range(num_ffns): + if 'embed_dims' not in ffn_cfgs[ffn_index]: + ffn_cfgs[ffn_index]['embed_dims'] = self.embed_dims + else: + assert ffn_cfgs[ffn_index]['embed_dims'] == self.embed_dims + self.ffns.append( + build_feedforward_network(ffn_cfgs[ffn_index], + dict(type='FFN'))) + + self.norms = ModuleList() + num_norms = operation_order.count('norm') + for _ in range(num_norms): + self.norms.append(build_norm_layer(norm_cfg, self.embed_dims)[1]) + + def forward(self, + query, + key=None, + value=None, + query_pos=None, + key_pos=None, + attn_masks=None, + query_key_padding_mask=None, + key_padding_mask=None, + **kwargs): + """Forward function for `TransformerDecoderLayer`. + + **kwargs contains some specific arguments of attentions. + + Args: + query (Tensor): The input query with shape + [num_queries, bs, embed_dims] if + self.batch_first is False, else + [bs, num_queries embed_dims]. + key (Tensor): The key tensor with shape [num_keys, bs, + embed_dims] if self.batch_first is False, else + [bs, num_keys, embed_dims] . + value (Tensor): The value tensor with same shape as `key`. + query_pos (Tensor): The positional encoding for `query`. + Default: None. + key_pos (Tensor): The positional encoding for `key`. + Default: None. + attn_masks (List[Tensor] | None): 2D Tensor used in + calculation of corresponding attention. The length of + it should equal to the number of `attention` in + `operation_order`. Default: None. + query_key_padding_mask (Tensor): ByteTensor for `query`, with + shape [bs, num_queries]. Only used in `self_attn` layer. + Defaults to None. + key_padding_mask (Tensor): ByteTensor for `query`, with + shape [bs, num_keys]. Default: None. + + Returns: + Tensor: forwarded results with shape [num_queries, bs, embed_dims]. + """ + + norm_index = 0 + attn_index = 0 + ffn_index = 0 + identity = query + if attn_masks is None: + attn_masks = [None for _ in range(self.num_attn)] + elif isinstance(attn_masks, torch.Tensor): + attn_masks = [ + copy.deepcopy(attn_masks) for _ in range(self.num_attn) + ] + warnings.warn(f'Use same attn_mask in all attentions in ' + f'{self.__class__.__name__} ') + else: + assert len(attn_masks) == self.num_attn, f'The length of ' \ + f'attn_masks {len(attn_masks)} must be equal ' \ + f'to the number of attention in ' \ + f'operation_order {self.num_attn}' + + for layer in self.operation_order: + if layer == 'self_attn': + temp_key = temp_value = query + query = self.attentions[attn_index]( + query, + temp_key, + temp_value, + identity if self.pre_norm else None, + query_pos=query_pos, + key_pos=query_pos, + attn_mask=attn_masks[attn_index], + key_padding_mask=query_key_padding_mask, + **kwargs) + attn_index += 1 + identity = query + + elif layer == 'norm': + query = self.norms[norm_index](query) + norm_index += 1 + + elif layer == 'cross_attn': + query = self.attentions[attn_index]( + query, + key, + value, + identity if self.pre_norm else None, + query_pos=query_pos, + key_pos=key_pos, + attn_mask=attn_masks[attn_index], + key_padding_mask=key_padding_mask, + **kwargs) + attn_index += 1 + identity = query + + elif layer == 'ffn': + query = self.ffns[ffn_index]( + query, identity if self.pre_norm else None) + ffn_index += 1 + + return query + + +@TRANSFORMER_LAYER_SEQUENCE.register_module() +class TransformerLayerSequence(BaseModule): + """Base class for TransformerEncoder and TransformerDecoder in vision + transformer. + + As base-class of Encoder and Decoder in vision transformer. + Support customization such as specifying different kind + of `transformer_layer` in `transformer_coder`. + + Args: + transformerlayer (list[obj:`mmcv.ConfigDict`] | + obj:`mmcv.ConfigDict`): Config of transformerlayer + in TransformerCoder. If it is obj:`mmcv.ConfigDict`, + it would be repeated `num_layer` times to a + list[`mmcv.ConfigDict`]. Default: None. + num_layers (int): The number of `TransformerLayer`. Default: None. + init_cfg (obj:`mmcv.ConfigDict`): The Config for initialization. + Default: None. + """ + + def __init__(self, transformerlayers=None, num_layers=None, init_cfg=None): + super(TransformerLayerSequence, self).__init__(init_cfg) + if isinstance(transformerlayers, dict): + transformerlayers = [ + copy.deepcopy(transformerlayers) for _ in range(num_layers) + ] + else: + assert isinstance(transformerlayers, list) and \ + len(transformerlayers) == num_layers + self.num_layers = num_layers + self.layers = ModuleList() + for i in range(num_layers): + self.layers.append(build_transformer_layer(transformerlayers[i])) + self.embed_dims = self.layers[0].embed_dims + self.pre_norm = self.layers[0].pre_norm + + def forward(self, + query, + key, + value, + query_pos=None, + key_pos=None, + attn_masks=None, + query_key_padding_mask=None, + key_padding_mask=None, + **kwargs): + """Forward function for `TransformerCoder`. + + Args: + query (Tensor): Input query with shape + `(num_queries, bs, embed_dims)`. + key (Tensor): The key tensor with shape + `(num_keys, bs, embed_dims)`. + value (Tensor): The value tensor with shape + `(num_keys, bs, embed_dims)`. + query_pos (Tensor): The positional encoding for `query`. + Default: None. + key_pos (Tensor): The positional encoding for `key`. + Default: None. + attn_masks (List[Tensor], optional): Each element is 2D Tensor + which is used in calculation of corresponding attention in + operation_order. Default: None. + query_key_padding_mask (Tensor): ByteTensor for `query`, with + shape [bs, num_queries]. Only used in self-attention + Default: None. + key_padding_mask (Tensor): ByteTensor for `query`, with + shape [bs, num_keys]. Default: None. + + Returns: + Tensor: results with shape [num_queries, bs, embed_dims]. + """ + for layer in self.layers: + query = layer( + query, + key, + value, + query_pos=query_pos, + key_pos=key_pos, + attn_masks=attn_masks, + query_key_padding_mask=query_key_padding_mask, + key_padding_mask=key_padding_mask, + **kwargs) + return query diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/upsample.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/upsample.py new file mode 100644 index 0000000000000000000000000000000000000000..0fd21fbf9dccca42f520d5474e22ac1e62048d57 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/upsample.py @@ -0,0 +1,84 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn +import torch.nn.functional as F + +from ..utils import xavier_init +from .registry import UPSAMPLE_LAYERS + +UPSAMPLE_LAYERS.register_module('nearest', module=nn.Upsample) +UPSAMPLE_LAYERS.register_module('bilinear', module=nn.Upsample) + + +@UPSAMPLE_LAYERS.register_module(name='pixel_shuffle') +class PixelShufflePack(nn.Module): + """Pixel Shuffle upsample layer. + + This module packs `F.pixel_shuffle()` and a nn.Conv2d module together to + achieve a simple upsampling with pixel shuffle. + + Args: + in_channels (int): Number of input channels. + out_channels (int): Number of output channels. + scale_factor (int): Upsample ratio. + upsample_kernel (int): Kernel size of the conv layer to expand the + channels. + """ + + def __init__(self, in_channels, out_channels, scale_factor, + upsample_kernel): + super(PixelShufflePack, self).__init__() + self.in_channels = in_channels + self.out_channels = out_channels + self.scale_factor = scale_factor + self.upsample_kernel = upsample_kernel + self.upsample_conv = nn.Conv2d( + self.in_channels, + self.out_channels * scale_factor * scale_factor, + self.upsample_kernel, + padding=(self.upsample_kernel - 1) // 2) + self.init_weights() + + def init_weights(self): + xavier_init(self.upsample_conv, distribution='uniform') + + def forward(self, x): + x = self.upsample_conv(x) + x = F.pixel_shuffle(x, self.scale_factor) + return x + + +def build_upsample_layer(cfg, *args, **kwargs): + """Build upsample layer. + + Args: + cfg (dict): The upsample layer config, which should contain: + + - type (str): Layer type. + - scale_factor (int): Upsample ratio, which is not applicable to + deconv. + - layer args: Args needed to instantiate a upsample layer. + args (argument list): Arguments passed to the ``__init__`` + method of the corresponding conv layer. + kwargs (keyword arguments): Keyword arguments passed to the + ``__init__`` method of the corresponding conv layer. + + Returns: + nn.Module: Created upsample layer. + """ + if not isinstance(cfg, dict): + raise TypeError(f'cfg must be a dict, but got {type(cfg)}') + if 'type' not in cfg: + raise KeyError( + f'the cfg dict must contain the key "type", but got {cfg}') + cfg_ = cfg.copy() + + layer_type = cfg_.pop('type') + if layer_type not in UPSAMPLE_LAYERS: + raise KeyError(f'Unrecognized upsample type {layer_type}') + else: + upsample = UPSAMPLE_LAYERS.get(layer_type) + + if upsample is nn.Upsample: + cfg_['mode'] = layer_type + layer = upsample(*args, **kwargs, **cfg_) + return layer diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/wrappers.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/wrappers.py new file mode 100644 index 0000000000000000000000000000000000000000..8aebf67bf52355a513f21756ee74fe510902d075 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/bricks/wrappers.py @@ -0,0 +1,180 @@ +# Copyright (c) OpenMMLab. All rights reserved. +r"""Modified from https://github.com/facebookresearch/detectron2/blob/master/detectron2/layers/wrappers.py # noqa: E501 + +Wrap some nn modules to support empty tensor input. Currently, these wrappers +are mainly used in mask heads like fcn_mask_head and maskiou_heads since mask +heads are trained on only positive RoIs. +""" +import math + +import torch +import torch.nn as nn +from torch.nn.modules.utils import _pair, _triple + +from .registry import CONV_LAYERS, UPSAMPLE_LAYERS + +if torch.__version__ == 'parrots': + TORCH_VERSION = torch.__version__ +else: + # torch.__version__ could be 1.3.1+cu92, we only need the first two + # for comparison + TORCH_VERSION = tuple(int(x) for x in torch.__version__.split('.')[:2]) + + +def obsolete_torch_version(torch_version, version_threshold): + return torch_version == 'parrots' or torch_version <= version_threshold + + +class NewEmptyTensorOp(torch.autograd.Function): + + @staticmethod + def forward(ctx, x, new_shape): + ctx.shape = x.shape + return x.new_empty(new_shape) + + @staticmethod + def backward(ctx, grad): + shape = ctx.shape + return NewEmptyTensorOp.apply(grad, shape), None + + +@CONV_LAYERS.register_module('Conv', force=True) +class Conv2d(nn.Conv2d): + + def forward(self, x): + if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 4)): + out_shape = [x.shape[0], self.out_channels] + for i, k, p, s, d in zip(x.shape[-2:], self.kernel_size, + self.padding, self.stride, self.dilation): + o = (i + 2 * p - (d * (k - 1) + 1)) // s + 1 + out_shape.append(o) + empty = NewEmptyTensorOp.apply(x, out_shape) + if self.training: + # produce dummy gradient to avoid DDP warning. + dummy = sum(x.view(-1)[0] for x in self.parameters()) * 0.0 + return empty + dummy + else: + return empty + + return super().forward(x) + + +@CONV_LAYERS.register_module('Conv3d', force=True) +class Conv3d(nn.Conv3d): + + def forward(self, x): + if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 4)): + out_shape = [x.shape[0], self.out_channels] + for i, k, p, s, d in zip(x.shape[-3:], self.kernel_size, + self.padding, self.stride, self.dilation): + o = (i + 2 * p - (d * (k - 1) + 1)) // s + 1 + out_shape.append(o) + empty = NewEmptyTensorOp.apply(x, out_shape) + if self.training: + # produce dummy gradient to avoid DDP warning. + dummy = sum(x.view(-1)[0] for x in self.parameters()) * 0.0 + return empty + dummy + else: + return empty + + return super().forward(x) + + +@CONV_LAYERS.register_module() +@CONV_LAYERS.register_module('deconv') +@UPSAMPLE_LAYERS.register_module('deconv', force=True) +class ConvTranspose2d(nn.ConvTranspose2d): + + def forward(self, x): + if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 4)): + out_shape = [x.shape[0], self.out_channels] + for i, k, p, s, d, op in zip(x.shape[-2:], self.kernel_size, + self.padding, self.stride, + self.dilation, self.output_padding): + out_shape.append((i - 1) * s - 2 * p + (d * (k - 1) + 1) + op) + empty = NewEmptyTensorOp.apply(x, out_shape) + if self.training: + # produce dummy gradient to avoid DDP warning. + dummy = sum(x.view(-1)[0] for x in self.parameters()) * 0.0 + return empty + dummy + else: + return empty + + return super().forward(x) + + +@CONV_LAYERS.register_module() +@CONV_LAYERS.register_module('deconv3d') +@UPSAMPLE_LAYERS.register_module('deconv3d', force=True) +class ConvTranspose3d(nn.ConvTranspose3d): + + def forward(self, x): + if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 4)): + out_shape = [x.shape[0], self.out_channels] + for i, k, p, s, d, op in zip(x.shape[-3:], self.kernel_size, + self.padding, self.stride, + self.dilation, self.output_padding): + out_shape.append((i - 1) * s - 2 * p + (d * (k - 1) + 1) + op) + empty = NewEmptyTensorOp.apply(x, out_shape) + if self.training: + # produce dummy gradient to avoid DDP warning. + dummy = sum(x.view(-1)[0] for x in self.parameters()) * 0.0 + return empty + dummy + else: + return empty + + return super().forward(x) + + +class MaxPool2d(nn.MaxPool2d): + + def forward(self, x): + # PyTorch 1.9 does not support empty tensor inference yet + if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 9)): + out_shape = list(x.shape[:2]) + for i, k, p, s, d in zip(x.shape[-2:], _pair(self.kernel_size), + _pair(self.padding), _pair(self.stride), + _pair(self.dilation)): + o = (i + 2 * p - (d * (k - 1) + 1)) / s + 1 + o = math.ceil(o) if self.ceil_mode else math.floor(o) + out_shape.append(o) + empty = NewEmptyTensorOp.apply(x, out_shape) + return empty + + return super().forward(x) + + +class MaxPool3d(nn.MaxPool3d): + + def forward(self, x): + # PyTorch 1.9 does not support empty tensor inference yet + if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 9)): + out_shape = list(x.shape[:2]) + for i, k, p, s, d in zip(x.shape[-3:], _triple(self.kernel_size), + _triple(self.padding), + _triple(self.stride), + _triple(self.dilation)): + o = (i + 2 * p - (d * (k - 1) + 1)) / s + 1 + o = math.ceil(o) if self.ceil_mode else math.floor(o) + out_shape.append(o) + empty = NewEmptyTensorOp.apply(x, out_shape) + return empty + + return super().forward(x) + + +class Linear(torch.nn.Linear): + + def forward(self, x): + # empty tensor forward of Linear layer is supported in Pytorch 1.6 + if x.numel() == 0 and obsolete_torch_version(TORCH_VERSION, (1, 5)): + out_shape = [x.shape[0], self.out_features] + empty = NewEmptyTensorOp.apply(x, out_shape) + if self.training: + # produce dummy gradient to avoid DDP warning. + dummy = sum(x.view(-1)[0] for x in self.parameters()) * 0.0 + return empty + dummy + else: + return empty + + return super().forward(x) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/builder.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/builder.py new file mode 100644 index 0000000000000000000000000000000000000000..7567316c566bd3aca6d8f65a84b00e9e890948a7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/builder.py @@ -0,0 +1,30 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from ..runner import Sequential +from ..utils import Registry, build_from_cfg + + +def build_model_from_cfg(cfg, registry, default_args=None): + """Build a PyTorch model from config dict(s). Different from + ``build_from_cfg``, if cfg is a list, a ``nn.Sequential`` will be built. + + Args: + cfg (dict, list[dict]): The config of modules, is is either a config + dict or a list of config dicts. If cfg is a list, a + the built modules will be wrapped with ``nn.Sequential``. + registry (:obj:`Registry`): A registry the module belongs to. + default_args (dict, optional): Default arguments to build the module. + Defaults to None. + + Returns: + nn.Module: A built nn module. + """ + if isinstance(cfg, list): + modules = [ + build_from_cfg(cfg_, registry, default_args) for cfg_ in cfg + ] + return Sequential(*modules) + else: + return build_from_cfg(cfg, registry, default_args) + + +MODELS = Registry('model', build_func=build_model_from_cfg) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..a263e31c1e3977712827ca229bbc04910b4e928e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/__init__.py @@ -0,0 +1,19 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .flops_counter import get_model_complexity_info +from .fuse_conv_bn import fuse_conv_bn +from .sync_bn import revert_sync_batchnorm +from .weight_init import (INITIALIZERS, Caffe2XavierInit, ConstantInit, + KaimingInit, NormalInit, PretrainedInit, + TruncNormalInit, UniformInit, XavierInit, + bias_init_with_prob, caffe2_xavier_init, + constant_init, initialize, kaiming_init, normal_init, + trunc_normal_init, uniform_init, xavier_init) + +__all__ = [ + 'get_model_complexity_info', 'bias_init_with_prob', 'caffe2_xavier_init', + 'constant_init', 'kaiming_init', 'normal_init', 'trunc_normal_init', + 'uniform_init', 'xavier_init', 'fuse_conv_bn', 'initialize', + 'INITIALIZERS', 'ConstantInit', 'XavierInit', 'NormalInit', + 'TruncNormalInit', 'UniformInit', 'KaimingInit', 'PretrainedInit', + 'Caffe2XavierInit', 'revert_sync_batchnorm' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/flops_counter.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/flops_counter.py new file mode 100644 index 0000000000000000000000000000000000000000..a6045db847883b539c2464f4359af5583707b053 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/flops_counter.py @@ -0,0 +1,599 @@ +# Modified from flops-counter.pytorch by Vladislav Sovrasov +# original repo: https://github.com/sovrasov/flops-counter.pytorch + +# MIT License + +# Copyright (c) 2018 Vladislav Sovrasov + +# Permission is hereby granted, free of charge, to any person obtaining a copy +# of this software and associated documentation files (the "Software"), to deal +# in the Software without restriction, including without limitation the rights +# to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +# copies of the Software, and to permit persons to whom the Software is +# furnished to do so, subject to the following conditions: + +# The above copyright notice and this permission notice shall be included in +# all copies or substantial portions of the Software. + +# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +# SOFTWARE. + +import sys +import warnings +from functools import partial + +import numpy as np +import torch +import torch.nn as nn + +import mmcv + + +def get_model_complexity_info(model, + input_shape, + print_per_layer_stat=True, + as_strings=True, + input_constructor=None, + flush=False, + ost=sys.stdout): + """Get complexity information of a model. + + This method can calculate FLOPs and parameter counts of a model with + corresponding input shape. It can also print complexity information for + each layer in a model. + + Supported layers are listed as below: + - Convolutions: ``nn.Conv1d``, ``nn.Conv2d``, ``nn.Conv3d``. + - Activations: ``nn.ReLU``, ``nn.PReLU``, ``nn.ELU``, + ``nn.LeakyReLU``, ``nn.ReLU6``. + - Poolings: ``nn.MaxPool1d``, ``nn.MaxPool2d``, ``nn.MaxPool3d``, + ``nn.AvgPool1d``, ``nn.AvgPool2d``, ``nn.AvgPool3d``, + ``nn.AdaptiveMaxPool1d``, ``nn.AdaptiveMaxPool2d``, + ``nn.AdaptiveMaxPool3d``, ``nn.AdaptiveAvgPool1d``, + ``nn.AdaptiveAvgPool2d``, ``nn.AdaptiveAvgPool3d``. + - BatchNorms: ``nn.BatchNorm1d``, ``nn.BatchNorm2d``, + ``nn.BatchNorm3d``, ``nn.GroupNorm``, ``nn.InstanceNorm1d``, + ``InstanceNorm2d``, ``InstanceNorm3d``, ``nn.LayerNorm``. + - Linear: ``nn.Linear``. + - Deconvolution: ``nn.ConvTranspose2d``. + - Upsample: ``nn.Upsample``. + + Args: + model (nn.Module): The model for complexity calculation. + input_shape (tuple): Input shape used for calculation. + print_per_layer_stat (bool): Whether to print complexity information + for each layer in a model. Default: True. + as_strings (bool): Output FLOPs and params counts in a string form. + Default: True. + input_constructor (None | callable): If specified, it takes a callable + method that generates input. otherwise, it will generate a random + tensor with input shape to calculate FLOPs. Default: None. + flush (bool): same as that in :func:`print`. Default: False. + ost (stream): same as ``file`` param in :func:`print`. + Default: sys.stdout. + + Returns: + tuple[float | str]: If ``as_strings`` is set to True, it will return + FLOPs and parameter counts in a string format. otherwise, it will + return those in a float number format. + """ + assert type(input_shape) is tuple + assert len(input_shape) >= 1 + assert isinstance(model, nn.Module) + flops_model = add_flops_counting_methods(model) + flops_model.eval() + flops_model.start_flops_count() + if input_constructor: + input = input_constructor(input_shape) + _ = flops_model(**input) + else: + try: + batch = torch.ones(()).new_empty( + (1, *input_shape), + dtype=next(flops_model.parameters()).dtype, + device=next(flops_model.parameters()).device) + except StopIteration: + # Avoid StopIteration for models which have no parameters, + # like `nn.Relu()`, `nn.AvgPool2d`, etc. + batch = torch.ones(()).new_empty((1, *input_shape)) + + _ = flops_model(batch) + + flops_count, params_count = flops_model.compute_average_flops_cost() + if print_per_layer_stat: + print_model_with_flops( + flops_model, flops_count, params_count, ost=ost, flush=flush) + flops_model.stop_flops_count() + + if as_strings: + return flops_to_string(flops_count), params_to_string(params_count) + + return flops_count, params_count + + +def flops_to_string(flops, units='GFLOPs', precision=2): + """Convert FLOPs number into a string. + + Note that Here we take a multiply-add counts as one FLOP. + + Args: + flops (float): FLOPs number to be converted. + units (str | None): Converted FLOPs units. Options are None, 'GFLOPs', + 'MFLOPs', 'KFLOPs', 'FLOPs'. If set to None, it will automatically + choose the most suitable unit for FLOPs. Default: 'GFLOPs'. + precision (int): Digit number after the decimal point. Default: 2. + + Returns: + str: The converted FLOPs number with units. + + Examples: + >>> flops_to_string(1e9) + '1.0 GFLOPs' + >>> flops_to_string(2e5, 'MFLOPs') + '0.2 MFLOPs' + >>> flops_to_string(3e-9, None) + '3e-09 FLOPs' + """ + if units is None: + if flops // 10**9 > 0: + return str(round(flops / 10.**9, precision)) + ' GFLOPs' + elif flops // 10**6 > 0: + return str(round(flops / 10.**6, precision)) + ' MFLOPs' + elif flops // 10**3 > 0: + return str(round(flops / 10.**3, precision)) + ' KFLOPs' + else: + return str(flops) + ' FLOPs' + else: + if units == 'GFLOPs': + return str(round(flops / 10.**9, precision)) + ' ' + units + elif units == 'MFLOPs': + return str(round(flops / 10.**6, precision)) + ' ' + units + elif units == 'KFLOPs': + return str(round(flops / 10.**3, precision)) + ' ' + units + else: + return str(flops) + ' FLOPs' + + +def params_to_string(num_params, units=None, precision=2): + """Convert parameter number into a string. + + Args: + num_params (float): Parameter number to be converted. + units (str | None): Converted FLOPs units. Options are None, 'M', + 'K' and ''. If set to None, it will automatically choose the most + suitable unit for Parameter number. Default: None. + precision (int): Digit number after the decimal point. Default: 2. + + Returns: + str: The converted parameter number with units. + + Examples: + >>> params_to_string(1e9) + '1000.0 M' + >>> params_to_string(2e5) + '200.0 k' + >>> params_to_string(3e-9) + '3e-09' + """ + if units is None: + if num_params // 10**6 > 0: + return str(round(num_params / 10**6, precision)) + ' M' + elif num_params // 10**3: + return str(round(num_params / 10**3, precision)) + ' k' + else: + return str(num_params) + else: + if units == 'M': + return str(round(num_params / 10.**6, precision)) + ' ' + units + elif units == 'K': + return str(round(num_params / 10.**3, precision)) + ' ' + units + else: + return str(num_params) + + +def print_model_with_flops(model, + total_flops, + total_params, + units='GFLOPs', + precision=3, + ost=sys.stdout, + flush=False): + """Print a model with FLOPs for each layer. + + Args: + model (nn.Module): The model to be printed. + total_flops (float): Total FLOPs of the model. + total_params (float): Total parameter counts of the model. + units (str | None): Converted FLOPs units. Default: 'GFLOPs'. + precision (int): Digit number after the decimal point. Default: 3. + ost (stream): same as `file` param in :func:`print`. + Default: sys.stdout. + flush (bool): same as that in :func:`print`. Default: False. + + Example: + >>> class ExampleModel(nn.Module): + + >>> def __init__(self): + >>> super().__init__() + >>> self.conv1 = nn.Conv2d(3, 8, 3) + >>> self.conv2 = nn.Conv2d(8, 256, 3) + >>> self.conv3 = nn.Conv2d(256, 8, 3) + >>> self.avg_pool = nn.AdaptiveAvgPool2d((1, 1)) + >>> self.flatten = nn.Flatten() + >>> self.fc = nn.Linear(8, 1) + + >>> def forward(self, x): + >>> x = self.conv1(x) + >>> x = self.conv2(x) + >>> x = self.conv3(x) + >>> x = self.avg_pool(x) + >>> x = self.flatten(x) + >>> x = self.fc(x) + >>> return x + + >>> model = ExampleModel() + >>> x = (3, 16, 16) + to print the complexity information state for each layer, you can use + >>> get_model_complexity_info(model, x) + or directly use + >>> print_model_with_flops(model, 4579784.0, 37361) + ExampleModel( + 0.037 M, 100.000% Params, 0.005 GFLOPs, 100.000% FLOPs, + (conv1): Conv2d(0.0 M, 0.600% Params, 0.0 GFLOPs, 0.959% FLOPs, 3, 8, kernel_size=(3, 3), stride=(1, 1)) # noqa: E501 + (conv2): Conv2d(0.019 M, 50.020% Params, 0.003 GFLOPs, 58.760% FLOPs, 8, 256, kernel_size=(3, 3), stride=(1, 1)) + (conv3): Conv2d(0.018 M, 49.356% Params, 0.002 GFLOPs, 40.264% FLOPs, 256, 8, kernel_size=(3, 3), stride=(1, 1)) + (avg_pool): AdaptiveAvgPool2d(0.0 M, 0.000% Params, 0.0 GFLOPs, 0.017% FLOPs, output_size=(1, 1)) + (flatten): Flatten(0.0 M, 0.000% Params, 0.0 GFLOPs, 0.000% FLOPs, ) + (fc): Linear(0.0 M, 0.024% Params, 0.0 GFLOPs, 0.000% FLOPs, in_features=8, out_features=1, bias=True) + ) + """ + + def accumulate_params(self): + if is_supported_instance(self): + return self.__params__ + else: + sum = 0 + for m in self.children(): + sum += m.accumulate_params() + return sum + + def accumulate_flops(self): + if is_supported_instance(self): + return self.__flops__ / model.__batch_counter__ + else: + sum = 0 + for m in self.children(): + sum += m.accumulate_flops() + return sum + + def flops_repr(self): + accumulated_num_params = self.accumulate_params() + accumulated_flops_cost = self.accumulate_flops() + return ', '.join([ + params_to_string( + accumulated_num_params, units='M', precision=precision), + '{:.3%} Params'.format(accumulated_num_params / total_params), + flops_to_string( + accumulated_flops_cost, units=units, precision=precision), + '{:.3%} FLOPs'.format(accumulated_flops_cost / total_flops), + self.original_extra_repr() + ]) + + def add_extra_repr(m): + m.accumulate_flops = accumulate_flops.__get__(m) + m.accumulate_params = accumulate_params.__get__(m) + flops_extra_repr = flops_repr.__get__(m) + if m.extra_repr != flops_extra_repr: + m.original_extra_repr = m.extra_repr + m.extra_repr = flops_extra_repr + assert m.extra_repr != m.original_extra_repr + + def del_extra_repr(m): + if hasattr(m, 'original_extra_repr'): + m.extra_repr = m.original_extra_repr + del m.original_extra_repr + if hasattr(m, 'accumulate_flops'): + del m.accumulate_flops + + model.apply(add_extra_repr) + print(model, file=ost, flush=flush) + model.apply(del_extra_repr) + + +def get_model_parameters_number(model): + """Calculate parameter number of a model. + + Args: + model (nn.module): The model for parameter number calculation. + + Returns: + float: Parameter number of the model. + """ + num_params = sum(p.numel() for p in model.parameters() if p.requires_grad) + return num_params + + +def add_flops_counting_methods(net_main_module): + # adding additional methods to the existing module object, + # this is done this way so that each function has access to self object + net_main_module.start_flops_count = start_flops_count.__get__( + net_main_module) + net_main_module.stop_flops_count = stop_flops_count.__get__( + net_main_module) + net_main_module.reset_flops_count = reset_flops_count.__get__( + net_main_module) + net_main_module.compute_average_flops_cost = compute_average_flops_cost.__get__( # noqa: E501 + net_main_module) + + net_main_module.reset_flops_count() + + return net_main_module + + +def compute_average_flops_cost(self): + """Compute average FLOPs cost. + + A method to compute average FLOPs cost, which will be available after + `add_flops_counting_methods()` is called on a desired net object. + + Returns: + float: Current mean flops consumption per image. + """ + batches_count = self.__batch_counter__ + flops_sum = 0 + for module in self.modules(): + if is_supported_instance(module): + flops_sum += module.__flops__ + params_sum = get_model_parameters_number(self) + return flops_sum / batches_count, params_sum + + +def start_flops_count(self): + """Activate the computation of mean flops consumption per image. + + A method to activate the computation of mean flops consumption per image. + which will be available after ``add_flops_counting_methods()`` is called on + a desired net object. It should be called before running the network. + """ + add_batch_counter_hook_function(self) + + def add_flops_counter_hook_function(module): + if is_supported_instance(module): + if hasattr(module, '__flops_handle__'): + return + + else: + handle = module.register_forward_hook( + get_modules_mapping()[type(module)]) + + module.__flops_handle__ = handle + + self.apply(partial(add_flops_counter_hook_function)) + + +def stop_flops_count(self): + """Stop computing the mean flops consumption per image. + + A method to stop computing the mean flops consumption per image, which will + be available after ``add_flops_counting_methods()`` is called on a desired + net object. It can be called to pause the computation whenever. + """ + remove_batch_counter_hook_function(self) + self.apply(remove_flops_counter_hook_function) + + +def reset_flops_count(self): + """Reset statistics computed so far. + + A method to Reset computed statistics, which will be available after + `add_flops_counting_methods()` is called on a desired net object. + """ + add_batch_counter_variables_or_reset(self) + self.apply(add_flops_counter_variable_or_reset) + + +# ---- Internal functions +def empty_flops_counter_hook(module, input, output): + module.__flops__ += 0 + + +def upsample_flops_counter_hook(module, input, output): + output_size = output[0] + batch_size = output_size.shape[0] + output_elements_count = batch_size + for val in output_size.shape[1:]: + output_elements_count *= val + module.__flops__ += int(output_elements_count) + + +def relu_flops_counter_hook(module, input, output): + active_elements_count = output.numel() + module.__flops__ += int(active_elements_count) + + +def linear_flops_counter_hook(module, input, output): + input = input[0] + output_last_dim = output.shape[ + -1] # pytorch checks dimensions, so here we don't care much + module.__flops__ += int(np.prod(input.shape) * output_last_dim) + + +def pool_flops_counter_hook(module, input, output): + input = input[0] + module.__flops__ += int(np.prod(input.shape)) + + +def norm_flops_counter_hook(module, input, output): + input = input[0] + + batch_flops = np.prod(input.shape) + if (getattr(module, 'affine', False) + or getattr(module, 'elementwise_affine', False)): + batch_flops *= 2 + module.__flops__ += int(batch_flops) + + +def deconv_flops_counter_hook(conv_module, input, output): + # Can have multiple inputs, getting the first one + input = input[0] + + batch_size = input.shape[0] + input_height, input_width = input.shape[2:] + + kernel_height, kernel_width = conv_module.kernel_size + in_channels = conv_module.in_channels + out_channels = conv_module.out_channels + groups = conv_module.groups + + filters_per_channel = out_channels // groups + conv_per_position_flops = ( + kernel_height * kernel_width * in_channels * filters_per_channel) + + active_elements_count = batch_size * input_height * input_width + overall_conv_flops = conv_per_position_flops * active_elements_count + bias_flops = 0 + if conv_module.bias is not None: + output_height, output_width = output.shape[2:] + bias_flops = out_channels * batch_size * output_height * output_width + overall_flops = overall_conv_flops + bias_flops + + conv_module.__flops__ += int(overall_flops) + + +def conv_flops_counter_hook(conv_module, input, output): + # Can have multiple inputs, getting the first one + input = input[0] + + batch_size = input.shape[0] + output_dims = list(output.shape[2:]) + + kernel_dims = list(conv_module.kernel_size) + in_channels = conv_module.in_channels + out_channels = conv_module.out_channels + groups = conv_module.groups + + filters_per_channel = out_channels // groups + conv_per_position_flops = int( + np.prod(kernel_dims)) * in_channels * filters_per_channel + + active_elements_count = batch_size * int(np.prod(output_dims)) + + overall_conv_flops = conv_per_position_flops * active_elements_count + + bias_flops = 0 + + if conv_module.bias is not None: + + bias_flops = out_channels * active_elements_count + + overall_flops = overall_conv_flops + bias_flops + + conv_module.__flops__ += int(overall_flops) + + +def batch_counter_hook(module, input, output): + batch_size = 1 + if len(input) > 0: + # Can have multiple inputs, getting the first one + input = input[0] + batch_size = len(input) + else: + warnings.warn('No positional inputs found for a module, ' + 'assuming batch size is 1.') + module.__batch_counter__ += batch_size + + +def add_batch_counter_variables_or_reset(module): + + module.__batch_counter__ = 0 + + +def add_batch_counter_hook_function(module): + if hasattr(module, '__batch_counter_handle__'): + return + + handle = module.register_forward_hook(batch_counter_hook) + module.__batch_counter_handle__ = handle + + +def remove_batch_counter_hook_function(module): + if hasattr(module, '__batch_counter_handle__'): + module.__batch_counter_handle__.remove() + del module.__batch_counter_handle__ + + +def add_flops_counter_variable_or_reset(module): + if is_supported_instance(module): + if hasattr(module, '__flops__') or hasattr(module, '__params__'): + warnings.warn('variables __flops__ or __params__ are already ' + 'defined for the module' + type(module).__name__ + + ' ptflops can affect your code!') + module.__flops__ = 0 + module.__params__ = get_model_parameters_number(module) + + +def is_supported_instance(module): + if type(module) in get_modules_mapping(): + return True + return False + + +def remove_flops_counter_hook_function(module): + if is_supported_instance(module): + if hasattr(module, '__flops_handle__'): + module.__flops_handle__.remove() + del module.__flops_handle__ + + +def get_modules_mapping(): + return { + # convolutions + nn.Conv1d: conv_flops_counter_hook, + nn.Conv2d: conv_flops_counter_hook, + mmcv.cnn.bricks.Conv2d: conv_flops_counter_hook, + nn.Conv3d: conv_flops_counter_hook, + mmcv.cnn.bricks.Conv3d: conv_flops_counter_hook, + # activations + nn.ReLU: relu_flops_counter_hook, + nn.PReLU: relu_flops_counter_hook, + nn.ELU: relu_flops_counter_hook, + nn.LeakyReLU: relu_flops_counter_hook, + nn.ReLU6: relu_flops_counter_hook, + # poolings + nn.MaxPool1d: pool_flops_counter_hook, + nn.AvgPool1d: pool_flops_counter_hook, + nn.AvgPool2d: pool_flops_counter_hook, + nn.MaxPool2d: pool_flops_counter_hook, + mmcv.cnn.bricks.MaxPool2d: pool_flops_counter_hook, + nn.MaxPool3d: pool_flops_counter_hook, + mmcv.cnn.bricks.MaxPool3d: pool_flops_counter_hook, + nn.AvgPool3d: pool_flops_counter_hook, + nn.AdaptiveMaxPool1d: pool_flops_counter_hook, + nn.AdaptiveAvgPool1d: pool_flops_counter_hook, + nn.AdaptiveMaxPool2d: pool_flops_counter_hook, + nn.AdaptiveAvgPool2d: pool_flops_counter_hook, + nn.AdaptiveMaxPool3d: pool_flops_counter_hook, + nn.AdaptiveAvgPool3d: pool_flops_counter_hook, + # normalizations + nn.BatchNorm1d: norm_flops_counter_hook, + nn.BatchNorm2d: norm_flops_counter_hook, + nn.BatchNorm3d: norm_flops_counter_hook, + nn.GroupNorm: norm_flops_counter_hook, + nn.InstanceNorm1d: norm_flops_counter_hook, + nn.InstanceNorm2d: norm_flops_counter_hook, + nn.InstanceNorm3d: norm_flops_counter_hook, + nn.LayerNorm: norm_flops_counter_hook, + # FC + nn.Linear: linear_flops_counter_hook, + mmcv.cnn.bricks.Linear: linear_flops_counter_hook, + # Upscale + nn.Upsample: upsample_flops_counter_hook, + # Deconvolution + nn.ConvTranspose2d: deconv_flops_counter_hook, + mmcv.cnn.bricks.ConvTranspose2d: deconv_flops_counter_hook, + } diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/fuse_conv_bn.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/fuse_conv_bn.py new file mode 100644 index 0000000000000000000000000000000000000000..cb7076f80bf37f7931185bf0293ffcc1ce19c8ef --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/fuse_conv_bn.py @@ -0,0 +1,59 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn + + +def _fuse_conv_bn(conv, bn): + """Fuse conv and bn into one module. + + Args: + conv (nn.Module): Conv to be fused. + bn (nn.Module): BN to be fused. + + Returns: + nn.Module: Fused module. + """ + conv_w = conv.weight + conv_b = conv.bias if conv.bias is not None else torch.zeros_like( + bn.running_mean) + + factor = bn.weight / torch.sqrt(bn.running_var + bn.eps) + conv.weight = nn.Parameter(conv_w * + factor.reshape([conv.out_channels, 1, 1, 1])) + conv.bias = nn.Parameter((conv_b - bn.running_mean) * factor + bn.bias) + return conv + + +def fuse_conv_bn(module): + """Recursively fuse conv and bn in a module. + + During inference, the functionary of batch norm layers is turned off + but only the mean and var alone channels are used, which exposes the + chance to fuse it with the preceding conv layers to save computations and + simplify network structures. + + Args: + module (nn.Module): Module to be fused. + + Returns: + nn.Module: Fused module. + """ + last_conv = None + last_conv_name = None + + for name, child in module.named_children(): + if isinstance(child, + (nn.modules.batchnorm._BatchNorm, nn.SyncBatchNorm)): + if last_conv is None: # only fuse BN that is after Conv + continue + fused_conv = _fuse_conv_bn(last_conv, child) + module._modules[last_conv_name] = fused_conv + # To reduce changes, set BN as Identity instead of deleting it. + module._modules[name] = nn.Identity() + last_conv = None + elif isinstance(child, nn.Conv2d): + last_conv = child + last_conv_name = name + else: + fuse_conv_bn(child) + return module diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/sync_bn.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/sync_bn.py new file mode 100644 index 0000000000000000000000000000000000000000..0c52526e95c4d7b3b0c0b7f70690326e5ae55169 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/sync_bn.py @@ -0,0 +1,60 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch + +import mmcv + + +class _BatchNormXd(torch.nn.modules.batchnorm._BatchNorm): + """A general BatchNorm layer without input dimension check. + + Reproduced from @kapily's work: + (https://github.com/pytorch/pytorch/issues/41081#issuecomment-783961547) + The only difference between BatchNorm1d, BatchNorm2d, BatchNorm3d, etc + is `_check_input_dim` that is designed for tensor sanity checks. + The check has been bypassed in this class for the convenience of converting + SyncBatchNorm. + """ + + def _check_input_dim(self, input): + return + + +def revert_sync_batchnorm(module): + """Helper function to convert all `SyncBatchNorm` (SyncBN) and + `mmcv.ops.sync_bn.SyncBatchNorm`(MMSyncBN) layers in the model to + `BatchNormXd` layers. + + Adapted from @kapily's work: + (https://github.com/pytorch/pytorch/issues/41081#issuecomment-783961547) + + Args: + module (nn.Module): The module containing `SyncBatchNorm` layers. + + Returns: + module_output: The converted module with `BatchNormXd` layers. + """ + module_output = module + module_checklist = [torch.nn.modules.batchnorm.SyncBatchNorm] + if hasattr(mmcv, 'ops'): + module_checklist.append(mmcv.ops.SyncBatchNorm) + if isinstance(module, tuple(module_checklist)): + module_output = _BatchNormXd(module.num_features, module.eps, + module.momentum, module.affine, + module.track_running_stats) + if module.affine: + # no_grad() may not be needed here but + # just to be consistent with `convert_sync_batchnorm()` + with torch.no_grad(): + module_output.weight = module.weight + module_output.bias = module.bias + module_output.running_mean = module.running_mean + module_output.running_var = module.running_var + module_output.num_batches_tracked = module.num_batches_tracked + module_output.training = module.training + # qconfig exists in quantized models + if hasattr(module, 'qconfig'): + module_output.qconfig = module.qconfig + for name, child in module.named_children(): + module_output.add_module(name, revert_sync_batchnorm(child)) + del module + return module_output diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/weight_init.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/weight_init.py new file mode 100644 index 0000000000000000000000000000000000000000..0ac08c87f4e2ffafb914603b1b6b8e57767950fe --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/cnn/utils/weight_init.py @@ -0,0 +1,685 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import math +import warnings + +import numpy as np +import torch +import torch.nn as nn +from torch import Tensor + +from mmcv.utils import Registry, build_from_cfg, get_logger, print_log + +INITIALIZERS = Registry('initializer') + + +def update_init_info(module, init_info): + """Update the `_params_init_info` in the module if the value of parameters + are changed. + + Args: + module (obj:`nn.Module`): The module of PyTorch with a user-defined + attribute `_params_init_info` which records the initialization + information. + init_info (str): The string that describes the initialization. + """ + assert hasattr( + module, + '_params_init_info'), f'Can not find `_params_init_info` in {module}' + for name, param in module.named_parameters(): + + assert param in module._params_init_info, ( + f'Find a new :obj:`Parameter` ' + f'named `{name}` during executing the ' + f'`init_weights` of ' + f'`{module.__class__.__name__}`. ' + f'Please do not add or ' + f'replace parameters during executing ' + f'the `init_weights`. ') + + # The parameter has been changed during executing the + # `init_weights` of module + mean_value = param.data.mean() + if module._params_init_info[param]['tmp_mean_value'] != mean_value: + module._params_init_info[param]['init_info'] = init_info + module._params_init_info[param]['tmp_mean_value'] = mean_value + + +def constant_init(module, val, bias=0): + if hasattr(module, 'weight') and module.weight is not None: + nn.init.constant_(module.weight, val) + if hasattr(module, 'bias') and module.bias is not None: + nn.init.constant_(module.bias, bias) + + +def xavier_init(module, gain=1, bias=0, distribution='normal'): + assert distribution in ['uniform', 'normal'] + if hasattr(module, 'weight') and module.weight is not None: + if distribution == 'uniform': + nn.init.xavier_uniform_(module.weight, gain=gain) + else: + nn.init.xavier_normal_(module.weight, gain=gain) + if hasattr(module, 'bias') and module.bias is not None: + nn.init.constant_(module.bias, bias) + + +def normal_init(module, mean=0, std=1, bias=0): + if hasattr(module, 'weight') and module.weight is not None: + nn.init.normal_(module.weight, mean, std) + if hasattr(module, 'bias') and module.bias is not None: + nn.init.constant_(module.bias, bias) + + +def trunc_normal_init(module: nn.Module, + mean: float = 0, + std: float = 1, + a: float = -2, + b: float = 2, + bias: float = 0) -> None: + if hasattr(module, 'weight') and module.weight is not None: + trunc_normal_(module.weight, mean, std, a, b) # type: ignore + if hasattr(module, 'bias') and module.bias is not None: + nn.init.constant_(module.bias, bias) # type: ignore + + +def uniform_init(module, a=0, b=1, bias=0): + if hasattr(module, 'weight') and module.weight is not None: + nn.init.uniform_(module.weight, a, b) + if hasattr(module, 'bias') and module.bias is not None: + nn.init.constant_(module.bias, bias) + + +def kaiming_init(module, + a=0, + mode='fan_out', + nonlinearity='relu', + bias=0, + distribution='normal'): + assert distribution in ['uniform', 'normal'] + if hasattr(module, 'weight') and module.weight is not None: + if distribution == 'uniform': + nn.init.kaiming_uniform_( + module.weight, a=a, mode=mode, nonlinearity=nonlinearity) + else: + nn.init.kaiming_normal_( + module.weight, a=a, mode=mode, nonlinearity=nonlinearity) + if hasattr(module, 'bias') and module.bias is not None: + nn.init.constant_(module.bias, bias) + + +def caffe2_xavier_init(module, bias=0): + # `XavierFill` in Caffe2 corresponds to `kaiming_uniform_` in PyTorch + # Acknowledgment to FAIR's internal code + kaiming_init( + module, + a=1, + mode='fan_in', + nonlinearity='leaky_relu', + bias=bias, + distribution='uniform') + + +def bias_init_with_prob(prior_prob): + """initialize conv/fc bias value according to a given probability value.""" + bias_init = float(-np.log((1 - prior_prob) / prior_prob)) + return bias_init + + +def _get_bases_name(m): + return [b.__name__ for b in m.__class__.__bases__] + + +class BaseInit(object): + + def __init__(self, *, bias=0, bias_prob=None, layer=None): + self.wholemodule = False + if not isinstance(bias, (int, float)): + raise TypeError(f'bias must be a number, but got a {type(bias)}') + + if bias_prob is not None: + if not isinstance(bias_prob, float): + raise TypeError(f'bias_prob type must be float, \ + but got {type(bias_prob)}') + + if layer is not None: + if not isinstance(layer, (str, list)): + raise TypeError(f'layer must be a str or a list of str, \ + but got a {type(layer)}') + else: + layer = [] + + if bias_prob is not None: + self.bias = bias_init_with_prob(bias_prob) + else: + self.bias = bias + self.layer = [layer] if isinstance(layer, str) else layer + + def _get_init_info(self): + info = f'{self.__class__.__name__}, bias={self.bias}' + return info + + +@INITIALIZERS.register_module(name='Constant') +class ConstantInit(BaseInit): + """Initialize module parameters with constant values. + + Args: + val (int | float): the value to fill the weights in the module with + bias (int | float): the value to fill the bias. Defaults to 0. + bias_prob (float, optional): the probability for bias initialization. + Defaults to None. + layer (str | list[str], optional): the layer will be initialized. + Defaults to None. + """ + + def __init__(self, val, **kwargs): + super().__init__(**kwargs) + self.val = val + + def __call__(self, module): + + def init(m): + if self.wholemodule: + constant_init(m, self.val, self.bias) + else: + layername = m.__class__.__name__ + basesname = _get_bases_name(m) + if len(set(self.layer) & set([layername] + basesname)): + constant_init(m, self.val, self.bias) + + module.apply(init) + if hasattr(module, '_params_init_info'): + update_init_info(module, init_info=self._get_init_info()) + + def _get_init_info(self): + info = f'{self.__class__.__name__}: val={self.val}, bias={self.bias}' + return info + + +@INITIALIZERS.register_module(name='Xavier') +class XavierInit(BaseInit): + r"""Initialize module parameters with values according to the method + described in `Understanding the difficulty of training deep feedforward + neural networks - Glorot, X. & Bengio, Y. (2010). + `_ + + Args: + gain (int | float): an optional scaling factor. Defaults to 1. + bias (int | float): the value to fill the bias. Defaults to 0. + bias_prob (float, optional): the probability for bias initialization. + Defaults to None. + distribution (str): distribution either be ``'normal'`` + or ``'uniform'``. Defaults to ``'normal'``. + layer (str | list[str], optional): the layer will be initialized. + Defaults to None. + """ + + def __init__(self, gain=1, distribution='normal', **kwargs): + super().__init__(**kwargs) + self.gain = gain + self.distribution = distribution + + def __call__(self, module): + + def init(m): + if self.wholemodule: + xavier_init(m, self.gain, self.bias, self.distribution) + else: + layername = m.__class__.__name__ + basesname = _get_bases_name(m) + if len(set(self.layer) & set([layername] + basesname)): + xavier_init(m, self.gain, self.bias, self.distribution) + + module.apply(init) + if hasattr(module, '_params_init_info'): + update_init_info(module, init_info=self._get_init_info()) + + def _get_init_info(self): + info = f'{self.__class__.__name__}: gain={self.gain}, ' \ + f'distribution={self.distribution}, bias={self.bias}' + return info + + +@INITIALIZERS.register_module(name='Normal') +class NormalInit(BaseInit): + r"""Initialize module parameters with the values drawn from the normal + distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)`. + + Args: + mean (int | float):the mean of the normal distribution. Defaults to 0. + std (int | float): the standard deviation of the normal distribution. + Defaults to 1. + bias (int | float): the value to fill the bias. Defaults to 0. + bias_prob (float, optional): the probability for bias initialization. + Defaults to None. + layer (str | list[str], optional): the layer will be initialized. + Defaults to None. + + """ + + def __init__(self, mean=0, std=1, **kwargs): + super().__init__(**kwargs) + self.mean = mean + self.std = std + + def __call__(self, module): + + def init(m): + if self.wholemodule: + normal_init(m, self.mean, self.std, self.bias) + else: + layername = m.__class__.__name__ + basesname = _get_bases_name(m) + if len(set(self.layer) & set([layername] + basesname)): + normal_init(m, self.mean, self.std, self.bias) + + module.apply(init) + if hasattr(module, '_params_init_info'): + update_init_info(module, init_info=self._get_init_info()) + + def _get_init_info(self): + info = f'{self.__class__.__name__}: mean={self.mean},' \ + f' std={self.std}, bias={self.bias}' + return info + + +@INITIALIZERS.register_module(name='TruncNormal') +class TruncNormalInit(BaseInit): + r"""Initialize module parameters with the values drawn from the normal + distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)` with values + outside :math:`[a, b]`. + + Args: + mean (float): the mean of the normal distribution. Defaults to 0. + std (float): the standard deviation of the normal distribution. + Defaults to 1. + a (float): The minimum cutoff value. + b ( float): The maximum cutoff value. + bias (float): the value to fill the bias. Defaults to 0. + bias_prob (float, optional): the probability for bias initialization. + Defaults to None. + layer (str | list[str], optional): the layer will be initialized. + Defaults to None. + + """ + + def __init__(self, + mean: float = 0, + std: float = 1, + a: float = -2, + b: float = 2, + **kwargs) -> None: + super().__init__(**kwargs) + self.mean = mean + self.std = std + self.a = a + self.b = b + + def __call__(self, module: nn.Module) -> None: + + def init(m): + if self.wholemodule: + trunc_normal_init(m, self.mean, self.std, self.a, self.b, + self.bias) + else: + layername = m.__class__.__name__ + basesname = _get_bases_name(m) + if len(set(self.layer) & set([layername] + basesname)): + trunc_normal_init(m, self.mean, self.std, self.a, self.b, + self.bias) + + module.apply(init) + if hasattr(module, '_params_init_info'): + update_init_info(module, init_info=self._get_init_info()) + + def _get_init_info(self): + info = f'{self.__class__.__name__}: a={self.a}, b={self.b},' \ + f' mean={self.mean}, std={self.std}, bias={self.bias}' + return info + + +@INITIALIZERS.register_module(name='Uniform') +class UniformInit(BaseInit): + r"""Initialize module parameters with values drawn from the uniform + distribution :math:`\mathcal{U}(a, b)`. + + Args: + a (int | float): the lower bound of the uniform distribution. + Defaults to 0. + b (int | float): the upper bound of the uniform distribution. + Defaults to 1. + bias (int | float): the value to fill the bias. Defaults to 0. + bias_prob (float, optional): the probability for bias initialization. + Defaults to None. + layer (str | list[str], optional): the layer will be initialized. + Defaults to None. + """ + + def __init__(self, a=0, b=1, **kwargs): + super().__init__(**kwargs) + self.a = a + self.b = b + + def __call__(self, module): + + def init(m): + if self.wholemodule: + uniform_init(m, self.a, self.b, self.bias) + else: + layername = m.__class__.__name__ + basesname = _get_bases_name(m) + if len(set(self.layer) & set([layername] + basesname)): + uniform_init(m, self.a, self.b, self.bias) + + module.apply(init) + if hasattr(module, '_params_init_info'): + update_init_info(module, init_info=self._get_init_info()) + + def _get_init_info(self): + info = f'{self.__class__.__name__}: a={self.a},' \ + f' b={self.b}, bias={self.bias}' + return info + + +@INITIALIZERS.register_module(name='Kaiming') +class KaimingInit(BaseInit): + r"""Initialize module parameters with the values according to the method + described in `Delving deep into rectifiers: Surpassing human-level + performance on ImageNet classification - He, K. et al. (2015). + `_ + + Args: + a (int | float): the negative slope of the rectifier used after this + layer (only used with ``'leaky_relu'``). Defaults to 0. + mode (str): either ``'fan_in'`` or ``'fan_out'``. Choosing + ``'fan_in'`` preserves the magnitude of the variance of the weights + in the forward pass. Choosing ``'fan_out'`` preserves the + magnitudes in the backwards pass. Defaults to ``'fan_out'``. + nonlinearity (str): the non-linear function (`nn.functional` name), + recommended to use only with ``'relu'`` or ``'leaky_relu'`` . + Defaults to 'relu'. + bias (int | float): the value to fill the bias. Defaults to 0. + bias_prob (float, optional): the probability for bias initialization. + Defaults to None. + distribution (str): distribution either be ``'normal'`` or + ``'uniform'``. Defaults to ``'normal'``. + layer (str | list[str], optional): the layer will be initialized. + Defaults to None. + """ + + def __init__(self, + a=0, + mode='fan_out', + nonlinearity='relu', + distribution='normal', + **kwargs): + super().__init__(**kwargs) + self.a = a + self.mode = mode + self.nonlinearity = nonlinearity + self.distribution = distribution + + def __call__(self, module): + + def init(m): + if self.wholemodule: + kaiming_init(m, self.a, self.mode, self.nonlinearity, + self.bias, self.distribution) + else: + layername = m.__class__.__name__ + basesname = _get_bases_name(m) + if len(set(self.layer) & set([layername] + basesname)): + kaiming_init(m, self.a, self.mode, self.nonlinearity, + self.bias, self.distribution) + + module.apply(init) + if hasattr(module, '_params_init_info'): + update_init_info(module, init_info=self._get_init_info()) + + def _get_init_info(self): + info = f'{self.__class__.__name__}: a={self.a}, mode={self.mode}, ' \ + f'nonlinearity={self.nonlinearity}, ' \ + f'distribution ={self.distribution}, bias={self.bias}' + return info + + +@INITIALIZERS.register_module(name='Caffe2Xavier') +class Caffe2XavierInit(KaimingInit): + # `XavierFill` in Caffe2 corresponds to `kaiming_uniform_` in PyTorch + # Acknowledgment to FAIR's internal code + def __init__(self, **kwargs): + super().__init__( + a=1, + mode='fan_in', + nonlinearity='leaky_relu', + distribution='uniform', + **kwargs) + + def __call__(self, module): + super().__call__(module) + + +@INITIALIZERS.register_module(name='Pretrained') +class PretrainedInit(object): + """Initialize module by loading a pretrained model. + + Args: + checkpoint (str): the checkpoint file of the pretrained model should + be load. + prefix (str, optional): the prefix of a sub-module in the pretrained + model. it is for loading a part of the pretrained model to + initialize. For example, if we would like to only load the + backbone of a detector model, we can set ``prefix='backbone.'``. + Defaults to None. + map_location (str): map tensors into proper locations. + """ + + def __init__(self, checkpoint, prefix=None, map_location=None): + self.checkpoint = checkpoint + self.prefix = prefix + self.map_location = map_location + + def __call__(self, module): + from mmcv.runner import (_load_checkpoint_with_prefix, load_checkpoint, + load_state_dict) + logger = get_logger('mmcv') + if self.prefix is None: + print_log(f'load model from: {self.checkpoint}', logger=logger) + load_checkpoint( + module, + self.checkpoint, + map_location=self.map_location, + strict=False, + logger=logger) + else: + print_log( + f'load {self.prefix} in model from: {self.checkpoint}', + logger=logger) + state_dict = _load_checkpoint_with_prefix( + self.prefix, self.checkpoint, map_location=self.map_location) + load_state_dict(module, state_dict, strict=False, logger=logger) + + if hasattr(module, '_params_init_info'): + update_init_info(module, init_info=self._get_init_info()) + + def _get_init_info(self): + info = f'{self.__class__.__name__}: load from {self.checkpoint}' + return info + + +def _initialize(module, cfg, wholemodule=False): + func = build_from_cfg(cfg, INITIALIZERS) + # wholemodule flag is for override mode, there is no layer key in override + # and initializer will give init values for the whole module with the name + # in override. + func.wholemodule = wholemodule + func(module) + + +def _initialize_override(module, override, cfg): + if not isinstance(override, (dict, list)): + raise TypeError(f'override must be a dict or a list of dict, \ + but got {type(override)}') + + override = [override] if isinstance(override, dict) else override + + for override_ in override: + + cp_override = copy.deepcopy(override_) + name = cp_override.pop('name', None) + if name is None: + raise ValueError('`override` must contain the key "name",' + f'but got {cp_override}') + # if override only has name key, it means use args in init_cfg + if not cp_override: + cp_override.update(cfg) + # if override has name key and other args except type key, it will + # raise error + elif 'type' not in cp_override.keys(): + raise ValueError( + f'`override` need "type" key, but got {cp_override}') + + if hasattr(module, name): + _initialize(getattr(module, name), cp_override, wholemodule=True) + else: + raise RuntimeError(f'module did not have attribute {name}, ' + f'but init_cfg is {cp_override}.') + + +def initialize(module, init_cfg): + r"""Initialize a module. + + Args: + module (``torch.nn.Module``): the module will be initialized. + init_cfg (dict | list[dict]): initialization configuration dict to + define initializer. OpenMMLab has implemented 6 initializers + including ``Constant``, ``Xavier``, ``Normal``, ``Uniform``, + ``Kaiming``, and ``Pretrained``. + + Example: + >>> module = nn.Linear(2, 3, bias=True) + >>> init_cfg = dict(type='Constant', layer='Linear', val =1 , bias =2) + >>> initialize(module, init_cfg) + + >>> module = nn.Sequential(nn.Conv1d(3, 1, 3), nn.Linear(1,2)) + >>> # define key ``'layer'`` for initializing layer with different + >>> # configuration + >>> init_cfg = [dict(type='Constant', layer='Conv1d', val=1), + dict(type='Constant', layer='Linear', val=2)] + >>> initialize(module, init_cfg) + + >>> # define key``'override'`` to initialize some specific part in + >>> # module + >>> class FooNet(nn.Module): + >>> def __init__(self): + >>> super().__init__() + >>> self.feat = nn.Conv2d(3, 16, 3) + >>> self.reg = nn.Conv2d(16, 10, 3) + >>> self.cls = nn.Conv2d(16, 5, 3) + >>> model = FooNet() + >>> init_cfg = dict(type='Constant', val=1, bias=2, layer='Conv2d', + >>> override=dict(type='Constant', name='reg', val=3, bias=4)) + >>> initialize(model, init_cfg) + + >>> model = ResNet(depth=50) + >>> # Initialize weights with the pretrained model. + >>> init_cfg = dict(type='Pretrained', + checkpoint='torchvision://resnet50') + >>> initialize(model, init_cfg) + + >>> # Initialize weights of a sub-module with the specific part of + >>> # a pretrained model by using "prefix". + >>> url = 'http://download.openmmlab.com/mmdetection/v2.0/retinanet/'\ + >>> 'retinanet_r50_fpn_1x_coco/'\ + >>> 'retinanet_r50_fpn_1x_coco_20200130-c2398f9e.pth' + >>> init_cfg = dict(type='Pretrained', + checkpoint=url, prefix='backbone.') + """ + if not isinstance(init_cfg, (dict, list)): + raise TypeError(f'init_cfg must be a dict or a list of dict, \ + but got {type(init_cfg)}') + + if isinstance(init_cfg, dict): + init_cfg = [init_cfg] + + for cfg in init_cfg: + # should deeply copy the original config because cfg may be used by + # other modules, e.g., one init_cfg shared by multiple bottleneck + # blocks, the expected cfg will be changed after pop and will change + # the initialization behavior of other modules + cp_cfg = copy.deepcopy(cfg) + override = cp_cfg.pop('override', None) + _initialize(module, cp_cfg) + + if override is not None: + cp_cfg.pop('layer', None) + _initialize_override(module, override, cp_cfg) + else: + # All attributes in module have same initialization. + pass + + +def _no_grad_trunc_normal_(tensor: Tensor, mean: float, std: float, a: float, + b: float) -> Tensor: + # Method based on + # https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf + # Modified from + # https://github.com/pytorch/pytorch/blob/master/torch/nn/init.py + def norm_cdf(x): + # Computes standard normal cumulative distribution function + return (1. + math.erf(x / math.sqrt(2.))) / 2. + + if (mean < a - 2 * std) or (mean > b + 2 * std): + warnings.warn( + 'mean is more than 2 std from [a, b] in nn.init.trunc_normal_. ' + 'The distribution of values may be incorrect.', + stacklevel=2) + + with torch.no_grad(): + # Values are generated by using a truncated uniform distribution and + # then using the inverse CDF for the normal distribution. + # Get upper and lower cdf values + lower = norm_cdf((a - mean) / std) + upper = norm_cdf((b - mean) / std) + + # Uniformly fill tensor with values from [lower, upper], then translate + # to [2lower-1, 2upper-1]. + tensor.uniform_(2 * lower - 1, 2 * upper - 1) + + # Use inverse cdf transform for normal distribution to get truncated + # standard normal + tensor.erfinv_() + + # Transform to proper mean, std + tensor.mul_(std * math.sqrt(2.)) + tensor.add_(mean) + + # Clamp to ensure it's in the proper range + tensor.clamp_(min=a, max=b) + return tensor + + +def trunc_normal_(tensor: Tensor, + mean: float = 0., + std: float = 1., + a: float = -2., + b: float = 2.) -> Tensor: + r"""Fills the input Tensor with values drawn from a truncated + normal distribution. The values are effectively drawn from the + normal distribution :math:`\mathcal{N}(\text{mean}, \text{std}^2)` + with values outside :math:`[a, b]` redrawn until they are within + the bounds. The method used for generating the random values works + best when :math:`a \leq \text{mean} \leq b`. + + Modified from + https://github.com/pytorch/pytorch/blob/master/torch/nn/init.py + + Args: + tensor (``torch.Tensor``): an n-dimensional `torch.Tensor`. + mean (float): the mean of the normal distribution. + std (float): the standard deviation of the normal distribution. + a (float): the minimum cutoff value. + b (float): the maximum cutoff value. + """ + return _no_grad_trunc_normal_(tensor, mean, std, a, b) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..6ac55e63b999ccabecedae617ab4bb33a6e10e3a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/__init__.py @@ -0,0 +1,4 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from . import ipu, mlu + +__all__ = ['mlu', 'ipu'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/__init__.py new file mode 100755 index 0000000000000000000000000000000000000000..d550865ad20790f0eb79015abc866548c0f2f83b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/__init__.py @@ -0,0 +1,14 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.utils import IS_IPU_AVAILABLE + +if IS_IPU_AVAILABLE: + from .dataloader import IPUDataLoader + from .hook_wrapper import IPUFp16OptimizerHook + from .model_wrapper import ipu_model_wrapper + from .runner import IPUBaseRunner, IPUEpochBasedRunner, IPUIterBasedRunner + from .utils import cfg2options + __all__ = [ + 'cfg2options', 'ipu_model_wrapper', 'IPUFp16OptimizerHook', + 'IPUDataLoader', 'IPUBaseRunner', 'IPUEpochBasedRunner', + 'IPUIterBasedRunner' + ] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/dataloader.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/dataloader.py new file mode 100755 index 0000000000000000000000000000000000000000..1485df2f31facff79238c70d89fdd9030fddcbce --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/dataloader.py @@ -0,0 +1,157 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from collections.abc import Mapping, Sequence +from functools import partial + +import poptorch +from torch.utils.data.dataloader import default_collate + +from mmcv.parallel import DataContainer + + +def collate(batch, samples_per_gpu=1): + """Put each data field into a tensor/DataContainer with outer dimension + batch size. + + TODO support for + :type:`~mmcv.parallel.DataContainer`. Currently, it will be ignored. + There are 3 cases. + + 1. cpu_only = True, e.g., meta data. + 2. cpu_only = False, stack = True, e.g., images tensors. + 3. cpu_only = False, stack = False, e.g., gt bboxes. + """ + + if not isinstance(batch, Sequence): + raise TypeError( + f'`batch` should be a sequence, but got {type(batch)}.') + + if isinstance(batch[0], DataContainer): + # TODO `DataContainer` will be supported in the future. + raise TypeError('DataContainer is not supported in ipu data loader.') + elif isinstance(batch[0], Sequence): + transposed = zip(*batch) + collated_batch = [] + for samples in transposed: + if not isinstance(samples[0], DataContainer): + # At present, we will skip the processing of datacontainer, + # which will reduce the performance of IPU DataLoder + collated_batch.append(collate(samples, samples_per_gpu)) + return collated_batch + elif isinstance(batch[0], Mapping): + collated_batch = {} + for key in batch[0]: + if not isinstance(batch[0][key], DataContainer): + # At present, we will skip the processing of datacontainer, + # which will reduce the performance of IPU DataLoder + collated_batch[key] = collate([d[key] for d in batch]) + return collated_batch + else: + return default_collate(batch) + + +class IPUDataLoader(poptorch.DataLoader): + """Thin wrapper of `torch.utils.data.DataLoader`. + + Compared with the pytorch DataLoder, this DataLoder changes the way of + calculation of batch size and adds the AsynchronousDataAccessor to + load and release data faster in cpu mode. + + If this data loader is used in a distributed execution environment, it will + ensure that each process uses a different subset of the dataset, providing + you first call ``options.randomSeed(N)`` with an integer N which is the + same across all hosts. + + Args: + dataset (torch.utils.data.Dataset): The dataset to get the data from. + options (poptorch.Options): Options that will be used to compile + and run the model. + batch_size (int, optional): This is the batch size in the conventional + sense of being the size that runs through an operation in the model + at any given time. + shuffle (bool, optional): set to ``True`` to have the data reshuffled + at every epoch (default: ``False``). + num_workers (int, optional): how many subprocesses to use for data + loading. ``0`` means that the data will be loaded in the main + process. (default: ``0``) + drop_last (bool, optional): If True and the number of elements in the + dataset is not a multiple of the combined batch size then the + incomplete batch at the end will be dropped. + persistent_workers (bool, optional): Re-use workers between + iterations if True. + auto_distributed_partitioning (bool, optional): If True, partitions the + dataset for distributed execution automatically. Otherwise, it is + assumed that partitioning has been handled manually. + mode (poptorch.DataLoaderMode, optional): If `DataLoaderMode.Async`, + uses an :py:class:`~poptorch.AsynchronousDataAccessor` to access + the dataset. If `DataLoaderMode.Sync`, accesses the dataset + synchronously. + async_options (Dict[str, Any], optional): Options to pass to + :py:class:`~poptorch.AsynchronousDataAccessor`. + rebatched_worker_size (int, optional): When using AsyncRebatched: batch + size of the tensors loaded by the workers. + Default to the combined batch size. + If specified the ``rebatched_worker_size`` must be less than + or equal to the combined batch size. + kwargs (Dict[str, Any], optional): Other options to pass to PyTorch's + ``DataLoader`` constructor. + """ + + def __init__(self, + dataset, + options, + batch_size=1, + shuffle=False, + num_workers=0, + drop_last=True, + persistent_workers=True, + auto_distributed_partitioning=True, + mode='sync', + async_options=None, + rebatched_worker_size=None, + **kwargs): + """Lazy init: + + In many frameworks, the dataloader will be constructed before the + initialization of the ipu options, so the lazy init method is used + here, and the real initialization will not be done until the dataloader + needs to be used and the options are input. + """ + # lazy init: sometimes, we cannot get IPU options when build data + # loader + self.kwargs = { + 'dataset': dataset, + 'batch_size': batch_size, + 'shuffle': shuffle, + 'num_workers': num_workers, + 'drop_last': drop_last, + 'persistent_workers': persistent_workers, + 'auto_distributed_partitioning': auto_distributed_partitioning, + 'mode': mode, + 'collate_fn': partial(collate, samples_per_gpu=batch_size), + 'async_options': async_options, + 'rebatched_worker_size': rebatched_worker_size, + **kwargs + } + self.dataset = dataset + self.initialized = False + if options: + self.init(options=options) + + def init(self, options, **kwargs): + if not self.initialized: + kwargs = {**self.kwargs, **kwargs, 'options': options} + if kwargs['mode'] == 'sync': + kwargs['mode'] = poptorch.DataLoaderMode.Sync + elif kwargs['mode'] == 'async': + kwargs['mode'] = poptorch.DataLoaderMode.AsyncRebatched + if kwargs['async_options'] is None: + kwargs['async_options'] = { + 'load_indefinitely': True, + 'buffer_size': 8 + } + if kwargs['rebatched_worker_size'] is None: + kwargs['rebatched_worker_size'] = 128 + super().__init__(**kwargs) + self.initialized = True + + return self diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/hierarchical_data_manager.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/hierarchical_data_manager.py new file mode 100755 index 0000000000000000000000000000000000000000..a6f3b3cd2a139bcbc7852e7849071ab4b9fbb76f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/hierarchical_data_manager.py @@ -0,0 +1,243 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import numpy as np +import torch + +from mmcv.parallel import DataContainer + +# A customized None type for HierarchicalDataManager +HierarchicalDataNone = object() + + +class HierarchicalDataManager: + """A class manage all the tensors in the hierarchical data. + + At present, the input data structure accepted by IPU is limited, + when the input data structure of mmcv varies. + Here, an intermediate class is needed to get and update tensors + from the original data. + + HierarchicalDataManager will record a hierarchical input/output data in + self._hierarchical_data. For example, we have an input data: + {'img': tensorA, 'label': tensorB, 'img_metas': [tensorC, tensorD]} + To enable IPU to use the input, HierarchicalDataManager will collect + the torch tensors from self._hierarchical_data into a tuple like: + (tensorA, tensorB, tensorC, tensorD). + Meanwhile, the return of IPU is a tuple of tensors, HierarchicalDataManager + also have a function named update_all_tensors to update tensors in + self._hierarchical_data which is the output for upper calls. + + Args: + logger (:obj:`logging.Logger`): Logger used during running. + Defaults to None. + """ + + def __init__(self, logger=None): + self.atomic_types = (int, str, float, np.ndarray, type(None)) + self.warning = warnings.warn if logger is None else logger.warning + # enable or disable input data's shape and value check + self.quick_mode = False + self._hierarchical_data = None + + def quick(self): + self.quick_mode = True + + def compare_atomic_type(self, a, b): + """Compare data, supported datatypes are numpy array and python basic + types.""" + if isinstance(a, np.ndarray): + return np.all(a == b) + else: + return a == b + + def record_hierarchical_data(self, data): + """Record a hierarchical data.""" + if self._hierarchical_data is not None: + if isinstance(data, torch.Tensor): + assert isinstance(self._hierarchical_data, torch.Tensor), \ + 'original hierarchical data is not torch.tensor' + self._hierarchical_data = data + else: + self.update_hierarchical_data(data) + else: + self._hierarchical_data = data + + @property + def hierarchical_data(self): + return self._hierarchical_data + + def update_hierarchical_data(self, + dataA, + dataB=HierarchicalDataNone, + strict=True, + address='data'): + """Update dataB with dataA in-place. + + Args: + dataA (list or dict or tuple): New hierarchical data. + dataB (list or dict or tuple): hierarchical data to update. + if not specified, self.hierarchical_data will be updated then. + strict (bool, optional): If true, an error will be reported + when the following conditions occur: + 1. Non-torch.Tensor data changed. + 2. Torch.Tensor data shape changed. + address (str): Record the address of current data to be updated. + Default: 'data'. + """ + if dataB is HierarchicalDataNone: + dataB = self.hierarchical_data + + # Update with a da ta with the same structure + # but different values(tensors and basic python data types) + if isinstance(dataA, (tuple, list)): + for idx, node in enumerate(dataA): + new_address = '' + if not self.quick_mode: + new_address = address + f'[{str(idx)}]' + assert isinstance(node, type(dataB[idx])),\ + f'data structure changed: {new_address}' + if isinstance(node, torch.Tensor): + dataB[idx] = node + else: + self.update_hierarchical_data( + node, dataB[idx], strict, address=new_address) + elif isinstance(dataA, dict): + for k, v in dataA.items(): + new_address = '' + if not self.quick_mode: + new_address = address + f'[{str(k)}]' + assert isinstance(v, type(dataB[k])),\ + f'data structure changed: {new_address}' + if isinstance(v, torch.Tensor): + dataB[k] = v + else: + self.update_hierarchical_data( + v, dataB[k], strict, address=new_address) + elif isinstance(dataA, self.atomic_types): + if not self.quick_mode: + is_equal = self.compare_atomic_type(dataA, dataB) + if not is_equal: + if strict: + raise ValueError( + 'all data except torch.Tensor should be same, ' + f'but data({address}) is changed.') + else: + self.warning( + f'find a non-torch.Tensor data({type(dataA)}) ' + f'changed, and the address is {address}') + elif isinstance(dataA, DataContainer): + if not self.quick_mode: + assert isinstance(dataB, DataContainer) + new_address = address + '.data' + self.update_hierarchical_data( + dataA.data, dataB.data, False, address=new_address) + else: + raise NotImplementedError( + f'not supported datatype:{type(dataA)}, address is {address}') + + def collect_all_tensors(self, hierarchical_data=None): + """Collect torch.Tensor data from self.hierarchical_data to a list and + return.""" + # get a list of tensor from self._hierarchical_data + if hierarchical_data is None: + hierarchical_data = self._hierarchical_data + tensors = [] + if isinstance(hierarchical_data, torch.Tensor): + tensors = [hierarchical_data] + else: + self._collect_tensors(hierarchical_data, tensors) + return tensors + + def _collect_tensors(self, data, tensors): + if isinstance(data, (tuple, list)): + for node in data: + if isinstance(node, torch.Tensor): + tensors.append(node) + else: + self._collect_tensors(node, tensors) + elif isinstance(data, dict): + for v in data.values(): + if isinstance(v, torch.Tensor): + tensors.append(v) + else: + self._collect_tensors(v, tensors) + elif isinstance(data, self.atomic_types): + pass + elif isinstance(data, DataContainer): + self._collect_tensors(data.data, tensors) + else: + raise NotImplementedError(f'not supported datatype:{type(data)}') + + def update_all_tensors(self, tensors): + """Put tensors from tuple back to self.hierarchical_data.""" + if isinstance(self._hierarchical_data, torch.Tensor): + print(tensors, len(tensors)) + assert len(tensors) == 1 + assert isinstance(tensors[0], torch.Tensor) + self._hierarchical_data = tensors[0] + else: + # convert to list if tensors is tuple + tensors = list(tensors) + self._set_tensors(self._hierarchical_data, tensors) + return self.hierarchical_data + + def _set_tensors(self, data, tensors): + if isinstance(data, tuple): + data = list(data) + for idx in range(len(data)): + if isinstance(data[idx], torch.Tensor): + data[idx] = tensors.pop(0) + else: + self._set_tensors(data[idx], tensors) + data = tuple(data) + elif isinstance(data, list): + for idx in range(len(data)): + if isinstance(data[idx], torch.Tensor): + data[idx] = tensors.pop(0) + else: + self._set_tensors(data[idx], tensors) + elif isinstance(data, dict): + for k, v in data.items(): + if isinstance(v, torch.Tensor): + data[k] = tensors.pop(0) + else: + self._set_tensors(v, tensors) + elif isinstance(data, self.atomic_types): + pass + elif isinstance(data, DataContainer): + self._set_tensors(data.data, tensors) + else: + raise NotImplementedError(f'not supported datatype:{type(data)}') + + def clean_all_tensors(self): + """Delete tensors from self.hierarchical_data.""" + self._clean_tensors(self._hierarchical_data) + + def _clean_tensors(self, data): + if isinstance(data, tuple): + data = list(data) + for idx in range(len(data)): + if isinstance(data[idx], torch.Tensor): + data[idx] = None + else: + self._clean_tensors(data[idx]) + data = tuple(data) + elif isinstance(data, list): + for idx in range(len(data)): + if isinstance(data[idx], torch.Tensor): + data[idx] = None + else: + self._clean_tensors(data[idx]) + elif isinstance(data, dict): + for k, v in data.items(): + if isinstance(v, torch.Tensor): + data[k] = None + else: + self._clean_tensors(v) + elif isinstance(data, self.atomic_types): + pass + elif isinstance(data, DataContainer): + self._clean_tensors(data.data) + else: + raise NotImplementedError(f'not supported datatype:{type(data)}') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/hook_wrapper.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/hook_wrapper.py new file mode 100755 index 0000000000000000000000000000000000000000..141afb86d05a42c06fb5c4355cb47cae18e9bb2f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/hook_wrapper.py @@ -0,0 +1,105 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.runner import HOOKS, LrUpdaterHook, OptimizerHook +from mmcv.utils import TORCH_VERSION, digit_version + + +def wrap_lr_updater_hook(lr_hook_class): + """A wrapper function to wrap any subclass of LrUpdaterHook. + + IPU needs extra operations to upload optimizer settings. This wrapper will + override function(_set_lr) of a subclass of LrUpdaterHook. + """ + assert issubclass(lr_hook_class, LrUpdaterHook) + + class ipu_lr_hook_class(lr_hook_class): + + def _set_lr(self, runner, *args, **kwargs): + super()._set_lr(runner, *args, **kwargs) + # convert torch optimizer to poptorch optimizer + runner.model.setOptimizer(runner.optimizer) + + return ipu_lr_hook_class + + +def wrap_optimizer_hook(optimizer_hook_class): + """A wrapper function to wrap OptimizerHook. + + This is an non-intrusive implementation of wrapping optimizer hook (or you + need to change every config file to use IPU optimizer hook) IPU's clip-norm + implementation is different from pytorch, so there should be an error + raised when using clip-norm. + """ + + class ipu_optimizer_hook_class(OptimizerHook): + + def __init__(self, **kwargs): + super().__init__(**kwargs) + if self.grad_clip is not None: + raise NotImplementedError('IPU does not support gradient clip') + + return ipu_optimizer_hook_class + + +if (TORCH_VERSION != 'parrots' + and digit_version(TORCH_VERSION) >= digit_version('1.6.0')): + + @HOOKS.register_module() + class IPUFp16OptimizerHook(OptimizerHook): + """FP16 optimizer hook (using PyTorch's implementation). + + If you are using PyTorch >= 1.6, torch.cuda.amp is used as the backend, + to take care of the optimization procedure. + + Args: + loss_scale (float | str | dict): Scale factor configuration. + If loss_scale is a float, static loss scaling will be used with + the specified scale. If loss_scale is a string, it must be + 'dynamic', then dynamic loss scaling will be used. + It can also be a dict containing arguments of GradScalar. + Defaults to 512. For Pytorch >= 1.6, mmcv uses official + implementation of GradScaler. If you use a dict version of + loss_scale to create GradScaler, please refer to: + https://pytorch.org/docs/stable/amp.html#torch.cuda.amp.GradScaler + for the parameters. + + Examples: + >>> loss_scale = dict( + ... init_scale=65536.0, + ... growth_factor=2.0, + ... backoff_factor=0.5, + ... growth_interval=2000 + ... ) + >>> optimizer_hook = Fp16OptimizerHook(loss_scale=loss_scale) + """ + + def __init__(self, + grad_clip=None, + coalesce=True, + bucket_size_mb=-1, + loss_scale=512., + distributed=True): + assert grad_clip is None,\ + 'IPU mode does not support `grad_clip` currently' + assert coalesce,\ + 'implemented all reduce in distributed training currently' + assert bucket_size_mb == -1,\ + '`bucket_size_mb` should not be set in IPU mode' + self.distributed = distributed + self._scale_update_param = None + if loss_scale == 'dynamic': + raise NotImplementedError( + 'IPU mode does not support dynamic loss scale currently') + elif isinstance(loss_scale, float): + self.loss_scale = loss_scale + elif isinstance(loss_scale, dict): + raise NotImplementedError( + 'IPU mode supports single scale currently') + else: + raise ValueError( + f'loss_scale should be float, but got {loss_scale} ') + + def after_train_iter(self, runner): + pass + +else: + raise RuntimeError('The IPU mode only supports torch 1.6 and above') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/model_wrapper.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/model_wrapper.py new file mode 100755 index 0000000000000000000000000000000000000000..c345537e29b27cf7fff740269da8643c9570cd36 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/model_wrapper.py @@ -0,0 +1,721 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import inspect +from collections import OrderedDict +from typing import Optional, Union + +import poptorch +import torch +import torch.nn as nn +from poptorch import PoplarExecutor, __version__, identity_loss +from poptorch._args_parser import ArgsParser + +from mmcv.runner import auto_fp16 +from .hierarchical_data_manager import HierarchicalDataManager +from .utils import compare_ndarray, model_sharding, recomputation_checkpoint + + +class DictArgsParser(ArgsParser): + """A helper class for handling model input. + + Args: + inputs (list): Inputs of model. + """ + + def __init__(self, inputs): + # Combine args and kwargs: + self._has_variadic_arguments = True + self._varnames = list(inputs.keys()) + self._defaults = [inspect.Parameter.empty for _ in self._varnames] + self._warned_not_contiguous_input = False + + +class WrappedNet(nn.Module): + """A net wrapper for model conversion. + + This wrapper will make some changes and add some extra functions to + training/inference model. + + Args: + model (:obj:`nn.Module`): The model to run. + inputs_manager (:obj:`HierarchicalDataManager`): A parser + converting inputs from tuple to dictionary. + outputs_manager (:obj:`HierarchicalDataManager`): A parser + converting outputs from dictionary to tuple. + inter_outputs_in_cpu (dict): Specify the features to be + recorded. + modules_to_record (mmcv.Config, list): Index or name of modules which + will be recorded for output. It is necessary to specify output for + static graph of model training or inference. + """ + + def __init__(self, + model, + inputs_manager, + outputs_manager, + inter_outputs_in_cpu, + modules_to_record=None): + super().__init__() + self.model = model + self.inputs_manager = inputs_manager + self.outputs_manager = outputs_manager + self.training = model.training + # Register a hook function to capture the intermediate features + # generated by the network to align the outputs between ipu and cpu + # Used to confirm whether the implementation of CPU is consistent + # with the implementation of IPU + self.inter_outputs_in_cpu = inter_outputs_in_cpu + if modules_to_record is None: + modules_to_record = [] + + for idx, (name, module) in enumerate(model.named_modules()): + if name in modules_to_record or idx in modules_to_record: + features_hook = self.get_input_output_hook( + name, idx, self.inter_outputs_in_cpu) + module.register_forward_hook(hook=features_hook) + + def get_input_output_hook(self, name, idx, save_dict): + + def input_output_hook(module, fea_in, fea_out): + if isinstance(fea_in, tuple): + fea_in = list(fea_in) + if isinstance(fea_out, tuple): + fea_out = list(fea_out) + save_dict[name] = { + 'fea_in': fea_in, + 'fea_out': fea_out, + 'idx': idx + } + return None + + return input_output_hook + + def forward(self, inputs_tuple): + """This function is used to be compiled to ipu, the inputs and outputs + need to be tuples, so here we need to restore the input back to a + dictionary and convert the output to a tuple.""" + self.inputs_manager.update_all_tensors(inputs_tuple) + kwargs = {**(self.inputs_manager.hierarchical_data)} + if self.training: + outputs = self.forward_train(kwargs) + # tell poptorch which loss will be used finally + identity_loss(outputs['loss'], reduction='none') + else: + outputs = self.forward_eval(kwargs) + + if isinstance(outputs, torch.Tensor): + # currently not support single tensor output, + # need to wrap it with a dictionary, + # use a keyword to identify this case + outputs = {'output of WrappedNet: single tensor': outputs} + + # if there are some features need to be record, add extra outputs + for name in self.inter_outputs_in_cpu: + outputs[name] = self.inter_outputs_in_cpu[name] + + # record all the places of return tensors in the converting stage + # while in the real run stage, all the tensor are changed in-place + # that means the output can be obtained directly outside this function + self.outputs_manager.record_hierarchical_data(outputs) + plain_outputs = self.outputs_manager.collect_all_tensors() + return plain_outputs + + def forward_train(self, kwargs): + optimizer = kwargs.pop('optimizer') + outputs = self.train_step(kwargs, optimizer) + return outputs + + def train_step(self, data, optimizer=None, **kwargs): + """The iteration step during training. + + This method defines an iteration step during training, except for the + back propagation and optimizer updating, which are done in an optimizer + hook. Note that in some complicated cases or models, the whole process + including back propagation and optimizer updating are also defined in + this method, such as GAN. + + Args: + data (dict): The output of dataloader. + optimizer (:obj:`torch.optim.Optimizer`, optional): The + optimizer of runner is passed to ``train_step()``. This + argument is unused and reserved. + + Returns: + dict: Dict of outputs. The following fields are contained. + - loss (torch.Tensor): A tensor for back propagation, which \ + can be a weighted sum of multiple losses. + - log_vars (dict): Dict contains all the variables to be sent \ + to the logger. + - num_samples (int): Indicates the batch size (when the model \ + is DDP, it means the batch size on each GPU), which is \ + used for averaging the logs. + """ + losses = self.model(**data) + loss, log_vars = self._parse_losses(losses) + + outputs = dict( + loss=loss, log_vars=log_vars, num_samples=len(data['img'].data)) + + return outputs + + def _parse_losses(self, losses): + log_vars = OrderedDict() + for loss_name, loss_value in losses.items(): + if isinstance(loss_value, torch.Tensor): + log_vars[loss_name] = loss_value.mean() + elif isinstance(loss_value, list): + log_vars[loss_name] = sum(loss.mean() for loss in loss_value) + elif isinstance(loss_value, dict): + for name, value in loss_value.items(): + log_vars[name] = value + else: + raise TypeError( + f'{loss_name} is not a tensor or list of tensors') + + loss = sum(value for key, value in log_vars.items() if 'loss' in key) + log_vars['loss'] = loss + + return loss, log_vars + + def forward_eval(self, kwargs): + img = kwargs.pop('img') + img_metas = kwargs.pop('img_metas', None) + return_loss = kwargs.pop('return_loss') + assert not return_loss + # TODO Temporarily hard-code to close post_process, + # otherwise, in the third trace(_check_trace), + # post_process will convert output tensor to numpy array automatically, + # resulting in _check_trace failure + outputs = self.model( + img, + img_metas=img_metas, + return_loss=return_loss, + post_process=False) + return outputs + + +class MMPoplarExecutor(PoplarExecutor): + """An executor for inputs/outputs parsing, model compilation, data + alignment and IPU upload/download. + + Args: + model (:obj:`nn.Module`): The model to be compiled. + logger (:obj:`logging.Logger`): Logger used during running. + Defaults to None. + training (bool): Model in training mode or eval mode. + modules_to_record (mmcv.Config, list): Index or name of modules which + will be recorded for output. It is necessary to specify output for + static graph of model training or inference. + args (argument list): Arguments passed to the `__init__` + method of PoplarExecutor. + kwargs (keyword arguments): Keyword arguments passed to the `__init__` + method of PoplarExecutor. + """ + + def __init__(self, + model, + logger=None, + training=True, + modules_to_record=None, + *args, + **kwargs): + # self.model == self._user_model: input pytorch model + # self._model: wrapped model which is used to compile + # and update weights, these two models use same weights + # wrapped model only accept and output tuple, so + # HierarchicalDataManager will convert dictionary + # to tuple and convert them back + self.inputs_manager = HierarchicalDataManager(logger=logger) + self.outputs_manager = HierarchicalDataManager(logger=logger) + self.logger = logger + # the features calculated by CPU + self.inter_outputs_in_cpu = {} + # the features calculated by IPU + self.inter_outputs_in_ipu = {} + if modules_to_record is None: + # It is possible that the IPU implementation of some operators + # is inconsistent with the expected (CPU), here you can use + # this method to confirm whether there is a problem + self.compare_with_cpu = False + else: + self.compare_with_cpu = True + # move model.fp16_enabled to self.fp16_enabled, + # modify the position where the input is automatically casted to half + if getattr(model, 'fp16_enabled', False): + model.fp16_enabled = False + self.fp16_enabled = True + # make torch.jit.trace convert self._model + model = WrappedNet( + model, + self.inputs_manager, + self.outputs_manager, + self.inter_outputs_in_cpu, + modules_to_record=modules_to_record) + super().__init__(model, training=training, *args, **kwargs) + # overwrite self._args_parser in train_step or val_step + self._args_parser = None + if training: + assert self.training + else: + assert not self.training + + @property + def training(self): + # If trying to get the attribute(training) of self, + # since the class has no training attribute, + # it will automatically look for the training attribute of self.model. + # However, the real attribute we want to check is self._training, + # self.model.training and self._training are often inconsistent. + # It is not clear whether it is a Poptorch bug or a special design, + # temporarily use this function to fix the problem + return self._training # comes from self.model._training + + @auto_fp16(supported_types=(PoplarExecutor, )) + def run_model(self, data_dict): + # this function is used to parse input_dict + # and convert to output_dict + if self.isCompiled(): + self.inputs_manager.record_hierarchical_data(data_dict) + inputs_tuple = tuple(self.inputs_manager.collect_all_tensors()) + else: + # get tensors out of data and put them in a tuple + self.inputs_manager.record_hierarchical_data(data_dict) + inputs_tuple = tuple(self.inputs_manager.collect_all_tensors()) + # turn logger in data manager off after compilation + self.inputs_manager.quick() + self.outputs_manager.quick() + + # parser args in the first iter + if self._args_parser is None: + self._args_parser = DictArgsParser({'args': inputs_tuple}) + + # run or convert model + # the plain_outputs will be used in converting stage + plain_outputs = self(inputs_tuple) + + self.inputs_manager.clean_all_tensors() + + # put list of tensors back to the output dict + # according to the same order + self.outputs_manager.update_all_tensors(plain_outputs) + # get the real output dictionary from self.outputs_manager + output_dict = self.outputs_manager.hierarchical_data + + # split output_dict into inter_outputs_in_ipu + # and output of the torch model + torch_model_output = {} + for name in output_dict: + if name in self.inter_outputs_in_cpu: + self.inter_outputs_in_ipu[name] = output_dict[name] + else: + torch_model_output[name] = output_dict[name] + + if 'output of WrappedNet: single tensor' in output_dict: + assert len(torch_model_output) == 1 + assert isinstance( + torch_model_output['output of WrappedNet: single tensor'], + torch.Tensor) + torch_model_output = \ + torch_model_output['output of WrappedNet: single tensor'] + + return torch_model_output + + def train_step(self, data, optimizer=None, **kwargs): + # arguments from mmcls/models/classifiers/base.py: + # BaseClassifier.train_step + assert self.training + assert len(kwargs) == 0 # TODO, support later if necessary + + # TODO support datacontainer as input + # currently, auto_fp16 and HierarchicalDataManager take too much + # time on traversing datacontainer + data['img_metas'] = None + num_samples = len(data['img'].data) + + # TODO we will ignore optimizer because it will not be used in model, + # support later if necessary + data['optimizer'] = None + output_dict = self.run_model(data) + + # outputs contained loss, log_vars, num_samples, + # only loss(torch.tensor) has been updated + # remove all unchanged vars, left torch.tensor + neat_output_dict = {'loss': output_dict['loss']} + + # re-parse outputs, get back log_vars and num_samples + loss, log_vars = self.model._parse_losses(neat_output_dict) + final_output_dict = dict( + loss=loss, log_vars=log_vars, num_samples=num_samples) + return final_output_dict + + def eval_call(self, img, img_metas=None, return_loss=True, **kwargs): + # arguments from mmdet/models/detectors/base.py:BaseDetector.forward + # tmp usssage for eval mode + assert not self.training + assert len(kwargs) == 0 # TODO, support later if necessary + assert not return_loss + data = {'img': img, 'img_metas': img_metas, 'return_loss': return_loss} + + output_dict = self.run_model(data) + + return output_dict + + def detachFromDevice(self): + if self.isCompiled() and self._is_attached: + super().detachFromDevice() + + def attachToDevice(self): + if self.isCompiled() and not self._is_attached: + super().attachToDevice() + + +class TrainEvalModel: + """A class maintaining training MMPoplarExecutor and inference + MMPoplarExecutor. + + Args: + train_model (:obj:`nn.Module`): The training model to be compiled. + ``train_model`` can be None if only executing validation. + eval_model (:obj:`nn.Module`): The inference model to be compiled. + options (mmcv.Config, dict): Options that will be used to compile + and run the model. + optimizer (:obj:`torch.optim.Optimizer`, optional): torch + optimizer, necessary if in training mode + logger (:obj:`logging.Logger`): Logger used during running. + Defaults to None. + modules_to_record (mmcv.Config, list): Index or name of modules which + will be recorded for output. It is necessary to specify output for + static graph of model training or inference. + """ + + def __init__(self, + train_model, + eval_model, + options, + optimizer, + modules_to_record=None, + logger=None): + if train_model is None: + self._train_executor = None + self.training = False + else: + self._train_executor = get_training_model( + train_model, + options=options['training'], + optimizer=optimizer, + logger=logger, + modules_to_record=modules_to_record) + self.training = True + self._eval_executor = get_inference_model( + eval_model, options=options['inference'], logger=logger) + + @property + def executor(self): + if self.training: + return self._train_executor + else: + return self._eval_executor + + def train(self, mode: bool = True): + """Sets the module in training mode. + + This has any effect only on certain modules. See documentations of + particular modules for details of their behaviors in + training/evaluation mode, if they are affected, + e.g. :class:`Dropout`, :class:`BatchNorm`, etc. + + Args: + mode (bool): whether to set training mode (``True``) or evaluation + mode (``False``). Default: ``True``. + + Returns: + Module: self + """ + if not isinstance(mode, bool): + raise ValueError('training mode is expected to be boolean, ' + f'but got {type(mode)}') + if self._train_executor is None and mode: + raise RuntimeError( + 'The train_executor is not initialized.' + 'If you want to initialize train_executor,' + 'you need to input optimizer when converting pytorch model') + + if mode == self.training: + self.model.train(mode) + return self + else: + if self.isCompiled(): + # copy weights from IPU to cpu before off-load current session + self.copyWeightsToHost() + # detach the current session before change the mode, + # if is training mode and weights are updated, + # poptorch will copy weights from IPU to host + self.detachFromDevice() + + self.training = mode # session will changed with mode changing + self.model.train(mode) + + # after changing mode, attach the current new session, + # and this function will copy weights of model to device + self.attachToDevice() + return self + + def eval(self): + """Sets the module in evaluation mode. + + This has any effect only on certain modules. + See documentations of particular modules + for details of their behaviors in training/evaluation mode, + if they are affected, e.g. :class:`Dropout`, :class:`BatchNorm`, etc. + + This is equivalent with :meth:`self.train(False) + `. + + See :ref:`locally-disable-grad-doc` for a comparison between + `.eval()` and several similar mechanisms that may be confused with it. + + Returns: + Module: self + """ + return self.train(False) + + def compare_data_between_ipu_and_cpu(self, inter_outputs_in_cpu, + inter_outputs_in_ipu): + for key, val in inter_outputs_in_cpu.items(): + is_tensor = isinstance(val['fea_in'], torch.Tensor) + fea_in_cpu = val['fea_in'] + fea_in_cpu_list = [fea_in_cpu] if is_tensor else fea_in_cpu + fea_in_ipu = inter_outputs_in_ipu[key]['fea_in'] + fea_in_ipu_list = [fea_in_ipu] if is_tensor else fea_in_ipu + + is_tensor = isinstance(val['fea_out'], torch.Tensor) + fea_out_cpu = val['fea_out'] + fea_out_cpu_list = [fea_out_cpu] if is_tensor else fea_out_cpu + fea_out_ipu = inter_outputs_in_ipu[key]['fea_out'] + fea_out_ipu_list = [fea_out_ipu] if is_tensor else fea_out_ipu + + print('comparing layer:', key) + for idx, (featA, featB) in \ + enumerate(zip(fea_in_cpu_list, fea_in_ipu_list)): + print('fea_in, tensor ', idx) + compare_ndarray(featA.detach().numpy(), featB.detach().numpy()) + for idx, (featA, featB) in \ + enumerate(zip(fea_out_cpu_list, fea_out_ipu_list)): + print('fea_out, tensor', idx) + compare_ndarray(featA.detach().numpy(), featB.detach().numpy()) + + # TODO Unified training and eval interface, + # merge train_step(train) and __call__(eval) together + def train_step(self, data, optimizer=None, **kwargs): + assert self.training, 'not supported train_step on eval mode' + inter_outputs_in_cpu = {} + if (self._train_executor.isCompiled() + and self._train_executor.compare_with_cpu): + self.copyWeightsToHost() + # run in CPU mode + self._train_executor.model.train_step(data, optimizer, **kwargs) + inter_outputs_in_cpu = { + **(self._train_executor.inter_outputs_in_cpu) + } + # run in IPU mode + result = self._train_executor.train_step(data, optimizer, **kwargs) + if (self._train_executor.isCompiled() + and self._train_executor.compare_with_cpu + and len(inter_outputs_in_cpu) > 0): + self.compare_data_between_ipu_and_cpu( + inter_outputs_in_cpu, + self._train_executor.inter_outputs_in_ipu) + return result + + # TODO Unified training and eval interface, + # merge train_step(train) and __call__(eval) together + def __call__(self, *args, **kwargs): + if self.training: + raise NotImplementedError('use train_step rather than __call__') + else: + return self._eval_executor.eval_call(*args, **kwargs) + + def __getattr__(self, attr): + return getattr(self.executor, attr) + + +def get_training_model(model: nn.Module, + options: Optional[poptorch.Options] = None, + optimizer: Optional[torch.optim.Optimizer] = None, + logger=None, + modules_to_record=None) -> poptorch.PoplarExecutor: + """Create a PopTorch training model from a PyTorch model, running on IPU + hardware in training mode. + + Note: + PopTorch makes a shallow copy of the model. Changes to the + parameters in the returned training model affect the original model + and vice versa. However, primitive variable types are not synced: for + example calling ``model.train()`` on the original model, which + changes the ``training`` bool of the model instance, will not alter the + model returned by this function. You may need to call ``model.train()`` + on your model before you call this function for correct behavior. + + Args: + model (:obj:`nn.Module`): The model to run. + options (poptorch.Options): Options that will be used to compile + and run the model. + optimizer (:obj:`torch.optim.Optimizer`, optional): The optimizers + to apply during training. + logger (:obj:`logging.Logger`): Logger used during running. + Defaults to None. + modules_to_record (mmcv.Config, list): Index or name of modules which + will be recorded for output. It is necessary to specify output for + static graph of model training or inference. + + Returns: + The :class:`poptorch.PoplarExecutor` wrapper to use in place + of ``model``. + """ + # Create a copy of the original model in case it needs to be wrapped + maybe_wrapped_model = copy.copy(model) + + return MMPoplarExecutor( + model=maybe_wrapped_model, + logger=logger, + options=options, + training=True, + optimizer=optimizer, + user_model=model, + modules_to_record=modules_to_record, + poptorch_version=__version__) + + +def get_inference_model(model: Union[nn.Module, poptorch.PoplarExecutor], + options: Optional[poptorch.Options] = None, + logger=None) -> poptorch.PoplarExecutor: + """Create a PopTorch inference model from a PyTorch model, running on IPU + hardware in inference mode. + + Note: + PopTorch makes a shallow copy of the model. Changes to the + parameters in the returned inference model affect the original model + and vice versa. However, primitive variable types are not synced: for + example calling ``model.eval()`` on the original model will not alter + the model returned by this function. You may need to call + ``model.eval()`` on your model before you call this function for + correct behavior. + + Args: + model (:obj:`nn.Module`): The model to run. + options (poptorch.Options): Options that will be used to compile + and run the model. + logger (:obj:`logging.Logger`): Logger used during running. + Defaults to None. + + Returns: + The :class:`poptorch.PoplarExecutor` wrapper to use in place of + ``model``. + """ + + return MMPoplarExecutor( + model=copy.copy(model), + logger=logger, + options=options, + training=False, + poptorch_version=__version__) + + +def ipu_model_wrapper(model, + options, + optimizer=None, + logger=None, + modules_to_record=None, + ipu_model_cfg=None, + fp16_cfg=None): + """Convert torch model to IPU model. + + Args: + model (nn.Module): The target model to be converted. + options (dict[str, poptorch.Options]): IPU options, generated + by :func:`cfg2options`. + optimizer (:obj:`torch.optim.Optimizer`, optional): torch + optimizer, necessary if in training mode + logger (:obj:`logging.Logger`): Logger used during training. + modules_to_record (mmcv.Config, list): Index or name of modules which + will be recorded for output. It is necessary to specify output for + static graph of model training or inference. + ipu_model_cfg (dict): A dictionary contains train_split_edges and + train_ckpt_nodes, See details in :func:`model_sharding` and + :func:`recomputation_checkpoint` functions. + fp16_cfg (dict): Config for IPU fp16 training. Currently supports + configs: `loss_scale`, `velocity_accum_type` and `accum_type`. + See details in + https://docs.graphcore.ai/projects/poptorch-user-guide/en/latest/index.html + + Returns: + TrainEvalModel: IPU wrapped model. + """ + if ipu_model_cfg is None: + ipu_model_cfg = {} + training = model.training if optimizer is not None else False + # set mixed-precision + if fp16_cfg is not None: + from mmcv.runner import wrap_fp16_model + loss_scale = fp16_cfg['loss_scale'] + wrap_fp16_model(model) + model.half() + # TODO tmp ussage to set loss scaling for torch original optimizer + if optimizer is not None: + optimizer.loss_scaling = loss_scale + if fp16_cfg.get('velocity_accum_type', False): + if fp16_cfg['velocity_accum_type'] == 'half': + optimizer.velocity_accum_type = torch.half + else: + optimizer.velocity_accum_type = torch.float32 + if fp16_cfg.get('accum_type', False): + if fp16_cfg['accum_type'] == 'half': + optimizer.accum_type = torch.half + else: + optimizer.accum_type = torch.float32 + # TODO support feature alignment for fp16 + if modules_to_record is not None: + raise NotImplementedError( + 'Feature alignment for fp16 is not implemented') + + # set model partition + if optimizer is None: + train_model = None + else: + # split model into multi-IPUs if specified + train_model = model_sharding( + copy.copy(model).train(), + ipu_model_cfg.get('train_split_edges', [])) + + recomputation_checkpoint(train_model, + ipu_model_cfg.get('train_ckpt_nodes', [])) + + # TODO support feature alignment for gradient accumulation mode + gradient_accumulation = \ + getattr(options['training'].Training, 'gradient_accumulation', 1) + if gradient_accumulation > 1: + assert modules_to_record is None, \ + 'Feature alignment for grad-accumulation mode not implemented' + + # TODO support feature alignment for multi-replica mode + replication_factor = \ + getattr(options['training'], 'replication_factor', 1) + if replication_factor > 1: + assert modules_to_record is None, \ + 'Feature alignment for multi-replica mode not implemented' + + # TODO supports different model partitions between train and eval mode + assert len(ipu_model_cfg.get('eval_split_edges', [])) == 0,\ + 'Currently, BeginBlock can only be used once on the same model' + eval_model = copy.copy(model).eval() + + # wrap model for compilation + model = TrainEvalModel( + train_model, + eval_model, + options=options, + optimizer=optimizer, + logger=logger, + modules_to_record=modules_to_record) + model.train(training) + return model diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/runner.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/runner.py new file mode 100755 index 0000000000000000000000000000000000000000..e2d4922677e08b2d6b5132a01034de8b043fa3f1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/runner.py @@ -0,0 +1,142 @@ +# Copyright (c) OpenMMLab. All rights reserved. + +from mmcv.runner import (HOOKS, RUNNERS, BaseRunner, EpochBasedRunner, + IterBasedRunner) +from mmcv.utils import IS_IPU_AVAILABLE + +if IS_IPU_AVAILABLE: + from .dataloader import IPUDataLoader + from .hook_wrapper import (IPUFp16OptimizerHook, wrap_lr_updater_hook, + wrap_optimizer_hook) + from .model_wrapper import ipu_model_wrapper + from .utils import build_from_cfg_with_wrapper, cfg2options + + +class IPUBaseRunner(BaseRunner): + """A base runner for IPU. + + This runner has some extra processes for IPU which are shown below: + + 1. Parse options for IPU + 2. wrap pytorch model for IPU + 3. Raise errors while encountering illegal usage + 4. Input IPU options and initialize dataloader if finding an instance + of IPUDataLoader + + Args: + model (:obj:`nn.Module`): The model to run. + options_cfg (mmcv.Config, dict): Options that will be used to compile + and run the model. + modules_to_record (mmcv.Config, list): Index or name of modules which + will be recorded for output. It is necessary to specify output for + static graph of model training or inference. + ipu_model_cfg (mmcv.Config, dict): Config of model partition and + recomputing checkpoint + fp16_cfg (mmcv.Config): Config for fp16 training. + batch_processor (callable): A callable method that process a data + batch. Should be None for IPU runner + kwargs (Dict[str, Any], optional): Keyword arguments will be passed to + ``base_runner.BaseRunner``. + """ + + def __init__(self, + model, + options_cfg=None, + modules_to_record=None, + ipu_model_cfg=None, + fp16_cfg=None, + batch_processor=None, + **kwargs): + assert hasattr(model, 'train_step') and batch_processor is None,\ + 'only support model with train_step' + + if options_cfg is None: + options_cfg = {} + # call BaseRunner.__init__() here + super().__init__(model, **kwargs) + + # process options of ipu + if IS_IPU_AVAILABLE: + self.options = cfg2options(options_cfg) + self.model = ipu_model_wrapper( + self.model, + self.options, + self.optimizer, + self.logger, + modules_to_record=modules_to_record, + ipu_model_cfg=ipu_model_cfg, + fp16_cfg=fp16_cfg) + else: + raise NotImplementedError('cpu mode on IPURunner is not supported') + + def register_lr_hook(self, lr_config): + if lr_config is None: + return + assert isinstance(lr_config, dict) + assert 'policy' in lr_config + policy_type = lr_config.pop('policy') + # If the type of policy is all in lower case, + # e.g., 'cyclic', then its first letter will be capitalized, + # e.g., to be 'Cyclic'. + # This is for the convenient usage of Lr updater. + # Since this is not applicable for ` + # CosineAnnealingLrUpdater`, the string will not be changed + # if it contains capital letters. + if policy_type == policy_type.lower(): + policy_type = policy_type.title() + hook_type = policy_type + 'LrUpdaterHook' + lr_config['type'] = hook_type + hook = build_from_cfg_with_wrapper(lr_config, HOOKS, + wrap_lr_updater_hook) + self.register_hook(hook, priority='VERY_HIGH') + + def register_optimizer_hook(self, optimizer_config): + if optimizer_config is None: + return + assert isinstance(optimizer_config, (dict, IPUFp16OptimizerHook)) + if isinstance(optimizer_config, dict): + optimizer_config.setdefault('type', 'OptimizerHook') + hook = build_from_cfg_with_wrapper(optimizer_config, HOOKS, + wrap_optimizer_hook) + else: + hook = optimizer_config + self.register_hook(hook, priority='ABOVE_NORMAL') + + def run(self, data_loaders, workflow, *args, **kwargs): + for i, flow in enumerate(workflow): + mode, _ = flow + # initialize IPU dataloader if not initialized + assert isinstance(data_loaders[i], IPUDataLoader),\ + 'IPU runner can only work with `IPUDataLoader`' + data_loaders[i].init(options=self.get_options(mode)) + + super().run(data_loaders, workflow, *args, **kwargs) + + def get_options(self, mode): + if mode == 'train': + return self.options['training'] + elif mode == 'val': + return self.options['inference'] + else: + raise ValueError(f'mode should be train or val but got {mode}') + + +@RUNNERS.register_module() +class IPUEpochBasedRunner(IPUBaseRunner, EpochBasedRunner): + """Epoch-based Runner for IPU. + + The Inheritance order(MRO) is: IPUEpochBasedRunner -> IPUBaseRunner -> + EpochBasedRunner -> BaseRunner This runner train models epoch by epoch. + """ + pass + + +@RUNNERS.register_module() +class IPUIterBasedRunner(IPUBaseRunner, IterBasedRunner): + """Iteration-based Runner for IPU. + + The Inheritance order(MRO) is: IPUIterBasedRunner -> IPUBaseRunner -> + IterBasedRunner -> BaseRunner This runner train models iteration by + iteration. + """ + pass diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/utils.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/utils.py new file mode 100755 index 0000000000000000000000000000000000000000..79709db1ee1282e8daa6614ceb23481d3cd58338 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/ipu/utils.py @@ -0,0 +1,244 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import inspect + +import numpy as np +import popart +import poptorch +import torch +import torch.nn as nn + +from mmcv.utils import Registry + + +def _options_assigner(cfg, options_node): + # set popart.options by config + # cfg: dict, python data type + # options_node: python module or function + if isinstance(cfg, dict): + for key in cfg: + _options_assigner(cfg[key], getattr(options_node, key)) + elif isinstance(cfg, (int, float, str, list)): + if callable(options_node): + options_node(cfg) + else: + error_msg = f'options_node type {type(options_node)} not supported' + raise NotImplementedError(error_msg) + else: + error_msg = f'cfg type {type(cfg)} not supported' + raise NotImplementedError(error_msg) + + +def cfg2options(cfg): + """Parse dictionary to ipu options. + + Args: + cfg (dict): A dictionary of ipu settings. + + Returns: + dict[str, poptorch.Options]: Training options and inference options + of IPU. + """ + # set ipu options for inference and training by config + train_cfg = cfg.pop('train_cfg', {}) + eval_cfg = cfg.pop('eval_cfg', {}) + eval_cfg['replicationFactor'] = 1 # eval mode only use one replica + eval_cfg['executionStrategy'] = 'ShardedExecution' + # overwrite default ipu cfg with specified train cfgs + training_ipu_cfg = {**cfg, **train_cfg} + # overwrite default ipu cfg with specified eval cfgs + inference_ipu_cfg = {**cfg, **eval_cfg} + + ipu_options = { + 'training': _cast_to_options(training_ipu_cfg), + 'inference': _cast_to_options(inference_ipu_cfg) + } + + # TODO configure these codes + ipu_options['training']._Popart.set('disableGradAccumulationTensorStreams', + True) + ipu_options['training']._Popart.set( + 'accumulateOuterFragmentSettings.schedule', + int(popart.AccumulateOuterFragmentSchedule.OverlapMemoryOptimized)) + ipu_options['training'].Precision.enableStochasticRounding(True) + + return ipu_options + + +def _cast_to_options(cfg): + # If it cannot be directly assigned, use if statement to parse it, + # and if it can be directly assigned, use _options_assigner to assign + options = poptorch.Options() + + if 'availableMemoryProportion' in cfg: + available_memory_proportion = cfg.pop('availableMemoryProportion') + mem_props = {} + for i, mem_prop in enumerate(available_memory_proportion): + mem_props[f'IPU{i}'] = mem_prop + options.setAvailableMemoryProportion(mem_props) + + if 'executionStrategy' in cfg: + execution_strategy = cfg.pop('executionStrategy') + if execution_strategy == 'SameAsIpu': + options.setExecutionStrategy( + poptorch.PipelinedExecution( + getattr(poptorch.AutoStage, execution_strategy))) + elif execution_strategy == 'ShardedExecution': + options.setExecutionStrategy(poptorch.ShardedExecution()) + else: + raise NotImplementedError( + 'executionStrategy should be "SameAsIpu" or "ShardedExecution"' + f', but got {execution_strategy}') + + if 'partialsType' in cfg: + partials_type = cfg.pop('partialsType') + options.Precision.setPartialsType(getattr( + torch, partials_type)) # half or float + + _options_assigner(cfg, options) + return options + + +def model_sharding(model, split_edges): + """split models in-place into multi-IPUs. + + Args: + model (nn.Module): The target model to be split. + split_edges (list of dict): Model layer names or layer numbers + of split edge. Each item of ``split_edges`` is a dictionary, + which may contain the following key-pairs: + + - layer_to_call: PyTorch module to assign to the block + - user_id (optional): A user defined identifier for the block. + - ipu_id: The id of the IPU to run on. + + Examples: + >>> split_edges = [ + ... dict(layer_to_call='model.conv1', ipu_id=0), + ... dict(layer_to_call='model.conv3', ipu_id=1)] + >>> sharding_model = model_sharding(torch_model, split_edges) + + Returns: + nn.Module: Split model. + """ + if len(split_edges) == 0: + return model + assert isinstance(split_edges, list) + spilt_edges_dict = {edge['layer_to_call']: edge for edge in split_edges} + + for idx, (name, module) in enumerate(model.named_modules()): + if idx in spilt_edges_dict and name in spilt_edges_dict: + raise ValueError( + 'The same layer is referenced twice while doing model' + f' partition: idx is {idx} and name is {name}') + + edge = spilt_edges_dict.pop(name, None) + edge = spilt_edges_dict.pop(idx, edge) + if edge is not None: + poptorch.BeginBlock(module, edge.get('user_id', name), + edge['ipu_id']) + + # ensure all split_edges are used + if len(spilt_edges_dict) > 0: + split_edge_names = list(spilt_edges_dict.keys()) + raise RuntimeError( + f'split_edges: {split_edge_names} are not contained in the model') + return model + + +def recomputation_checkpoint(model: nn.Module, module_names: list): + """Annotates the output of a module to be checkpointed instead of + recomputed. + + If recomputation mode is enabled, ipu will release the activations of + the middle layers to save memory. During the backward of gradient, + the activation of the middle layer will be recalculated again. + This function is used to declare the activations of some intermediate + layers that need to be saved in order to skip the recomputation of + some layers. + + Args: + model (nn.Module): The target model to apply recomputation + checkpoint. + module_names (list): Layer names of module. + """ + + def recompute_outputs(module, inputs, outputs): + if isinstance(outputs, tuple): + return tuple(poptorch.recomputationCheckpoint(y) for y in outputs) + else: + return poptorch.recomputationCheckpoint(outputs) + + for name, module in model.named_modules(): + if name in module_names: + module.register_forward_hook(recompute_outputs) + module_names.remove(name) + + # check all module_names are used + assert len(module_names) == 0,\ + f'recomputed nodes: {module_names} are not contained in the model' + + +def compare_ndarray(featA, featB, rtol=1e-3, atol=1e-5): + """Align data between two activations or weights.""" + try: + np.testing.assert_allclose(featA, featB, rtol=rtol, atol=atol) + except AssertionError as e: + print(e) + + +def build_from_cfg_with_wrapper(cfg, + registry, + wrapper_func=None, + default_args=None): + """Build a module from config dict and wrap module with "wrapper_func". + + Args: + cfg (dict): Config dict. It should at least contain the key "type". + registry (:obj:`Registry`): The registry to search the type from. + default_args (dict, optional): Default initialization arguments. + wrapper_func (function): Used to wrap class + + Returns: + object: The constructed object. + """ + if not isinstance(cfg, dict): + raise TypeError(f'cfg must be a dict, but got {type(cfg)}') + if 'type' not in cfg: + if default_args is None or 'type' not in default_args: + raise KeyError( + '`cfg` or `default_args` must contain the key "type", ' + f'but got {cfg}\n{default_args}') + if not isinstance(registry, Registry): + raise TypeError('registry must be an mmcv.Registry object, ' + f'but got {type(registry)}') + if not (isinstance(default_args, dict) or default_args is None): + raise TypeError('default_args must be a dict or None, ' + f'but got {type(default_args)}') + + args = cfg.copy() + + if default_args is not None: + for name, value in default_args.items(): + args.setdefault(name, value) + + obj_type = args.pop('type') + if isinstance(obj_type, str): + obj_cls = registry.get(obj_type) + if obj_cls is None: + raise KeyError( + f'{obj_type} is not in the {registry.name} registry') + elif inspect.isclass(obj_type): + obj_cls = obj_type + else: + raise TypeError( + f'type must be a str or valid type, but got {type(obj_type)}') + + if wrapper_func is None: + wrapped_obj_cls = obj_cls + else: + wrapped_obj_cls = wrapper_func(obj_cls) + try: + return wrapped_obj_cls(**args) + except Exception as e: + # Normal TypeError does not print class name. + raise type(e)(f'{wrapped_obj_cls.__name__}: {e}') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..572c4da7ee898b370f833baf8c3e3e7b68cd0050 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/__init__.py @@ -0,0 +1,9 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .data_parallel import MLUDataParallel +from .distributed import MLUDistributedDataParallel +from .scatter_gather import scatter, scatter_kwargs + +__all__ = [ + 'MLUDataParallel', 'MLUDistributedDataParallel', 'scatter', + 'scatter_kwargs' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/_functions.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/_functions.py new file mode 100644 index 0000000000000000000000000000000000000000..7c35e65a217b5160cdd89b28e6c3194d2fbfcff2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/_functions.py @@ -0,0 +1,22 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch + + +def scatter(input, devices): + """scatter copies tensor to MLU directly.""" + if isinstance(input, list): + outputs = [scatter(_input, devices) for _input in input] + return outputs + elif isinstance(input, torch.Tensor): + output = input.contiguous() + return output.to('mlu') if devices != [-1] else output + else: + raise Exception(f'Unknown type {type(input)}.') + + +class Scatter: + + @staticmethod + def forward(target_mlus, input): + outputs = scatter(input, target_mlus) + return tuple(outputs) if isinstance(outputs, list) else (outputs, ) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/data_parallel.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/data_parallel.py new file mode 100644 index 0000000000000000000000000000000000000000..b2d09d0b04dcce5e10d5344618423a256b157c86 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/data_parallel.py @@ -0,0 +1,41 @@ +# Copyright (c) OpenMMLab. All rights reserved. + +import torch + +from mmcv.parallel import MMDataParallel +from .scatter_gather import scatter_kwargs + + +class MLUDataParallel(MMDataParallel): + """The MLUDataParallel module that supports DataContainer. + + MLUDataParallel is a class inherited from MMDataParall, which supports + MLU training and inference only. + + The main differences with MMDataParallel: + + - It only supports single-card of MLU, and only use first card to + run training and inference. + + - It uses direct host-to-device copy instead of stream-background + scatter. + + .. warning:: + MLUDataParallel only supports single MLU training, if you need to + train with multiple MLUs, please use MLUDistributedDataParallel + instead. If you have multiple MLUs, you can set the environment + variable ``MLU_VISIBLE_DEVICES=0`` (or any other card number(s)) + to specify the running device. + + Args: + module (:class:`nn.Module`): Module to be encapsulated. + dim (int): Dimension used to scatter the data. Defaults to 0. + """ + + def __init__(self, *args, dim=0, **kwargs): + super(MLUDataParallel, self).__init__(*args, dim=dim, **kwargs) + self.device_ids = [0] + self.src_device_obj = torch.device('mlu:0') + + def scatter(self, inputs, kwargs, device_ids): + return scatter_kwargs(inputs, kwargs, device_ids, dim=self.dim) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/distributed.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/distributed.py new file mode 100644 index 0000000000000000000000000000000000000000..3768c754c908b219fd5a770d69e6ed5416781ba8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/distributed.py @@ -0,0 +1,20 @@ +# Copyright (c) OpenMMLab. All rights reserved. + +from mmcv.parallel import MMDistributedDataParallel +from .scatter_gather import scatter_kwargs + + +class MLUDistributedDataParallel(MMDistributedDataParallel): + """The DDP module supports DataContainer. + + MLUDDP has one difference from MMDDP which moves data to MLU with coping + instead of scattering. + """ + + def to_kwargs(self, inputs, kwargs, device_id): + # Use `self.to_kwargs` instead of `self.scatter` in pytorch1.8 + # to move all tensors to device_id + return scatter_kwargs(inputs, kwargs, [device_id], dim=self.dim) + + def scatter(self, inputs, kwargs, device_ids): + return scatter_kwargs(inputs, kwargs, device_ids, dim=self.dim) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/scatter_gather.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/scatter_gather.py new file mode 100644 index 0000000000000000000000000000000000000000..0b0c9b96f51252e4c510f66a2ec5fb7522716e29 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/device/mlu/scatter_gather.py @@ -0,0 +1,59 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch + +from mmcv.parallel.data_container import DataContainer +from ._functions import Scatter + + +def scatter(inputs, target_mlus, dim=0): + """Scatter inputs to target mlu. + + The only difference from original :func:`scatter` is to add support for + :type:`~mmcv.parallel.DataContainer`. + """ + + def scatter_map(obj): + if isinstance(obj, torch.Tensor): + if target_mlus != [-1]: + obj = obj.to('mlu') + return [obj] + else: + # for CPU inference we use self-implemented scatter + return Scatter.forward(target_mlus, obj) + if isinstance(obj, DataContainer): + if obj.cpu_only: + return obj.data + else: + return Scatter.forward(target_mlus, obj.data) + if isinstance(obj, tuple) and len(obj) > 0: + return list(zip(*map(scatter_map, obj))) + if isinstance(obj, list) and len(obj) > 0: + out = list(map(list, zip(*map(scatter_map, obj)))) + return out + if isinstance(obj, dict) and len(obj) > 0: + out = list(map(type(obj), zip(*map(scatter_map, obj.items())))) + return out + return [obj for targets in target_mlus] + + # After scatter_map is called, a scatter_map cell will exist. This cell + # has a reference to the actual function scatter_map, which has references + # to a closure that has a reference to the scatter_map cell (because the + # fn is recursive). To avoid this reference cycle, we set the function to + # None, clearing the cell + try: + return scatter_map(inputs) + finally: + scatter_map = None + + +def scatter_kwargs(inputs, kwargs, target_mlus, dim=0): + """Scatter with support for kwargs dictionary.""" + inputs = scatter(inputs, target_mlus, dim) if inputs else [] + kwargs = scatter(kwargs, target_mlus, dim) if kwargs else [] + if len(inputs) < len(kwargs): + inputs.extend([() for _ in range(len(kwargs) - len(inputs))]) + elif len(kwargs) < len(inputs): + kwargs.extend([{} for _ in range(len(inputs) - len(kwargs))]) + inputs = tuple(inputs) + kwargs = tuple(kwargs) + return inputs, kwargs diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/engine/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/engine/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..3193b7f664e19ce2458d81c836597fa22e4bb082 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/engine/__init__.py @@ -0,0 +1,8 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .test import (collect_results_cpu, collect_results_gpu, multi_gpu_test, + single_gpu_test) + +__all__ = [ + 'collect_results_cpu', 'collect_results_gpu', 'multi_gpu_test', + 'single_gpu_test' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/engine/test.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/engine/test.py new file mode 100644 index 0000000000000000000000000000000000000000..f236b1cda2f39517bda3e4cce9badc19c6cbf190 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/engine/test.py @@ -0,0 +1,202 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import pickle +import shutil +import tempfile +import time + +import torch +import torch.distributed as dist + +import mmcv +from mmcv.runner import get_dist_info + + +def single_gpu_test(model, data_loader): + """Test model with a single gpu. + + This method tests model with a single gpu and displays test progress bar. + + Args: + model (nn.Module): Model to be tested. + data_loader (nn.Dataloader): Pytorch data loader. + + Returns: + list: The prediction results. + """ + model.eval() + results = [] + dataset = data_loader.dataset + prog_bar = mmcv.ProgressBar(len(dataset)) + for data in data_loader: + with torch.no_grad(): + result = model(return_loss=False, **data) + results.extend(result) + + # Assume result has the same length of batch_size + # refer to https://github.com/open-mmlab/mmcv/issues/985 + batch_size = len(result) + for _ in range(batch_size): + prog_bar.update() + return results + + +def multi_gpu_test(model, data_loader, tmpdir=None, gpu_collect=False): + """Test model with multiple gpus. + + This method tests model with multiple gpus and collects the results + under two different modes: gpu and cpu modes. By setting + ``gpu_collect=True``, it encodes results to gpu tensors and use gpu + communication for results collection. On cpu mode it saves the results on + different gpus to ``tmpdir`` and collects them by the rank 0 worker. + + Args: + model (nn.Module): Model to be tested. + data_loader (nn.Dataloader): Pytorch data loader. + tmpdir (str): Path of directory to save the temporary results from + different gpus under cpu mode. + gpu_collect (bool): Option to use either gpu or cpu to collect results. + + Returns: + list: The prediction results. + """ + model.eval() + results = [] + dataset = data_loader.dataset + rank, world_size = get_dist_info() + if rank == 0: + prog_bar = mmcv.ProgressBar(len(dataset)) + time.sleep(2) # This line can prevent deadlock problem in some cases. + for i, data in enumerate(data_loader): + with torch.no_grad(): + result = model(return_loss=False, **data) + results.extend(result) + + if rank == 0: + batch_size = len(result) + batch_size_all = batch_size * world_size + if batch_size_all + prog_bar.completed > len(dataset): + batch_size_all = len(dataset) - prog_bar.completed + for _ in range(batch_size_all): + prog_bar.update() + + # collect results from all ranks + if gpu_collect: + results = collect_results_gpu(results, len(dataset)) + else: + results = collect_results_cpu(results, len(dataset), tmpdir) + return results + + +def collect_results_cpu(result_part, size, tmpdir=None): + """Collect results under cpu mode. + + On cpu mode, this function will save the results on different gpus to + ``tmpdir`` and collect them by the rank 0 worker. + + Args: + result_part (list): Result list containing result parts + to be collected. + size (int): Size of the results, commonly equal to length of + the results. + tmpdir (str | None): temporal directory for collected results to + store. If set to None, it will create a random temporal directory + for it. + + Returns: + list: The collected results. + """ + rank, world_size = get_dist_info() + # create a tmp dir if it is not specified + if tmpdir is None: + MAX_LEN = 512 + # 32 is whitespace + dir_tensor = torch.full((MAX_LEN, ), + 32, + dtype=torch.uint8, + device='cuda') + if rank == 0: + mmcv.mkdir_or_exist('.dist_test') + tmpdir = tempfile.mkdtemp(dir='.dist_test') + tmpdir = torch.tensor( + bytearray(tmpdir.encode()), dtype=torch.uint8, device='cuda') + dir_tensor[:len(tmpdir)] = tmpdir + dist.broadcast(dir_tensor, 0) + tmpdir = dir_tensor.cpu().numpy().tobytes().decode().rstrip() + else: + mmcv.mkdir_or_exist(tmpdir) + # dump the part result to the dir + mmcv.dump(result_part, osp.join(tmpdir, f'part_{rank}.pkl')) + dist.barrier() + # collect all parts + if rank != 0: + return None + else: + # load results of all parts from tmp dir + part_list = [] + for i in range(world_size): + part_file = osp.join(tmpdir, f'part_{i}.pkl') + part_result = mmcv.load(part_file) + # When data is severely insufficient, an empty part_result + # on a certain gpu could makes the overall outputs empty. + if part_result: + part_list.append(part_result) + # sort the results + ordered_results = [] + for res in zip(*part_list): + ordered_results.extend(list(res)) + # the dataloader may pad some samples + ordered_results = ordered_results[:size] + # remove tmp dir + shutil.rmtree(tmpdir) + return ordered_results + + +def collect_results_gpu(result_part, size): + """Collect results under gpu mode. + + On gpu mode, this function will encode results to gpu tensors and use gpu + communication for results collection. + + Args: + result_part (list): Result list containing result parts + to be collected. + size (int): Size of the results, commonly equal to length of + the results. + + Returns: + list: The collected results. + """ + rank, world_size = get_dist_info() + # dump result part to tensor with pickle + part_tensor = torch.tensor( + bytearray(pickle.dumps(result_part)), dtype=torch.uint8, device='cuda') + # gather all result part tensor shape + shape_tensor = torch.tensor(part_tensor.shape, device='cuda') + shape_list = [shape_tensor.clone() for _ in range(world_size)] + dist.all_gather(shape_list, shape_tensor) + # padding result part tensor to max length + shape_max = torch.tensor(shape_list).max() + part_send = torch.zeros(shape_max, dtype=torch.uint8, device='cuda') + part_send[:shape_tensor[0]] = part_tensor + part_recv_list = [ + part_tensor.new_zeros(shape_max) for _ in range(world_size) + ] + # gather all result part + dist.all_gather(part_recv_list, part_send) + + if rank == 0: + part_list = [] + for recv, shape in zip(part_recv_list, shape_list): + part_result = pickle.loads(recv[:shape[0]].cpu().numpy().tobytes()) + # When data is severely insufficient, an empty part_result + # on a certain gpu could makes the overall outputs empty. + if part_result: + part_list.append(part_result) + # sort the results + ordered_results = [] + for res in zip(*part_list): + ordered_results.extend(list(res)) + # the dataloader may pad some samples + ordered_results = ordered_results[:size] + return ordered_results diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..2051b85f7e59bff7bdbaa131849ce8cd31f059a4 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/__init__.py @@ -0,0 +1,11 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .file_client import BaseStorageBackend, FileClient +from .handlers import BaseFileHandler, JsonHandler, PickleHandler, YamlHandler +from .io import dump, load, register_handler +from .parse import dict_from_file, list_from_file + +__all__ = [ + 'BaseStorageBackend', 'FileClient', 'load', 'dump', 'register_handler', + 'BaseFileHandler', 'JsonHandler', 'PickleHandler', 'YamlHandler', + 'list_from_file', 'dict_from_file' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/file_client.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/file_client.py new file mode 100644 index 0000000000000000000000000000000000000000..e7fd7cdfa7dacb4604270b8a3492ca3af3ba6afc --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/file_client.py @@ -0,0 +1,1163 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import inspect +import os +import os.path as osp +import re +import tempfile +import warnings +from abc import ABCMeta, abstractmethod +from contextlib import contextmanager +from pathlib import Path +from typing import Iterable, Iterator, Optional, Tuple, Union +from urllib.request import urlopen + +import mmcv +from mmcv.utils.misc import has_method +from mmcv.utils.path import is_filepath + + +class BaseStorageBackend(metaclass=ABCMeta): + """Abstract class of storage backends. + + All backends need to implement two apis: ``get()`` and ``get_text()``. + ``get()`` reads the file as a byte stream and ``get_text()`` reads the file + as texts. + """ + + # a flag to indicate whether the backend can create a symlink for a file + _allow_symlink = False + + @property + def name(self): + return self.__class__.__name__ + + @property + def allow_symlink(self): + return self._allow_symlink + + @abstractmethod + def get(self, filepath): + pass + + @abstractmethod + def get_text(self, filepath): + pass + + +class CephBackend(BaseStorageBackend): + """Ceph storage backend (for internal use). + + Args: + path_mapping (dict|None): path mapping dict from local path to Petrel + path. When ``path_mapping={'src': 'dst'}``, ``src`` in ``filepath`` + will be replaced by ``dst``. Default: None. + + .. warning:: + :class:`mmcv.fileio.file_client.CephBackend` will be deprecated, + please use :class:`mmcv.fileio.file_client.PetrelBackend` instead. + """ + + def __init__(self, path_mapping=None): + try: + import ceph + except ImportError: + raise ImportError('Please install ceph to enable CephBackend.') + + warnings.warn( + 'CephBackend will be deprecated, please use PetrelBackend instead', + DeprecationWarning) + self._client = ceph.S3Client() + assert isinstance(path_mapping, dict) or path_mapping is None + self.path_mapping = path_mapping + + def get(self, filepath): + filepath = str(filepath) + if self.path_mapping is not None: + for k, v in self.path_mapping.items(): + filepath = filepath.replace(k, v) + value = self._client.Get(filepath) + value_buf = memoryview(value) + return value_buf + + def get_text(self, filepath, encoding=None): + raise NotImplementedError + + +class PetrelBackend(BaseStorageBackend): + """Petrel storage backend (for internal use). + + PetrelBackend supports reading and writing data to multiple clusters. + If the file path contains the cluster name, PetrelBackend will read data + from specified cluster or write data to it. Otherwise, PetrelBackend will + access the default cluster. + + Args: + path_mapping (dict, optional): Path mapping dict from local path to + Petrel path. When ``path_mapping={'src': 'dst'}``, ``src`` in + ``filepath`` will be replaced by ``dst``. Default: None. + enable_mc (bool, optional): Whether to enable memcached support. + Default: True. + + Examples: + >>> filepath1 = 's3://path/of/file' + >>> filepath2 = 'cluster-name:s3://path/of/file' + >>> client = PetrelBackend() + >>> client.get(filepath1) # get data from default cluster + >>> client.get(filepath2) # get data from 'cluster-name' cluster + """ + + def __init__(self, + path_mapping: Optional[dict] = None, + enable_mc: bool = True): + try: + from petrel_client import client + except ImportError: + raise ImportError('Please install petrel_client to enable ' + 'PetrelBackend.') + + self._client = client.Client(enable_mc=enable_mc) + assert isinstance(path_mapping, dict) or path_mapping is None + self.path_mapping = path_mapping + + def _map_path(self, filepath: Union[str, Path]) -> str: + """Map ``filepath`` to a string path whose prefix will be replaced by + :attr:`self.path_mapping`. + + Args: + filepath (str): Path to be mapped. + """ + filepath = str(filepath) + if self.path_mapping is not None: + for k, v in self.path_mapping.items(): + filepath = filepath.replace(k, v) + return filepath + + def _format_path(self, filepath: str) -> str: + """Convert a ``filepath`` to standard format of petrel oss. + + If the ``filepath`` is concatenated by ``os.path.join``, in a Windows + environment, the ``filepath`` will be the format of + 's3://bucket_name\\image.jpg'. By invoking :meth:`_format_path`, the + above ``filepath`` will be converted to 's3://bucket_name/image.jpg'. + + Args: + filepath (str): Path to be formatted. + """ + return re.sub(r'\\+', '/', filepath) + + def get(self, filepath: Union[str, Path]) -> memoryview: + """Read data from a given ``filepath`` with 'rb' mode. + + Args: + filepath (str or Path): Path to read data. + + Returns: + memoryview: A memory view of expected bytes object to avoid + copying. The memoryview object can be converted to bytes by + ``value_buf.tobytes()``. + """ + filepath = self._map_path(filepath) + filepath = self._format_path(filepath) + value = self._client.Get(filepath) + value_buf = memoryview(value) + return value_buf + + def get_text(self, + filepath: Union[str, Path], + encoding: str = 'utf-8') -> str: + """Read data from a given ``filepath`` with 'r' mode. + + Args: + filepath (str or Path): Path to read data. + encoding (str): The encoding format used to open the ``filepath``. + Default: 'utf-8'. + + Returns: + str: Expected text reading from ``filepath``. + """ + return str(self.get(filepath), encoding=encoding) + + def put(self, obj: bytes, filepath: Union[str, Path]) -> None: + """Save data to a given ``filepath``. + + Args: + obj (bytes): Data to be saved. + filepath (str or Path): Path to write data. + """ + filepath = self._map_path(filepath) + filepath = self._format_path(filepath) + self._client.put(filepath, obj) + + def put_text(self, + obj: str, + filepath: Union[str, Path], + encoding: str = 'utf-8') -> None: + """Save data to a given ``filepath``. + + Args: + obj (str): Data to be written. + filepath (str or Path): Path to write data. + encoding (str): The encoding format used to encode the ``obj``. + Default: 'utf-8'. + """ + self.put(bytes(obj, encoding=encoding), filepath) + + def remove(self, filepath: Union[str, Path]) -> None: + """Remove a file. + + Args: + filepath (str or Path): Path to be removed. + """ + if not has_method(self._client, 'delete'): + raise NotImplementedError( + ('Current version of Petrel Python SDK has not supported ' + 'the `delete` method, please use a higher version or dev' + ' branch instead.')) + + filepath = self._map_path(filepath) + filepath = self._format_path(filepath) + self._client.delete(filepath) + + def exists(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path exists. + + Args: + filepath (str or Path): Path to be checked whether exists. + + Returns: + bool: Return ``True`` if ``filepath`` exists, ``False`` otherwise. + """ + if not (has_method(self._client, 'contains') + and has_method(self._client, 'isdir')): + raise NotImplementedError( + ('Current version of Petrel Python SDK has not supported ' + 'the `contains` and `isdir` methods, please use a higher' + 'version or dev branch instead.')) + + filepath = self._map_path(filepath) + filepath = self._format_path(filepath) + return self._client.contains(filepath) or self._client.isdir(filepath) + + def isdir(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path is a directory. + + Args: + filepath (str or Path): Path to be checked whether it is a + directory. + + Returns: + bool: Return ``True`` if ``filepath`` points to a directory, + ``False`` otherwise. + """ + if not has_method(self._client, 'isdir'): + raise NotImplementedError( + ('Current version of Petrel Python SDK has not supported ' + 'the `isdir` method, please use a higher version or dev' + ' branch instead.')) + + filepath = self._map_path(filepath) + filepath = self._format_path(filepath) + return self._client.isdir(filepath) + + def isfile(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path is a file. + + Args: + filepath (str or Path): Path to be checked whether it is a file. + + Returns: + bool: Return ``True`` if ``filepath`` points to a file, ``False`` + otherwise. + """ + if not has_method(self._client, 'contains'): + raise NotImplementedError( + ('Current version of Petrel Python SDK has not supported ' + 'the `contains` method, please use a higher version or ' + 'dev branch instead.')) + + filepath = self._map_path(filepath) + filepath = self._format_path(filepath) + return self._client.contains(filepath) + + def join_path(self, filepath: Union[str, Path], + *filepaths: Union[str, Path]) -> str: + """Concatenate all file paths. + + Args: + filepath (str or Path): Path to be concatenated. + + Returns: + str: The result after concatenation. + """ + filepath = self._format_path(self._map_path(filepath)) + if filepath.endswith('/'): + filepath = filepath[:-1] + formatted_paths = [filepath] + for path in filepaths: + formatted_paths.append(self._format_path(self._map_path(path))) + return '/'.join(formatted_paths) + + @contextmanager + def get_local_path(self, filepath: Union[str, Path]) -> Iterable[str]: + """Download a file from ``filepath`` and return a temporary path. + + ``get_local_path`` is decorated by :meth:`contxtlib.contextmanager`. It + can be called with ``with`` statement, and when exists from the + ``with`` statement, the temporary path will be released. + + Args: + filepath (str | Path): Download a file from ``filepath``. + + Examples: + >>> client = PetrelBackend() + >>> # After existing from the ``with`` clause, + >>> # the path will be removed + >>> with client.get_local_path('s3://path/of/your/file') as path: + ... # do something here + + Yields: + Iterable[str]: Only yield one temporary path. + """ + filepath = self._map_path(filepath) + filepath = self._format_path(filepath) + assert self.isfile(filepath) + try: + f = tempfile.NamedTemporaryFile(delete=False) + f.write(self.get(filepath)) + f.close() + yield f.name + finally: + os.remove(f.name) + + def list_dir_or_file(self, + dir_path: Union[str, Path], + list_dir: bool = True, + list_file: bool = True, + suffix: Optional[Union[str, Tuple[str]]] = None, + recursive: bool = False) -> Iterator[str]: + """Scan a directory to find the interested directories or files in + arbitrary order. + + Note: + Petrel has no concept of directories but it simulates the directory + hierarchy in the filesystem through public prefixes. In addition, + if the returned path ends with '/', it means the path is a public + prefix which is a logical directory. + + Note: + :meth:`list_dir_or_file` returns the path relative to ``dir_path``. + In addition, the returned path of directory will not contains the + suffix '/' which is consistent with other backends. + + Args: + dir_path (str | Path): Path of the directory. + list_dir (bool): List the directories. Default: True. + list_file (bool): List the path of files. Default: True. + suffix (str or tuple[str], optional): File suffix + that we are interested in. Default: None. + recursive (bool): If set to True, recursively scan the + directory. Default: False. + + Yields: + Iterable[str]: A relative path to ``dir_path``. + """ + if not has_method(self._client, 'list'): + raise NotImplementedError( + ('Current version of Petrel Python SDK has not supported ' + 'the `list` method, please use a higher version or dev' + ' branch instead.')) + + dir_path = self._map_path(dir_path) + dir_path = self._format_path(dir_path) + if list_dir and suffix is not None: + raise TypeError( + '`list_dir` should be False when `suffix` is not None') + + if (suffix is not None) and not isinstance(suffix, (str, tuple)): + raise TypeError('`suffix` must be a string or tuple of strings') + + # Petrel's simulated directory hierarchy assumes that directory paths + # should end with `/` + if not dir_path.endswith('/'): + dir_path += '/' + + root = dir_path + + def _list_dir_or_file(dir_path, list_dir, list_file, suffix, + recursive): + for path in self._client.list(dir_path): + # the `self.isdir` is not used here to determine whether path + # is a directory, because `self.isdir` relies on + # `self._client.list` + if path.endswith('/'): # a directory path + next_dir_path = self.join_path(dir_path, path) + if list_dir: + # get the relative path and exclude the last + # character '/' + rel_dir = next_dir_path[len(root):-1] + yield rel_dir + if recursive: + yield from _list_dir_or_file(next_dir_path, list_dir, + list_file, suffix, + recursive) + else: # a file path + absolute_path = self.join_path(dir_path, path) + rel_path = absolute_path[len(root):] + if (suffix is None + or rel_path.endswith(suffix)) and list_file: + yield rel_path + + return _list_dir_or_file(dir_path, list_dir, list_file, suffix, + recursive) + + +class MemcachedBackend(BaseStorageBackend): + """Memcached storage backend. + + Attributes: + server_list_cfg (str): Config file for memcached server list. + client_cfg (str): Config file for memcached client. + sys_path (str | None): Additional path to be appended to `sys.path`. + Default: None. + """ + + def __init__(self, server_list_cfg, client_cfg, sys_path=None): + if sys_path is not None: + import sys + sys.path.append(sys_path) + try: + import mc + except ImportError: + raise ImportError( + 'Please install memcached to enable MemcachedBackend.') + + self.server_list_cfg = server_list_cfg + self.client_cfg = client_cfg + self._client = mc.MemcachedClient.GetInstance(self.server_list_cfg, + self.client_cfg) + # mc.pyvector servers as a point which points to a memory cache + self._mc_buffer = mc.pyvector() + + def get(self, filepath): + filepath = str(filepath) + import mc + self._client.Get(filepath, self._mc_buffer) + value_buf = mc.ConvertBuffer(self._mc_buffer) + return value_buf + + def get_text(self, filepath, encoding=None): + raise NotImplementedError + + +class LmdbBackend(BaseStorageBackend): + """Lmdb storage backend. + + Args: + db_path (str): Lmdb database path. + readonly (bool, optional): Lmdb environment parameter. If True, + disallow any write operations. Default: True. + lock (bool, optional): Lmdb environment parameter. If False, when + concurrent access occurs, do not lock the database. Default: False. + readahead (bool, optional): Lmdb environment parameter. If False, + disable the OS filesystem readahead mechanism, which may improve + random read performance when a database is larger than RAM. + Default: False. + + Attributes: + db_path (str): Lmdb database path. + """ + + def __init__(self, + db_path, + readonly=True, + lock=False, + readahead=False, + **kwargs): + try: + import lmdb # NOQA + except ImportError: + raise ImportError('Please install lmdb to enable LmdbBackend.') + + self.db_path = str(db_path) + self.readonly = readonly + self.lock = lock + self.readahead = readahead + self.kwargs = kwargs + self._client = None + + def get(self, filepath): + """Get values according to the filepath. + + Args: + filepath (str | obj:`Path`): Here, filepath is the lmdb key. + """ + if self._client is None: + self._client = self._get_client() + + with self._client.begin(write=False) as txn: + value_buf = txn.get(str(filepath).encode('utf-8')) + return value_buf + + def get_text(self, filepath, encoding=None): + raise NotImplementedError + + def _get_client(self): + import lmdb + + return lmdb.open( + self.db_path, + readonly=self.readonly, + lock=self.lock, + readahead=self.readahead, + **self.kwargs) + + def __del__(self): + self._client.close() + + +class HardDiskBackend(BaseStorageBackend): + """Raw hard disks storage backend.""" + + _allow_symlink = True + + def get(self, filepath: Union[str, Path]) -> bytes: + """Read data from a given ``filepath`` with 'rb' mode. + + Args: + filepath (str or Path): Path to read data. + + Returns: + bytes: Expected bytes object. + """ + with open(filepath, 'rb') as f: + value_buf = f.read() + return value_buf + + def get_text(self, + filepath: Union[str, Path], + encoding: str = 'utf-8') -> str: + """Read data from a given ``filepath`` with 'r' mode. + + Args: + filepath (str or Path): Path to read data. + encoding (str): The encoding format used to open the ``filepath``. + Default: 'utf-8'. + + Returns: + str: Expected text reading from ``filepath``. + """ + with open(filepath, 'r', encoding=encoding) as f: + value_buf = f.read() + return value_buf + + def put(self, obj: bytes, filepath: Union[str, Path]) -> None: + """Write data to a given ``filepath`` with 'wb' mode. + + Note: + ``put`` will create a directory if the directory of ``filepath`` + does not exist. + + Args: + obj (bytes): Data to be written. + filepath (str or Path): Path to write data. + """ + mmcv.mkdir_or_exist(osp.dirname(filepath)) + with open(filepath, 'wb') as f: + f.write(obj) + + def put_text(self, + obj: str, + filepath: Union[str, Path], + encoding: str = 'utf-8') -> None: + """Write data to a given ``filepath`` with 'w' mode. + + Note: + ``put_text`` will create a directory if the directory of + ``filepath`` does not exist. + + Args: + obj (str): Data to be written. + filepath (str or Path): Path to write data. + encoding (str): The encoding format used to open the ``filepath``. + Default: 'utf-8'. + """ + mmcv.mkdir_or_exist(osp.dirname(filepath)) + with open(filepath, 'w', encoding=encoding) as f: + f.write(obj) + + def remove(self, filepath: Union[str, Path]) -> None: + """Remove a file. + + Args: + filepath (str or Path): Path to be removed. + """ + os.remove(filepath) + + def exists(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path exists. + + Args: + filepath (str or Path): Path to be checked whether exists. + + Returns: + bool: Return ``True`` if ``filepath`` exists, ``False`` otherwise. + """ + return osp.exists(filepath) + + def isdir(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path is a directory. + + Args: + filepath (str or Path): Path to be checked whether it is a + directory. + + Returns: + bool: Return ``True`` if ``filepath`` points to a directory, + ``False`` otherwise. + """ + return osp.isdir(filepath) + + def isfile(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path is a file. + + Args: + filepath (str or Path): Path to be checked whether it is a file. + + Returns: + bool: Return ``True`` if ``filepath`` points to a file, ``False`` + otherwise. + """ + return osp.isfile(filepath) + + def join_path(self, filepath: Union[str, Path], + *filepaths: Union[str, Path]) -> str: + """Concatenate all file paths. + + Join one or more filepath components intelligently. The return value + is the concatenation of filepath and any members of *filepaths. + + Args: + filepath (str or Path): Path to be concatenated. + + Returns: + str: The result of concatenation. + """ + return osp.join(filepath, *filepaths) + + @contextmanager + def get_local_path( + self, filepath: Union[str, Path]) -> Iterable[Union[str, Path]]: + """Only for unified API and do nothing.""" + yield filepath + + def list_dir_or_file(self, + dir_path: Union[str, Path], + list_dir: bool = True, + list_file: bool = True, + suffix: Optional[Union[str, Tuple[str]]] = None, + recursive: bool = False) -> Iterator[str]: + """Scan a directory to find the interested directories or files in + arbitrary order. + + Note: + :meth:`list_dir_or_file` returns the path relative to ``dir_path``. + + Args: + dir_path (str | Path): Path of the directory. + list_dir (bool): List the directories. Default: True. + list_file (bool): List the path of files. Default: True. + suffix (str or tuple[str], optional): File suffix + that we are interested in. Default: None. + recursive (bool): If set to True, recursively scan the + directory. Default: False. + + Yields: + Iterable[str]: A relative path to ``dir_path``. + """ + if list_dir and suffix is not None: + raise TypeError('`suffix` should be None when `list_dir` is True') + + if (suffix is not None) and not isinstance(suffix, (str, tuple)): + raise TypeError('`suffix` must be a string or tuple of strings') + + root = dir_path + + def _list_dir_or_file(dir_path, list_dir, list_file, suffix, + recursive): + for entry in os.scandir(dir_path): + if not entry.name.startswith('.') and entry.is_file(): + rel_path = osp.relpath(entry.path, root) + if (suffix is None + or rel_path.endswith(suffix)) and list_file: + yield rel_path + elif osp.isdir(entry.path): + if list_dir: + rel_dir = osp.relpath(entry.path, root) + yield rel_dir + if recursive: + yield from _list_dir_or_file(entry.path, list_dir, + list_file, suffix, + recursive) + + return _list_dir_or_file(dir_path, list_dir, list_file, suffix, + recursive) + + +class HTTPBackend(BaseStorageBackend): + """HTTP and HTTPS storage bachend.""" + + def get(self, filepath): + value_buf = urlopen(filepath).read() + return value_buf + + def get_text(self, filepath, encoding='utf-8'): + value_buf = urlopen(filepath).read() + return value_buf.decode(encoding) + + @contextmanager + def get_local_path(self, filepath: str) -> Iterable[str]: + """Download a file from ``filepath``. + + ``get_local_path`` is decorated by :meth:`contxtlib.contextmanager`. It + can be called with ``with`` statement, and when exists from the + ``with`` statement, the temporary path will be released. + + Args: + filepath (str): Download a file from ``filepath``. + + Examples: + >>> client = HTTPBackend() + >>> # After existing from the ``with`` clause, + >>> # the path will be removed + >>> with client.get_local_path('http://path/of/your/file') as path: + ... # do something here + """ + try: + f = tempfile.NamedTemporaryFile(delete=False) + f.write(self.get(filepath)) + f.close() + yield f.name + finally: + os.remove(f.name) + + +class FileClient: + """A general file client to access files in different backends. + + The client loads a file or text in a specified backend from its path + and returns it as a binary or text file. There are two ways to choose a + backend, the name of backend and the prefix of path. Although both of them + can be used to choose a storage backend, ``backend`` has a higher priority + that is if they are all set, the storage backend will be chosen by the + backend argument. If they are all `None`, the disk backend will be chosen. + Note that It can also register other backend accessor with a given name, + prefixes, and backend class. In addition, We use the singleton pattern to + avoid repeated object creation. If the arguments are the same, the same + object will be returned. + + Args: + backend (str, optional): The storage backend type. Options are "disk", + "ceph", "memcached", "lmdb", "http" and "petrel". Default: None. + prefix (str, optional): The prefix of the registered storage backend. + Options are "s3", "http", "https". Default: None. + + Examples: + >>> # only set backend + >>> file_client = FileClient(backend='petrel') + >>> # only set prefix + >>> file_client = FileClient(prefix='s3') + >>> # set both backend and prefix but use backend to choose client + >>> file_client = FileClient(backend='petrel', prefix='s3') + >>> # if the arguments are the same, the same object is returned + >>> file_client1 = FileClient(backend='petrel') + >>> file_client1 is file_client + True + + Attributes: + client (:obj:`BaseStorageBackend`): The backend object. + """ + + _backends = { + 'disk': HardDiskBackend, + 'ceph': CephBackend, + 'memcached': MemcachedBackend, + 'lmdb': LmdbBackend, + 'petrel': PetrelBackend, + 'http': HTTPBackend, + } + # This collection is used to record the overridden backends, and when a + # backend appears in the collection, the singleton pattern is disabled for + # that backend, because if the singleton pattern is used, then the object + # returned will be the backend before overwriting + _overridden_backends = set() + _prefix_to_backends = { + 's3': PetrelBackend, + 'http': HTTPBackend, + 'https': HTTPBackend, + } + _overridden_prefixes = set() + + _instances = {} + + def __new__(cls, backend=None, prefix=None, **kwargs): + if backend is None and prefix is None: + backend = 'disk' + if backend is not None and backend not in cls._backends: + raise ValueError( + f'Backend {backend} is not supported. Currently supported ones' + f' are {list(cls._backends.keys())}') + if prefix is not None and prefix not in cls._prefix_to_backends: + raise ValueError( + f'prefix {prefix} is not supported. Currently supported ones ' + f'are {list(cls._prefix_to_backends.keys())}') + + # concatenate the arguments to a unique key for determining whether + # objects with the same arguments were created + arg_key = f'{backend}:{prefix}' + for key, value in kwargs.items(): + arg_key += f':{key}:{value}' + + # if a backend was overridden, it will create a new object + if (arg_key in cls._instances + and backend not in cls._overridden_backends + and prefix not in cls._overridden_prefixes): + _instance = cls._instances[arg_key] + else: + # create a new object and put it to _instance + _instance = super().__new__(cls) + if backend is not None: + _instance.client = cls._backends[backend](**kwargs) + else: + _instance.client = cls._prefix_to_backends[prefix](**kwargs) + + cls._instances[arg_key] = _instance + + return _instance + + @property + def name(self): + return self.client.name + + @property + def allow_symlink(self): + return self.client.allow_symlink + + @staticmethod + def parse_uri_prefix(uri: Union[str, Path]) -> Optional[str]: + """Parse the prefix of a uri. + + Args: + uri (str | Path): Uri to be parsed that contains the file prefix. + + Examples: + >>> FileClient.parse_uri_prefix('s3://path/of/your/file') + 's3' + + Returns: + str | None: Return the prefix of uri if the uri contains '://' else + ``None``. + """ + assert is_filepath(uri) + uri = str(uri) + if '://' not in uri: + return None + else: + prefix, _ = uri.split('://') + # In the case of PetrelBackend, the prefix may contains the cluster + # name like clusterName:s3 + if ':' in prefix: + _, prefix = prefix.split(':') + return prefix + + @classmethod + def infer_client(cls, + file_client_args: Optional[dict] = None, + uri: Optional[Union[str, Path]] = None) -> 'FileClient': + """Infer a suitable file client based on the URI and arguments. + + Args: + file_client_args (dict, optional): Arguments to instantiate a + FileClient. Default: None. + uri (str | Path, optional): Uri to be parsed that contains the file + prefix. Default: None. + + Examples: + >>> uri = 's3://path/of/your/file' + >>> file_client = FileClient.infer_client(uri=uri) + >>> file_client_args = {'backend': 'petrel'} + >>> file_client = FileClient.infer_client(file_client_args) + + Returns: + FileClient: Instantiated FileClient object. + """ + assert file_client_args is not None or uri is not None + if file_client_args is None: + file_prefix = cls.parse_uri_prefix(uri) # type: ignore + return cls(prefix=file_prefix) + else: + return cls(**file_client_args) + + @classmethod + def _register_backend(cls, name, backend, force=False, prefixes=None): + if not isinstance(name, str): + raise TypeError('the backend name should be a string, ' + f'but got {type(name)}') + if not inspect.isclass(backend): + raise TypeError( + f'backend should be a class but got {type(backend)}') + if not issubclass(backend, BaseStorageBackend): + raise TypeError( + f'backend {backend} is not a subclass of BaseStorageBackend') + if not force and name in cls._backends: + raise KeyError( + f'{name} is already registered as a storage backend, ' + 'add "force=True" if you want to override it') + + if name in cls._backends and force: + cls._overridden_backends.add(name) + cls._backends[name] = backend + + if prefixes is not None: + if isinstance(prefixes, str): + prefixes = [prefixes] + else: + assert isinstance(prefixes, (list, tuple)) + for prefix in prefixes: + if prefix not in cls._prefix_to_backends: + cls._prefix_to_backends[prefix] = backend + elif (prefix in cls._prefix_to_backends) and force: + cls._overridden_prefixes.add(prefix) + cls._prefix_to_backends[prefix] = backend + else: + raise KeyError( + f'{prefix} is already registered as a storage backend,' + ' add "force=True" if you want to override it') + + @classmethod + def register_backend(cls, name, backend=None, force=False, prefixes=None): + """Register a backend to FileClient. + + This method can be used as a normal class method or a decorator. + + .. code-block:: python + + class NewBackend(BaseStorageBackend): + + def get(self, filepath): + return filepath + + def get_text(self, filepath): + return filepath + + FileClient.register_backend('new', NewBackend) + + or + + .. code-block:: python + + @FileClient.register_backend('new') + class NewBackend(BaseStorageBackend): + + def get(self, filepath): + return filepath + + def get_text(self, filepath): + return filepath + + Args: + name (str): The name of the registered backend. + backend (class, optional): The backend class to be registered, + which must be a subclass of :class:`BaseStorageBackend`. + When this method is used as a decorator, backend is None. + Defaults to None. + force (bool, optional): Whether to override the backend if the name + has already been registered. Defaults to False. + prefixes (str or list[str] or tuple[str], optional): The prefixes + of the registered storage backend. Default: None. + `New in version 1.3.15.` + """ + if backend is not None: + cls._register_backend( + name, backend, force=force, prefixes=prefixes) + return + + def _register(backend_cls): + cls._register_backend( + name, backend_cls, force=force, prefixes=prefixes) + return backend_cls + + return _register + + def get(self, filepath: Union[str, Path]) -> Union[bytes, memoryview]: + """Read data from a given ``filepath`` with 'rb' mode. + + Note: + There are two types of return values for ``get``, one is ``bytes`` + and the other is ``memoryview``. The advantage of using memoryview + is that you can avoid copying, and if you want to convert it to + ``bytes``, you can use ``.tobytes()``. + + Args: + filepath (str or Path): Path to read data. + + Returns: + bytes | memoryview: Expected bytes object or a memory view of the + bytes object. + """ + return self.client.get(filepath) + + def get_text(self, filepath: Union[str, Path], encoding='utf-8') -> str: + """Read data from a given ``filepath`` with 'r' mode. + + Args: + filepath (str or Path): Path to read data. + encoding (str): The encoding format used to open the ``filepath``. + Default: 'utf-8'. + + Returns: + str: Expected text reading from ``filepath``. + """ + return self.client.get_text(filepath, encoding) + + def put(self, obj: bytes, filepath: Union[str, Path]) -> None: + """Write data to a given ``filepath`` with 'wb' mode. + + Note: + ``put`` should create a directory if the directory of ``filepath`` + does not exist. + + Args: + obj (bytes): Data to be written. + filepath (str or Path): Path to write data. + """ + self.client.put(obj, filepath) + + def put_text(self, obj: str, filepath: Union[str, Path]) -> None: + """Write data to a given ``filepath`` with 'w' mode. + + Note: + ``put_text`` should create a directory if the directory of + ``filepath`` does not exist. + + Args: + obj (str): Data to be written. + filepath (str or Path): Path to write data. + encoding (str, optional): The encoding format used to open the + `filepath`. Default: 'utf-8'. + """ + self.client.put_text(obj, filepath) + + def remove(self, filepath: Union[str, Path]) -> None: + """Remove a file. + + Args: + filepath (str, Path): Path to be removed. + """ + self.client.remove(filepath) + + def exists(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path exists. + + Args: + filepath (str or Path): Path to be checked whether exists. + + Returns: + bool: Return ``True`` if ``filepath`` exists, ``False`` otherwise. + """ + return self.client.exists(filepath) + + def isdir(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path is a directory. + + Args: + filepath (str or Path): Path to be checked whether it is a + directory. + + Returns: + bool: Return ``True`` if ``filepath`` points to a directory, + ``False`` otherwise. + """ + return self.client.isdir(filepath) + + def isfile(self, filepath: Union[str, Path]) -> bool: + """Check whether a file path is a file. + + Args: + filepath (str or Path): Path to be checked whether it is a file. + + Returns: + bool: Return ``True`` if ``filepath`` points to a file, ``False`` + otherwise. + """ + return self.client.isfile(filepath) + + def join_path(self, filepath: Union[str, Path], + *filepaths: Union[str, Path]) -> str: + """Concatenate all file paths. + + Join one or more filepath components intelligently. The return value + is the concatenation of filepath and any members of *filepaths. + + Args: + filepath (str or Path): Path to be concatenated. + + Returns: + str: The result of concatenation. + """ + return self.client.join_path(filepath, *filepaths) + + @contextmanager + def get_local_path(self, filepath: Union[str, Path]) -> Iterable[str]: + """Download data from ``filepath`` and write the data to local path. + + ``get_local_path`` is decorated by :meth:`contxtlib.contextmanager`. It + can be called with ``with`` statement, and when exists from the + ``with`` statement, the temporary path will be released. + + Note: + If the ``filepath`` is a local path, just return itself. + + .. warning:: + ``get_local_path`` is an experimental interface that may change in + the future. + + Args: + filepath (str or Path): Path to be read data. + + Examples: + >>> file_client = FileClient(prefix='s3') + >>> with file_client.get_local_path('s3://bucket/abc.jpg') as path: + ... # do something here + + Yields: + Iterable[str]: Only yield one path. + """ + with self.client.get_local_path(str(filepath)) as local_path: + yield local_path + + def list_dir_or_file(self, + dir_path: Union[str, Path], + list_dir: bool = True, + list_file: bool = True, + suffix: Optional[Union[str, Tuple[str]]] = None, + recursive: bool = False) -> Iterator[str]: + """Scan a directory to find the interested directories or files in + arbitrary order. + + Note: + :meth:`list_dir_or_file` returns the path relative to ``dir_path``. + + Args: + dir_path (str | Path): Path of the directory. + list_dir (bool): List the directories. Default: True. + list_file (bool): List the path of files. Default: True. + suffix (str or tuple[str], optional): File suffix + that we are interested in. Default: None. + recursive (bool): If set to True, recursively scan the + directory. Default: False. + + Yields: + Iterable[str]: A relative path to ``dir_path``. + """ + yield from self.client.list_dir_or_file(dir_path, list_dir, list_file, + suffix, recursive) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..aa24d91972837b8756b225f4879bac20436eb72a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/__init__.py @@ -0,0 +1,7 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .base import BaseFileHandler +from .json_handler import JsonHandler +from .pickle_handler import PickleHandler +from .yaml_handler import YamlHandler + +__all__ = ['BaseFileHandler', 'JsonHandler', 'PickleHandler', 'YamlHandler'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/base.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/base.py new file mode 100644 index 0000000000000000000000000000000000000000..288878bc57282fbb2f12b32290152ca8e9d3cab0 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/base.py @@ -0,0 +1,30 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from abc import ABCMeta, abstractmethod + + +class BaseFileHandler(metaclass=ABCMeta): + # `str_like` is a flag to indicate whether the type of file object is + # str-like object or bytes-like object. Pickle only processes bytes-like + # objects but json only processes str-like object. If it is str-like + # object, `StringIO` will be used to process the buffer. + str_like = True + + @abstractmethod + def load_from_fileobj(self, file, **kwargs): + pass + + @abstractmethod + def dump_to_fileobj(self, obj, file, **kwargs): + pass + + @abstractmethod + def dump_to_str(self, obj, **kwargs): + pass + + def load_from_path(self, filepath, mode='r', **kwargs): + with open(filepath, mode) as f: + return self.load_from_fileobj(f, **kwargs) + + def dump_to_path(self, obj, filepath, mode='w', **kwargs): + with open(filepath, mode) as f: + self.dump_to_fileobj(obj, f, **kwargs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/json_handler.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/json_handler.py new file mode 100644 index 0000000000000000000000000000000000000000..18d4f15f74139d20adff18b20be5529c592a66b6 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/json_handler.py @@ -0,0 +1,36 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import json + +import numpy as np + +from .base import BaseFileHandler + + +def set_default(obj): + """Set default json values for non-serializable values. + + It helps convert ``set``, ``range`` and ``np.ndarray`` data types to list. + It also converts ``np.generic`` (including ``np.int32``, ``np.float32``, + etc.) into plain numbers of plain python built-in types. + """ + if isinstance(obj, (set, range)): + return list(obj) + elif isinstance(obj, np.ndarray): + return obj.tolist() + elif isinstance(obj, np.generic): + return obj.item() + raise TypeError(f'{type(obj)} is unsupported for json dump') + + +class JsonHandler(BaseFileHandler): + + def load_from_fileobj(self, file): + return json.load(file) + + def dump_to_fileobj(self, obj, file, **kwargs): + kwargs.setdefault('default', set_default) + json.dump(obj, file, **kwargs) + + def dump_to_str(self, obj, **kwargs): + kwargs.setdefault('default', set_default) + return json.dumps(obj, **kwargs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/pickle_handler.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/pickle_handler.py new file mode 100644 index 0000000000000000000000000000000000000000..b37c79bed4ef9fd8913715e62dbe3fc5cafdc3aa --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/pickle_handler.py @@ -0,0 +1,28 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import pickle + +from .base import BaseFileHandler + + +class PickleHandler(BaseFileHandler): + + str_like = False + + def load_from_fileobj(self, file, **kwargs): + return pickle.load(file, **kwargs) + + def load_from_path(self, filepath, **kwargs): + return super(PickleHandler, self).load_from_path( + filepath, mode='rb', **kwargs) + + def dump_to_str(self, obj, **kwargs): + kwargs.setdefault('protocol', 2) + return pickle.dumps(obj, **kwargs) + + def dump_to_fileobj(self, obj, file, **kwargs): + kwargs.setdefault('protocol', 2) + pickle.dump(obj, file, **kwargs) + + def dump_to_path(self, obj, filepath, **kwargs): + super(PickleHandler, self).dump_to_path( + obj, filepath, mode='wb', **kwargs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/yaml_handler.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/yaml_handler.py new file mode 100644 index 0000000000000000000000000000000000000000..60911e7e605a4d0f770c040ca726f7cbac9727de --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/handlers/yaml_handler.py @@ -0,0 +1,25 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import yaml + +try: + from yaml import CDumper as Dumper + from yaml import CLoader as Loader +except ImportError: + from yaml import Loader, Dumper + +from .base import BaseFileHandler # isort:skip + + +class YamlHandler(BaseFileHandler): + + def load_from_fileobj(self, file, **kwargs): + kwargs.setdefault('Loader', Loader) + return yaml.load(file, **kwargs) + + def dump_to_fileobj(self, obj, file, **kwargs): + kwargs.setdefault('Dumper', Dumper) + yaml.dump(obj, file, **kwargs) + + def dump_to_str(self, obj, **kwargs): + kwargs.setdefault('Dumper', Dumper) + return yaml.dump(obj, **kwargs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/io.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/io.py new file mode 100644 index 0000000000000000000000000000000000000000..aaefde58aa3ea5b58f86249ce7e1c40c186eb8dd --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/io.py @@ -0,0 +1,151 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from io import BytesIO, StringIO +from pathlib import Path + +from ..utils import is_list_of, is_str +from .file_client import FileClient +from .handlers import BaseFileHandler, JsonHandler, PickleHandler, YamlHandler + +file_handlers = { + 'json': JsonHandler(), + 'yaml': YamlHandler(), + 'yml': YamlHandler(), + 'pickle': PickleHandler(), + 'pkl': PickleHandler() +} + + +def load(file, file_format=None, file_client_args=None, **kwargs): + """Load data from json/yaml/pickle files. + + This method provides a unified api for loading data from serialized files. + + Note: + In v1.3.16 and later, ``load`` supports loading data from serialized + files those can be storaged in different backends. + + Args: + file (str or :obj:`Path` or file-like object): Filename or a file-like + object. + file_format (str, optional): If not specified, the file format will be + inferred from the file extension, otherwise use the specified one. + Currently supported formats include "json", "yaml/yml" and + "pickle/pkl". + file_client_args (dict, optional): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + + Examples: + >>> load('/path/of/your/file') # file is storaged in disk + >>> load('https://path/of/your/file') # file is storaged in Internet + >>> load('s3://path/of/your/file') # file is storaged in petrel + + Returns: + The content from the file. + """ + if isinstance(file, Path): + file = str(file) + if file_format is None and is_str(file): + file_format = file.split('.')[-1] + if file_format not in file_handlers: + raise TypeError(f'Unsupported format: {file_format}') + + handler = file_handlers[file_format] + if is_str(file): + file_client = FileClient.infer_client(file_client_args, file) + if handler.str_like: + with StringIO(file_client.get_text(file)) as f: + obj = handler.load_from_fileobj(f, **kwargs) + else: + with BytesIO(file_client.get(file)) as f: + obj = handler.load_from_fileobj(f, **kwargs) + elif hasattr(file, 'read'): + obj = handler.load_from_fileobj(file, **kwargs) + else: + raise TypeError('"file" must be a filepath str or a file-object') + return obj + + +def dump(obj, file=None, file_format=None, file_client_args=None, **kwargs): + """Dump data to json/yaml/pickle strings or files. + + This method provides a unified api for dumping data as strings or to files, + and also supports custom arguments for each file format. + + Note: + In v1.3.16 and later, ``dump`` supports dumping data as strings or to + files which is saved to different backends. + + Args: + obj (any): The python object to be dumped. + file (str or :obj:`Path` or file-like object, optional): If not + specified, then the object is dumped to a str, otherwise to a file + specified by the filename or file-like object. + file_format (str, optional): Same as :func:`load`. + file_client_args (dict, optional): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + + Examples: + >>> dump('hello world', '/path/of/your/file') # disk + >>> dump('hello world', 's3://path/of/your/file') # ceph or petrel + + Returns: + bool: True for success, False otherwise. + """ + if isinstance(file, Path): + file = str(file) + if file_format is None: + if is_str(file): + file_format = file.split('.')[-1] + elif file is None: + raise ValueError( + 'file_format must be specified since file is None') + if file_format not in file_handlers: + raise TypeError(f'Unsupported format: {file_format}') + + handler = file_handlers[file_format] + if file is None: + return handler.dump_to_str(obj, **kwargs) + elif is_str(file): + file_client = FileClient.infer_client(file_client_args, file) + if handler.str_like: + with StringIO() as f: + handler.dump_to_fileobj(obj, f, **kwargs) + file_client.put_text(f.getvalue(), file) + else: + with BytesIO() as f: + handler.dump_to_fileobj(obj, f, **kwargs) + file_client.put(f.getvalue(), file) + elif hasattr(file, 'write'): + handler.dump_to_fileobj(obj, file, **kwargs) + else: + raise TypeError('"file" must be a filename str or a file-object') + + +def _register_handler(handler, file_formats): + """Register a handler for some file extensions. + + Args: + handler (:obj:`BaseFileHandler`): Handler to be registered. + file_formats (str or list[str]): File formats to be handled by this + handler. + """ + if not isinstance(handler, BaseFileHandler): + raise TypeError( + f'handler must be a child of BaseFileHandler, not {type(handler)}') + if isinstance(file_formats, str): + file_formats = [file_formats] + if not is_list_of(file_formats, str): + raise TypeError('file_formats must be a str or a list of str') + for ext in file_formats: + file_handlers[ext] = handler + + +def register_handler(file_formats, **kwargs): + + def wrap(cls): + _register_handler(cls(**kwargs), file_formats) + return cls + + return wrap diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/parse.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/parse.py new file mode 100644 index 0000000000000000000000000000000000000000..f60f0d611b8d75692221d0edd7dc993b0a6445c9 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/fileio/parse.py @@ -0,0 +1,97 @@ +# Copyright (c) OpenMMLab. All rights reserved. + +from io import StringIO + +from .file_client import FileClient + + +def list_from_file(filename, + prefix='', + offset=0, + max_num=0, + encoding='utf-8', + file_client_args=None): + """Load a text file and parse the content as a list of strings. + + Note: + In v1.3.16 and later, ``list_from_file`` supports loading a text file + which can be storaged in different backends and parsing the content as + a list for strings. + + Args: + filename (str): Filename. + prefix (str): The prefix to be inserted to the beginning of each item. + offset (int): The offset of lines. + max_num (int): The maximum number of lines to be read, + zeros and negatives mean no limitation. + encoding (str): Encoding used to open the file. Default utf-8. + file_client_args (dict, optional): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + + Examples: + >>> list_from_file('/path/of/your/file') # disk + ['hello', 'world'] + >>> list_from_file('s3://path/of/your/file') # ceph or petrel + ['hello', 'world'] + + Returns: + list[str]: A list of strings. + """ + cnt = 0 + item_list = [] + file_client = FileClient.infer_client(file_client_args, filename) + with StringIO(file_client.get_text(filename, encoding)) as f: + for _ in range(offset): + f.readline() + for line in f: + if 0 < max_num <= cnt: + break + item_list.append(prefix + line.rstrip('\n\r')) + cnt += 1 + return item_list + + +def dict_from_file(filename, + key_type=str, + encoding='utf-8', + file_client_args=None): + """Load a text file and parse the content as a dict. + + Each line of the text file will be two or more columns split by + whitespaces or tabs. The first column will be parsed as dict keys, and + the following columns will be parsed as dict values. + + Note: + In v1.3.16 and later, ``dict_from_file`` supports loading a text file + which can be storaged in different backends and parsing the content as + a dict. + + Args: + filename(str): Filename. + key_type(type): Type of the dict keys. str is user by default and + type conversion will be performed if specified. + encoding (str): Encoding used to open the file. Default utf-8. + file_client_args (dict, optional): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + + Examples: + >>> dict_from_file('/path/of/your/file') # disk + {'key1': 'value1', 'key2': 'value2'} + >>> dict_from_file('s3://path/of/your/file') # ceph or petrel + {'key1': 'value1', 'key2': 'value2'} + + Returns: + dict: The parsed contents. + """ + mapping = {} + file_client = FileClient.infer_client(file_client_args, filename) + with StringIO(file_client.get_text(filename, encoding)) as f: + for line in f: + items = line.rstrip('\n').split() + assert len(items) >= 2 + key = key_type(items[0]) + val = items[1:] if len(items) > 2 else items[1] + mapping[key] = val + return mapping diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..d0051d609d3de4e7562e3fe638335c66617c4d91 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/__init__.py @@ -0,0 +1,28 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .colorspace import (bgr2gray, bgr2hls, bgr2hsv, bgr2rgb, bgr2ycbcr, + gray2bgr, gray2rgb, hls2bgr, hsv2bgr, imconvert, + rgb2bgr, rgb2gray, rgb2ycbcr, ycbcr2bgr, ycbcr2rgb) +from .geometric import (cutout, imcrop, imflip, imflip_, impad, + impad_to_multiple, imrescale, imresize, imresize_like, + imresize_to_multiple, imrotate, imshear, imtranslate, + rescale_size) +from .io import imfrombytes, imread, imwrite, supported_backends, use_backend +from .misc import tensor2imgs +from .photometric import (adjust_brightness, adjust_color, adjust_contrast, + adjust_lighting, adjust_sharpness, auto_contrast, + clahe, imdenormalize, imequalize, iminvert, + imnormalize, imnormalize_, lut_transform, posterize, + solarize) + +__all__ = [ + 'bgr2gray', 'bgr2hls', 'bgr2hsv', 'bgr2rgb', 'gray2bgr', 'gray2rgb', + 'hls2bgr', 'hsv2bgr', 'imconvert', 'rgb2bgr', 'rgb2gray', 'imrescale', + 'imresize', 'imresize_like', 'imresize_to_multiple', 'rescale_size', + 'imcrop', 'imflip', 'imflip_', 'impad', 'impad_to_multiple', 'imrotate', + 'imfrombytes', 'imread', 'imwrite', 'supported_backends', 'use_backend', + 'imdenormalize', 'imnormalize', 'imnormalize_', 'iminvert', 'posterize', + 'solarize', 'rgb2ycbcr', 'bgr2ycbcr', 'ycbcr2rgb', 'ycbcr2bgr', + 'tensor2imgs', 'imshear', 'imtranslate', 'adjust_color', 'imequalize', + 'adjust_brightness', 'adjust_contrast', 'lut_transform', 'clahe', + 'adjust_sharpness', 'auto_contrast', 'cutout', 'adjust_lighting' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/colorspace.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/colorspace.py new file mode 100644 index 0000000000000000000000000000000000000000..4337720ead3417b328319aebbfcc9c0c43059996 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/colorspace.py @@ -0,0 +1,306 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import cv2 +import numpy as np + + +def imconvert(img, src, dst): + """Convert an image from the src colorspace to dst colorspace. + + Args: + img (ndarray): The input image. + src (str): The source colorspace, e.g., 'rgb', 'hsv'. + dst (str): The destination colorspace, e.g., 'rgb', 'hsv'. + + Returns: + ndarray: The converted image. + """ + code = getattr(cv2, f'COLOR_{src.upper()}2{dst.upper()}') + out_img = cv2.cvtColor(img, code) + return out_img + + +def bgr2gray(img, keepdim=False): + """Convert a BGR image to grayscale image. + + Args: + img (ndarray): The input image. + keepdim (bool): If False (by default), then return the grayscale image + with 2 dims, otherwise 3 dims. + + Returns: + ndarray: The converted grayscale image. + """ + out_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY) + if keepdim: + out_img = out_img[..., None] + return out_img + + +def rgb2gray(img, keepdim=False): + """Convert a RGB image to grayscale image. + + Args: + img (ndarray): The input image. + keepdim (bool): If False (by default), then return the grayscale image + with 2 dims, otherwise 3 dims. + + Returns: + ndarray: The converted grayscale image. + """ + out_img = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY) + if keepdim: + out_img = out_img[..., None] + return out_img + + +def gray2bgr(img): + """Convert a grayscale image to BGR image. + + Args: + img (ndarray): The input image. + + Returns: + ndarray: The converted BGR image. + """ + img = img[..., None] if img.ndim == 2 else img + out_img = cv2.cvtColor(img, cv2.COLOR_GRAY2BGR) + return out_img + + +def gray2rgb(img): + """Convert a grayscale image to RGB image. + + Args: + img (ndarray): The input image. + + Returns: + ndarray: The converted RGB image. + """ + img = img[..., None] if img.ndim == 2 else img + out_img = cv2.cvtColor(img, cv2.COLOR_GRAY2RGB) + return out_img + + +def _convert_input_type_range(img): + """Convert the type and range of the input image. + + It converts the input image to np.float32 type and range of [0, 1]. + It is mainly used for pre-processing the input image in colorspace + conversion functions such as rgb2ycbcr and ycbcr2rgb. + + Args: + img (ndarray): The input image. It accepts: + 1. np.uint8 type with range [0, 255]; + 2. np.float32 type with range [0, 1]. + + Returns: + (ndarray): The converted image with type of np.float32 and range of + [0, 1]. + """ + img_type = img.dtype + img = img.astype(np.float32) + if img_type == np.float32: + pass + elif img_type == np.uint8: + img /= 255. + else: + raise TypeError('The img type should be np.float32 or np.uint8, ' + f'but got {img_type}') + return img + + +def _convert_output_type_range(img, dst_type): + """Convert the type and range of the image according to dst_type. + + It converts the image to desired type and range. If `dst_type` is np.uint8, + images will be converted to np.uint8 type with range [0, 255]. If + `dst_type` is np.float32, it converts the image to np.float32 type with + range [0, 1]. + It is mainly used for post-processing images in colorspace conversion + functions such as rgb2ycbcr and ycbcr2rgb. + + Args: + img (ndarray): The image to be converted with np.float32 type and + range [0, 255]. + dst_type (np.uint8 | np.float32): If dst_type is np.uint8, it + converts the image to np.uint8 type with range [0, 255]. If + dst_type is np.float32, it converts the image to np.float32 type + with range [0, 1]. + + Returns: + (ndarray): The converted image with desired type and range. + """ + if dst_type not in (np.uint8, np.float32): + raise TypeError('The dst_type should be np.float32 or np.uint8, ' + f'but got {dst_type}') + if dst_type == np.uint8: + img = img.round() + else: + img /= 255. + return img.astype(dst_type) + + +def rgb2ycbcr(img, y_only=False): + """Convert a RGB image to YCbCr image. + + This function produces the same results as Matlab's `rgb2ycbcr` function. + It implements the ITU-R BT.601 conversion for standard-definition + television. See more details in + https://en.wikipedia.org/wiki/YCbCr#ITU-R_BT.601_conversion. + + It differs from a similar function in cv2.cvtColor: `RGB <-> YCrCb`. + In OpenCV, it implements a JPEG conversion. See more details in + https://en.wikipedia.org/wiki/YCbCr#JPEG_conversion. + + Args: + img (ndarray): The input image. It accepts: + 1. np.uint8 type with range [0, 255]; + 2. np.float32 type with range [0, 1]. + y_only (bool): Whether to only return Y channel. Default: False. + + Returns: + ndarray: The converted YCbCr image. The output image has the same type + and range as input image. + """ + img_type = img.dtype + img = _convert_input_type_range(img) + if y_only: + out_img = np.dot(img, [65.481, 128.553, 24.966]) + 16.0 + else: + out_img = np.matmul( + img, [[65.481, -37.797, 112.0], [128.553, -74.203, -93.786], + [24.966, 112.0, -18.214]]) + [16, 128, 128] + out_img = _convert_output_type_range(out_img, img_type) + return out_img + + +def bgr2ycbcr(img, y_only=False): + """Convert a BGR image to YCbCr image. + + The bgr version of rgb2ycbcr. + It implements the ITU-R BT.601 conversion for standard-definition + television. See more details in + https://en.wikipedia.org/wiki/YCbCr#ITU-R_BT.601_conversion. + + It differs from a similar function in cv2.cvtColor: `BGR <-> YCrCb`. + In OpenCV, it implements a JPEG conversion. See more details in + https://en.wikipedia.org/wiki/YCbCr#JPEG_conversion. + + Args: + img (ndarray): The input image. It accepts: + 1. np.uint8 type with range [0, 255]; + 2. np.float32 type with range [0, 1]. + y_only (bool): Whether to only return Y channel. Default: False. + + Returns: + ndarray: The converted YCbCr image. The output image has the same type + and range as input image. + """ + img_type = img.dtype + img = _convert_input_type_range(img) + if y_only: + out_img = np.dot(img, [24.966, 128.553, 65.481]) + 16.0 + else: + out_img = np.matmul( + img, [[24.966, 112.0, -18.214], [128.553, -74.203, -93.786], + [65.481, -37.797, 112.0]]) + [16, 128, 128] + out_img = _convert_output_type_range(out_img, img_type) + return out_img + + +def ycbcr2rgb(img): + """Convert a YCbCr image to RGB image. + + This function produces the same results as Matlab's ycbcr2rgb function. + It implements the ITU-R BT.601 conversion for standard-definition + television. See more details in + https://en.wikipedia.org/wiki/YCbCr#ITU-R_BT.601_conversion. + + It differs from a similar function in cv2.cvtColor: `YCrCb <-> RGB`. + In OpenCV, it implements a JPEG conversion. See more details in + https://en.wikipedia.org/wiki/YCbCr#JPEG_conversion. + + Args: + img (ndarray): The input image. It accepts: + 1. np.uint8 type with range [0, 255]; + 2. np.float32 type with range [0, 1]. + + Returns: + ndarray: The converted RGB image. The output image has the same type + and range as input image. + """ + img_type = img.dtype + img = _convert_input_type_range(img) * 255 + out_img = np.matmul(img, [[0.00456621, 0.00456621, 0.00456621], + [0, -0.00153632, 0.00791071], + [0.00625893, -0.00318811, 0]]) * 255.0 + [ + -222.921, 135.576, -276.836 + ] + out_img = _convert_output_type_range(out_img, img_type) + return out_img + + +def ycbcr2bgr(img): + """Convert a YCbCr image to BGR image. + + The bgr version of ycbcr2rgb. + It implements the ITU-R BT.601 conversion for standard-definition + television. See more details in + https://en.wikipedia.org/wiki/YCbCr#ITU-R_BT.601_conversion. + + It differs from a similar function in cv2.cvtColor: `YCrCb <-> BGR`. + In OpenCV, it implements a JPEG conversion. See more details in + https://en.wikipedia.org/wiki/YCbCr#JPEG_conversion. + + Args: + img (ndarray): The input image. It accepts: + 1. np.uint8 type with range [0, 255]; + 2. np.float32 type with range [0, 1]. + + Returns: + ndarray: The converted BGR image. The output image has the same type + and range as input image. + """ + img_type = img.dtype + img = _convert_input_type_range(img) * 255 + out_img = np.matmul(img, [[0.00456621, 0.00456621, 0.00456621], + [0.00791071, -0.00153632, 0], + [0, -0.00318811, 0.00625893]]) * 255.0 + [ + -276.836, 135.576, -222.921 + ] + out_img = _convert_output_type_range(out_img, img_type) + return out_img + + +def convert_color_factory(src, dst): + + code = getattr(cv2, f'COLOR_{src.upper()}2{dst.upper()}') + + def convert_color(img): + out_img = cv2.cvtColor(img, code) + return out_img + + convert_color.__doc__ = f"""Convert a {src.upper()} image to {dst.upper()} + image. + + Args: + img (ndarray or str): The input image. + + Returns: + ndarray: The converted {dst.upper()} image. + """ + + return convert_color + + +bgr2rgb = convert_color_factory('bgr', 'rgb') + +rgb2bgr = convert_color_factory('rgb', 'bgr') + +bgr2hsv = convert_color_factory('bgr', 'hsv') + +hsv2bgr = convert_color_factory('hsv', 'bgr') + +bgr2hls = convert_color_factory('bgr', 'hls') + +hls2bgr = convert_color_factory('hls', 'bgr') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/geometric.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/geometric.py new file mode 100644 index 0000000000000000000000000000000000000000..4c423bf2a1df3eb247512d429a53d5a0dd68db35 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/geometric.py @@ -0,0 +1,741 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import numbers + +import cv2 +import numpy as np + +from ..utils import to_2tuple +from .io import imread_backend + +try: + from PIL import Image +except ImportError: + Image = None + + +def _scale_size(size, scale): + """Rescale a size by a ratio. + + Args: + size (tuple[int]): (w, h). + scale (float | tuple(float)): Scaling factor. + + Returns: + tuple[int]: scaled size. + """ + if isinstance(scale, (float, int)): + scale = (scale, scale) + w, h = size + return int(w * float(scale[0]) + 0.5), int(h * float(scale[1]) + 0.5) + + +cv2_interp_codes = { + 'nearest': cv2.INTER_NEAREST, + 'bilinear': cv2.INTER_LINEAR, + 'bicubic': cv2.INTER_CUBIC, + 'area': cv2.INTER_AREA, + 'lanczos': cv2.INTER_LANCZOS4 +} + +# Pillow >=v9.1.0 use a slightly different naming scheme for filters. +# Set pillow_interp_codes according to the naming scheme used. +if Image is not None: + if hasattr(Image, 'Resampling'): + pillow_interp_codes = { + 'nearest': Image.Resampling.NEAREST, + 'bilinear': Image.Resampling.BILINEAR, + 'bicubic': Image.Resampling.BICUBIC, + 'box': Image.Resampling.BOX, + 'lanczos': Image.Resampling.LANCZOS, + 'hamming': Image.Resampling.HAMMING + } + else: + pillow_interp_codes = { + 'nearest': Image.NEAREST, + 'bilinear': Image.BILINEAR, + 'bicubic': Image.BICUBIC, + 'box': Image.BOX, + 'lanczos': Image.LANCZOS, + 'hamming': Image.HAMMING + } + + +def imresize(img, + size, + return_scale=False, + interpolation='bilinear', + out=None, + backend=None): + """Resize image to a given size. + + Args: + img (ndarray): The input image. + size (tuple[int]): Target size (w, h). + return_scale (bool): Whether to return `w_scale` and `h_scale`. + interpolation (str): Interpolation method, accepted values are + "nearest", "bilinear", "bicubic", "area", "lanczos" for 'cv2' + backend, "nearest", "bilinear" for 'pillow' backend. + out (ndarray): The output destination. + backend (str | None): The image resize backend type. Options are `cv2`, + `pillow`, `None`. If backend is None, the global imread_backend + specified by ``mmcv.use_backend()`` will be used. Default: None. + + Returns: + tuple | ndarray: (`resized_img`, `w_scale`, `h_scale`) or + `resized_img`. + """ + h, w = img.shape[:2] + if backend is None: + backend = imread_backend + if backend not in ['cv2', 'pillow']: + raise ValueError(f'backend: {backend} is not supported for resize.' + f"Supported backends are 'cv2', 'pillow'") + + if backend == 'pillow': + assert img.dtype == np.uint8, 'Pillow backend only support uint8 type' + pil_image = Image.fromarray(img) + pil_image = pil_image.resize(size, pillow_interp_codes[interpolation]) + resized_img = np.array(pil_image) + else: + resized_img = cv2.resize( + img, size, dst=out, interpolation=cv2_interp_codes[interpolation]) + if not return_scale: + return resized_img + else: + w_scale = size[0] / w + h_scale = size[1] / h + return resized_img, w_scale, h_scale + + +def imresize_to_multiple(img, + divisor, + size=None, + scale_factor=None, + keep_ratio=False, + return_scale=False, + interpolation='bilinear', + out=None, + backend=None): + """Resize image according to a given size or scale factor and then rounds + up the the resized or rescaled image size to the nearest value that can be + divided by the divisor. + + Args: + img (ndarray): The input image. + divisor (int | tuple): Resized image size will be a multiple of + divisor. If divisor is a tuple, divisor should be + (w_divisor, h_divisor). + size (None | int | tuple[int]): Target size (w, h). Default: None. + scale_factor (None | float | tuple[float]): Multiplier for spatial + size. Should match input size if it is a tuple and the 2D style is + (w_scale_factor, h_scale_factor). Default: None. + keep_ratio (bool): Whether to keep the aspect ratio when resizing the + image. Default: False. + return_scale (bool): Whether to return `w_scale` and `h_scale`. + interpolation (str): Interpolation method, accepted values are + "nearest", "bilinear", "bicubic", "area", "lanczos" for 'cv2' + backend, "nearest", "bilinear" for 'pillow' backend. + out (ndarray): The output destination. + backend (str | None): The image resize backend type. Options are `cv2`, + `pillow`, `None`. If backend is None, the global imread_backend + specified by ``mmcv.use_backend()`` will be used. Default: None. + + Returns: + tuple | ndarray: (`resized_img`, `w_scale`, `h_scale`) or + `resized_img`. + """ + h, w = img.shape[:2] + if size is not None and scale_factor is not None: + raise ValueError('only one of size or scale_factor should be defined') + elif size is None and scale_factor is None: + raise ValueError('one of size or scale_factor should be defined') + elif size is not None: + size = to_2tuple(size) + if keep_ratio: + size = rescale_size((w, h), size, return_scale=False) + else: + size = _scale_size((w, h), scale_factor) + + divisor = to_2tuple(divisor) + size = tuple([int(np.ceil(s / d)) * d for s, d in zip(size, divisor)]) + resized_img, w_scale, h_scale = imresize( + img, + size, + return_scale=True, + interpolation=interpolation, + out=out, + backend=backend) + if return_scale: + return resized_img, w_scale, h_scale + else: + return resized_img + + +def imresize_like(img, + dst_img, + return_scale=False, + interpolation='bilinear', + backend=None): + """Resize image to the same size of a given image. + + Args: + img (ndarray): The input image. + dst_img (ndarray): The target image. + return_scale (bool): Whether to return `w_scale` and `h_scale`. + interpolation (str): Same as :func:`resize`. + backend (str | None): Same as :func:`resize`. + + Returns: + tuple or ndarray: (`resized_img`, `w_scale`, `h_scale`) or + `resized_img`. + """ + h, w = dst_img.shape[:2] + return imresize(img, (w, h), return_scale, interpolation, backend=backend) + + +def rescale_size(old_size, scale, return_scale=False): + """Calculate the new size to be rescaled to. + + Args: + old_size (tuple[int]): The old size (w, h) of image. + scale (float | tuple[int]): The scaling factor or maximum size. + If it is a float number, then the image will be rescaled by this + factor, else if it is a tuple of 2 integers, then the image will + be rescaled as large as possible within the scale. + return_scale (bool): Whether to return the scaling factor besides the + rescaled image size. + + Returns: + tuple[int]: The new rescaled image size. + """ + w, h = old_size + if isinstance(scale, (float, int)): + if scale <= 0: + raise ValueError(f'Invalid scale {scale}, must be positive.') + scale_factor = scale + elif isinstance(scale, tuple): + max_long_edge = max(scale) + max_short_edge = min(scale) + scale_factor = min(max_long_edge / max(h, w), + max_short_edge / min(h, w)) + else: + raise TypeError( + f'Scale must be a number or tuple of int, but got {type(scale)}') + + new_size = _scale_size((w, h), scale_factor) + + if return_scale: + return new_size, scale_factor + else: + return new_size + + +def imrescale(img, + scale, + return_scale=False, + interpolation='bilinear', + backend=None): + """Resize image while keeping the aspect ratio. + + Args: + img (ndarray): The input image. + scale (float | tuple[int]): The scaling factor or maximum size. + If it is a float number, then the image will be rescaled by this + factor, else if it is a tuple of 2 integers, then the image will + be rescaled as large as possible within the scale. + return_scale (bool): Whether to return the scaling factor besides the + rescaled image. + interpolation (str): Same as :func:`resize`. + backend (str | None): Same as :func:`resize`. + + Returns: + ndarray: The rescaled image. + """ + h, w = img.shape[:2] + new_size, scale_factor = rescale_size((w, h), scale, return_scale=True) + rescaled_img = imresize( + img, new_size, interpolation=interpolation, backend=backend) + if return_scale: + return rescaled_img, scale_factor + else: + return rescaled_img + + +def imflip(img, direction='horizontal'): + """Flip an image horizontally or vertically. + + Args: + img (ndarray): Image to be flipped. + direction (str): The flip direction, either "horizontal" or + "vertical" or "diagonal". + + Returns: + ndarray: The flipped image. + """ + assert direction in ['horizontal', 'vertical', 'diagonal'] + if direction == 'horizontal': + return np.flip(img, axis=1) + elif direction == 'vertical': + return np.flip(img, axis=0) + else: + return np.flip(img, axis=(0, 1)) + + +def imflip_(img, direction='horizontal'): + """Inplace flip an image horizontally or vertically. + + Args: + img (ndarray): Image to be flipped. + direction (str): The flip direction, either "horizontal" or + "vertical" or "diagonal". + + Returns: + ndarray: The flipped image (inplace). + """ + assert direction in ['horizontal', 'vertical', 'diagonal'] + if direction == 'horizontal': + return cv2.flip(img, 1, img) + elif direction == 'vertical': + return cv2.flip(img, 0, img) + else: + return cv2.flip(img, -1, img) + + +def imrotate(img, + angle, + center=None, + scale=1.0, + border_value=0, + interpolation='bilinear', + auto_bound=False): + """Rotate an image. + + Args: + img (ndarray): Image to be rotated. + angle (float): Rotation angle in degrees, positive values mean + clockwise rotation. + center (tuple[float], optional): Center point (w, h) of the rotation in + the source image. If not specified, the center of the image will be + used. + scale (float): Isotropic scale factor. + border_value (int): Border value. + interpolation (str): Same as :func:`resize`. + auto_bound (bool): Whether to adjust the image size to cover the whole + rotated image. + + Returns: + ndarray: The rotated image. + """ + if center is not None and auto_bound: + raise ValueError('`auto_bound` conflicts with `center`') + h, w = img.shape[:2] + if center is None: + center = ((w - 1) * 0.5, (h - 1) * 0.5) + assert isinstance(center, tuple) + + matrix = cv2.getRotationMatrix2D(center, -angle, scale) + if auto_bound: + cos = np.abs(matrix[0, 0]) + sin = np.abs(matrix[0, 1]) + new_w = h * sin + w * cos + new_h = h * cos + w * sin + matrix[0, 2] += (new_w - w) * 0.5 + matrix[1, 2] += (new_h - h) * 0.5 + w = int(np.round(new_w)) + h = int(np.round(new_h)) + rotated = cv2.warpAffine( + img, + matrix, (w, h), + flags=cv2_interp_codes[interpolation], + borderValue=border_value) + return rotated + + +def bbox_clip(bboxes, img_shape): + """Clip bboxes to fit the image shape. + + Args: + bboxes (ndarray): Shape (..., 4*k) + img_shape (tuple[int]): (height, width) of the image. + + Returns: + ndarray: Clipped bboxes. + """ + assert bboxes.shape[-1] % 4 == 0 + cmin = np.empty(bboxes.shape[-1], dtype=bboxes.dtype) + cmin[0::2] = img_shape[1] - 1 + cmin[1::2] = img_shape[0] - 1 + clipped_bboxes = np.maximum(np.minimum(bboxes, cmin), 0) + return clipped_bboxes + + +def bbox_scaling(bboxes, scale, clip_shape=None): + """Scaling bboxes w.r.t the box center. + + Args: + bboxes (ndarray): Shape(..., 4). + scale (float): Scaling factor. + clip_shape (tuple[int], optional): If specified, bboxes that exceed the + boundary will be clipped according to the given shape (h, w). + + Returns: + ndarray: Scaled bboxes. + """ + if float(scale) == 1.0: + scaled_bboxes = bboxes.copy() + else: + w = bboxes[..., 2] - bboxes[..., 0] + 1 + h = bboxes[..., 3] - bboxes[..., 1] + 1 + dw = (w * (scale - 1)) * 0.5 + dh = (h * (scale - 1)) * 0.5 + scaled_bboxes = bboxes + np.stack((-dw, -dh, dw, dh), axis=-1) + if clip_shape is not None: + return bbox_clip(scaled_bboxes, clip_shape) + else: + return scaled_bboxes + + +def imcrop(img, bboxes, scale=1.0, pad_fill=None): + """Crop image patches. + + 3 steps: scale the bboxes -> clip bboxes -> crop and pad. + + Args: + img (ndarray): Image to be cropped. + bboxes (ndarray): Shape (k, 4) or (4, ), location of cropped bboxes. + scale (float, optional): Scale ratio of bboxes, the default value + 1.0 means no padding. + pad_fill (Number | list[Number]): Value to be filled for padding. + Default: None, which means no padding. + + Returns: + list[ndarray] | ndarray: The cropped image patches. + """ + chn = 1 if img.ndim == 2 else img.shape[2] + if pad_fill is not None: + if isinstance(pad_fill, (int, float)): + pad_fill = [pad_fill for _ in range(chn)] + assert len(pad_fill) == chn + + _bboxes = bboxes[None, ...] if bboxes.ndim == 1 else bboxes + scaled_bboxes = bbox_scaling(_bboxes, scale).astype(np.int32) + clipped_bbox = bbox_clip(scaled_bboxes, img.shape) + + patches = [] + for i in range(clipped_bbox.shape[0]): + x1, y1, x2, y2 = tuple(clipped_bbox[i, :]) + if pad_fill is None: + patch = img[y1:y2 + 1, x1:x2 + 1, ...] + else: + _x1, _y1, _x2, _y2 = tuple(scaled_bboxes[i, :]) + if chn == 1: + patch_shape = (_y2 - _y1 + 1, _x2 - _x1 + 1) + else: + patch_shape = (_y2 - _y1 + 1, _x2 - _x1 + 1, chn) + patch = np.array( + pad_fill, dtype=img.dtype) * np.ones( + patch_shape, dtype=img.dtype) + x_start = 0 if _x1 >= 0 else -_x1 + y_start = 0 if _y1 >= 0 else -_y1 + w = x2 - x1 + 1 + h = y2 - y1 + 1 + patch[y_start:y_start + h, x_start:x_start + w, + ...] = img[y1:y1 + h, x1:x1 + w, ...] + patches.append(patch) + + if bboxes.ndim == 1: + return patches[0] + else: + return patches + + +def impad(img, + *, + shape=None, + padding=None, + pad_val=0, + padding_mode='constant'): + """Pad the given image to a certain shape or pad on all sides with + specified padding mode and padding value. + + Args: + img (ndarray): Image to be padded. + shape (tuple[int]): Expected padding shape (h, w). Default: None. + padding (int or tuple[int]): Padding on each border. If a single int is + provided this is used to pad all borders. If tuple of length 2 is + provided this is the padding on left/right and top/bottom + respectively. If a tuple of length 4 is provided this is the + padding for the left, top, right and bottom borders respectively. + Default: None. Note that `shape` and `padding` can not be both + set. + pad_val (Number | Sequence[Number]): Values to be filled in padding + areas when padding_mode is 'constant'. Default: 0. + padding_mode (str): Type of padding. Should be: constant, edge, + reflect or symmetric. Default: constant. + - constant: pads with a constant value, this value is specified + with pad_val. + - edge: pads with the last value at the edge of the image. + - reflect: pads with reflection of image without repeating the last + value on the edge. For example, padding [1, 2, 3, 4] with 2 + elements on both sides in reflect mode will result in + [3, 2, 1, 2, 3, 4, 3, 2]. + - symmetric: pads with reflection of image repeating the last value + on the edge. For example, padding [1, 2, 3, 4] with 2 elements on + both sides in symmetric mode will result in + [2, 1, 1, 2, 3, 4, 4, 3] + + Returns: + ndarray: The padded image. + """ + + assert (shape is not None) ^ (padding is not None) + if shape is not None: + width = max(shape[1] - img.shape[1], 0) + height = max(shape[0] - img.shape[0], 0) + padding = (0, 0, width, height) + + # check pad_val + if isinstance(pad_val, tuple): + assert len(pad_val) == img.shape[-1] + elif not isinstance(pad_val, numbers.Number): + raise TypeError('pad_val must be a int or a tuple. ' + f'But received {type(pad_val)}') + + # check padding + if isinstance(padding, tuple) and len(padding) in [2, 4]: + if len(padding) == 2: + padding = (padding[0], padding[1], padding[0], padding[1]) + elif isinstance(padding, numbers.Number): + padding = (padding, padding, padding, padding) + else: + raise ValueError('Padding must be a int or a 2, or 4 element tuple.' + f'But received {padding}') + + # check padding mode + assert padding_mode in ['constant', 'edge', 'reflect', 'symmetric'] + + border_type = { + 'constant': cv2.BORDER_CONSTANT, + 'edge': cv2.BORDER_REPLICATE, + 'reflect': cv2.BORDER_REFLECT_101, + 'symmetric': cv2.BORDER_REFLECT + } + img = cv2.copyMakeBorder( + img, + padding[1], + padding[3], + padding[0], + padding[2], + border_type[padding_mode], + value=pad_val) + + return img + + +def impad_to_multiple(img, divisor, pad_val=0): + """Pad an image to ensure each edge to be multiple to some number. + + Args: + img (ndarray): Image to be padded. + divisor (int): Padded image edges will be multiple to divisor. + pad_val (Number | Sequence[Number]): Same as :func:`impad`. + + Returns: + ndarray: The padded image. + """ + pad_h = int(np.ceil(img.shape[0] / divisor)) * divisor + pad_w = int(np.ceil(img.shape[1] / divisor)) * divisor + return impad(img, shape=(pad_h, pad_w), pad_val=pad_val) + + +def cutout(img, shape, pad_val=0): + """Randomly cut out a rectangle from the original img. + + Args: + img (ndarray): Image to be cutout. + shape (int | tuple[int]): Expected cutout shape (h, w). If given as a + int, the value will be used for both h and w. + pad_val (int | float | tuple[int | float]): Values to be filled in the + cut area. Defaults to 0. + + Returns: + ndarray: The cutout image. + """ + + channels = 1 if img.ndim == 2 else img.shape[2] + if isinstance(shape, int): + cut_h, cut_w = shape, shape + else: + assert isinstance(shape, tuple) and len(shape) == 2, \ + f'shape must be a int or a tuple with length 2, but got type ' \ + f'{type(shape)} instead.' + cut_h, cut_w = shape + if isinstance(pad_val, (int, float)): + pad_val = tuple([pad_val] * channels) + elif isinstance(pad_val, tuple): + assert len(pad_val) == channels, \ + 'Expected the num of elements in tuple equals the channels' \ + 'of input image. Found {} vs {}'.format( + len(pad_val), channels) + else: + raise TypeError(f'Invalid type {type(pad_val)} for `pad_val`') + + img_h, img_w = img.shape[:2] + y0 = np.random.uniform(img_h) + x0 = np.random.uniform(img_w) + + y1 = int(max(0, y0 - cut_h / 2.)) + x1 = int(max(0, x0 - cut_w / 2.)) + y2 = min(img_h, y1 + cut_h) + x2 = min(img_w, x1 + cut_w) + + if img.ndim == 2: + patch_shape = (y2 - y1, x2 - x1) + else: + patch_shape = (y2 - y1, x2 - x1, channels) + + img_cutout = img.copy() + patch = np.array( + pad_val, dtype=img.dtype) * np.ones( + patch_shape, dtype=img.dtype) + img_cutout[y1:y2, x1:x2, ...] = patch + + return img_cutout + + +def _get_shear_matrix(magnitude, direction='horizontal'): + """Generate the shear matrix for transformation. + + Args: + magnitude (int | float): The magnitude used for shear. + direction (str): The flip direction, either "horizontal" + or "vertical". + + Returns: + ndarray: The shear matrix with dtype float32. + """ + if direction == 'horizontal': + shear_matrix = np.float32([[1, magnitude, 0], [0, 1, 0]]) + elif direction == 'vertical': + shear_matrix = np.float32([[1, 0, 0], [magnitude, 1, 0]]) + return shear_matrix + + +def imshear(img, + magnitude, + direction='horizontal', + border_value=0, + interpolation='bilinear'): + """Shear an image. + + Args: + img (ndarray): Image to be sheared with format (h, w) + or (h, w, c). + magnitude (int | float): The magnitude used for shear. + direction (str): The flip direction, either "horizontal" + or "vertical". + border_value (int | tuple[int]): Value used in case of a + constant border. + interpolation (str): Same as :func:`resize`. + + Returns: + ndarray: The sheared image. + """ + assert direction in ['horizontal', + 'vertical'], f'Invalid direction: {direction}' + height, width = img.shape[:2] + if img.ndim == 2: + channels = 1 + elif img.ndim == 3: + channels = img.shape[-1] + if isinstance(border_value, int): + border_value = tuple([border_value] * channels) + elif isinstance(border_value, tuple): + assert len(border_value) == channels, \ + 'Expected the num of elements in tuple equals the channels' \ + 'of input image. Found {} vs {}'.format( + len(border_value), channels) + else: + raise ValueError( + f'Invalid type {type(border_value)} for `border_value`') + shear_matrix = _get_shear_matrix(magnitude, direction) + sheared = cv2.warpAffine( + img, + shear_matrix, + (width, height), + # Note case when the number elements in `border_value` + # greater than 3 (e.g. shearing masks whose channels large + # than 3) will raise TypeError in `cv2.warpAffine`. + # Here simply slice the first 3 values in `border_value`. + borderValue=border_value[:3], + flags=cv2_interp_codes[interpolation]) + return sheared + + +def _get_translate_matrix(offset, direction='horizontal'): + """Generate the translate matrix. + + Args: + offset (int | float): The offset used for translate. + direction (str): The translate direction, either + "horizontal" or "vertical". + + Returns: + ndarray: The translate matrix with dtype float32. + """ + if direction == 'horizontal': + translate_matrix = np.float32([[1, 0, offset], [0, 1, 0]]) + elif direction == 'vertical': + translate_matrix = np.float32([[1, 0, 0], [0, 1, offset]]) + return translate_matrix + + +def imtranslate(img, + offset, + direction='horizontal', + border_value=0, + interpolation='bilinear'): + """Translate an image. + + Args: + img (ndarray): Image to be translated with format + (h, w) or (h, w, c). + offset (int | float): The offset used for translate. + direction (str): The translate direction, either "horizontal" + or "vertical". + border_value (int | tuple[int]): Value used in case of a + constant border. + interpolation (str): Same as :func:`resize`. + + Returns: + ndarray: The translated image. + """ + assert direction in ['horizontal', + 'vertical'], f'Invalid direction: {direction}' + height, width = img.shape[:2] + if img.ndim == 2: + channels = 1 + elif img.ndim == 3: + channels = img.shape[-1] + if isinstance(border_value, int): + border_value = tuple([border_value] * channels) + elif isinstance(border_value, tuple): + assert len(border_value) == channels, \ + 'Expected the num of elements in tuple equals the channels' \ + 'of input image. Found {} vs {}'.format( + len(border_value), channels) + else: + raise ValueError( + f'Invalid type {type(border_value)} for `border_value`.') + translate_matrix = _get_translate_matrix(offset, direction) + translated = cv2.warpAffine( + img, + translate_matrix, + (width, height), + # Note case when the number elements in `border_value` + # greater than 3 (e.g. translating masks whose channels + # large than 3) will raise TypeError in `cv2.warpAffine`. + # Here simply slice the first 3 values in `border_value`. + borderValue=border_value[:3], + flags=cv2_interp_codes[interpolation]) + return translated diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/io.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/io.py new file mode 100644 index 0000000000000000000000000000000000000000..ae81b561a84cccfa4923364679dce56d762db1bc --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/io.py @@ -0,0 +1,314 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import io +import os.path as osp +import warnings +from pathlib import Path + +import cv2 +import numpy as np +from cv2 import (IMREAD_COLOR, IMREAD_GRAYSCALE, IMREAD_IGNORE_ORIENTATION, + IMREAD_UNCHANGED) + +from mmcv.fileio import FileClient +from mmcv.utils import is_filepath, is_str + +try: + from turbojpeg import TJCS_RGB, TJPF_BGR, TJPF_GRAY, TurboJPEG +except ImportError: + TJCS_RGB = TJPF_GRAY = TJPF_BGR = TurboJPEG = None + +try: + from PIL import Image, ImageOps +except ImportError: + Image = None + +try: + import tifffile +except ImportError: + tifffile = None + +jpeg = None +supported_backends = ['cv2', 'turbojpeg', 'pillow', 'tifffile'] + +imread_flags = { + 'color': IMREAD_COLOR, + 'grayscale': IMREAD_GRAYSCALE, + 'unchanged': IMREAD_UNCHANGED, + 'color_ignore_orientation': IMREAD_IGNORE_ORIENTATION | IMREAD_COLOR, + 'grayscale_ignore_orientation': + IMREAD_IGNORE_ORIENTATION | IMREAD_GRAYSCALE +} + +imread_backend = 'cv2' + + +def use_backend(backend): + """Select a backend for image decoding. + + Args: + backend (str): The image decoding backend type. Options are `cv2`, + `pillow`, `turbojpeg` (see https://github.com/lilohuang/PyTurboJPEG) + and `tifffile`. `turbojpeg` is faster but it only supports `.jpeg` + file format. + """ + assert backend in supported_backends + global imread_backend + imread_backend = backend + if imread_backend == 'turbojpeg': + if TurboJPEG is None: + raise ImportError('`PyTurboJPEG` is not installed') + global jpeg + if jpeg is None: + jpeg = TurboJPEG() + elif imread_backend == 'pillow': + if Image is None: + raise ImportError('`Pillow` is not installed') + elif imread_backend == 'tifffile': + if tifffile is None: + raise ImportError('`tifffile` is not installed') + + +def _jpegflag(flag='color', channel_order='bgr'): + channel_order = channel_order.lower() + if channel_order not in ['rgb', 'bgr']: + raise ValueError('channel order must be either "rgb" or "bgr"') + + if flag == 'color': + if channel_order == 'bgr': + return TJPF_BGR + elif channel_order == 'rgb': + return TJCS_RGB + elif flag == 'grayscale': + return TJPF_GRAY + else: + raise ValueError('flag must be "color" or "grayscale"') + + +def _pillow2array(img, flag='color', channel_order='bgr'): + """Convert a pillow image to numpy array. + + Args: + img (:obj:`PIL.Image.Image`): The image loaded using PIL + flag (str): Flags specifying the color type of a loaded image, + candidates are 'color', 'grayscale' and 'unchanged'. + Default to 'color'. + channel_order (str): The channel order of the output image array, + candidates are 'bgr' and 'rgb'. Default to 'bgr'. + + Returns: + np.ndarray: The converted numpy array + """ + channel_order = channel_order.lower() + if channel_order not in ['rgb', 'bgr']: + raise ValueError('channel order must be either "rgb" or "bgr"') + + if flag == 'unchanged': + array = np.array(img) + if array.ndim >= 3 and array.shape[2] >= 3: # color image + array[:, :, :3] = array[:, :, (2, 1, 0)] # RGB to BGR + else: + # Handle exif orientation tag + if flag in ['color', 'grayscale']: + img = ImageOps.exif_transpose(img) + # If the image mode is not 'RGB', convert it to 'RGB' first. + if img.mode != 'RGB': + if img.mode != 'LA': + # Most formats except 'LA' can be directly converted to RGB + img = img.convert('RGB') + else: + # When the mode is 'LA', the default conversion will fill in + # the canvas with black, which sometimes shadows black objects + # in the foreground. + # + # Therefore, a random color (124, 117, 104) is used for canvas + img_rgba = img.convert('RGBA') + img = Image.new('RGB', img_rgba.size, (124, 117, 104)) + img.paste(img_rgba, mask=img_rgba.split()[3]) # 3 is alpha + if flag in ['color', 'color_ignore_orientation']: + array = np.array(img) + if channel_order != 'rgb': + array = array[:, :, ::-1] # RGB to BGR + elif flag in ['grayscale', 'grayscale_ignore_orientation']: + img = img.convert('L') + array = np.array(img) + else: + raise ValueError( + 'flag must be "color", "grayscale", "unchanged", ' + f'"color_ignore_orientation" or "grayscale_ignore_orientation"' + f' but got {flag}') + return array + + +def imread(img_or_path, + flag='color', + channel_order='bgr', + backend=None, + file_client_args=None): + """Read an image. + + Note: + In v1.4.1 and later, add `file_client_args` parameters. + + Args: + img_or_path (ndarray or str or Path): Either a numpy array or str or + pathlib.Path. If it is a numpy array (loaded image), then + it will be returned as is. + flag (str): Flags specifying the color type of a loaded image, + candidates are `color`, `grayscale`, `unchanged`, + `color_ignore_orientation` and `grayscale_ignore_orientation`. + By default, `cv2` and `pillow` backend would rotate the image + according to its EXIF info unless called with `unchanged` or + `*_ignore_orientation` flags. `turbojpeg` and `tifffile` backend + always ignore image's EXIF info regardless of the flag. + The `turbojpeg` backend only supports `color` and `grayscale`. + channel_order (str): Order of channel, candidates are `bgr` and `rgb`. + backend (str | None): The image decoding backend type. Options are + `cv2`, `pillow`, `turbojpeg`, `tifffile`, `None`. + If backend is None, the global imread_backend specified by + ``mmcv.use_backend()`` will be used. Default: None. + file_client_args (dict | None): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + + Returns: + ndarray: Loaded image array. + + Examples: + >>> import mmcv + >>> img_path = '/path/to/img.jpg' + >>> img = mmcv.imread(img_path) + >>> img = mmcv.imread(img_path, flag='color', channel_order='rgb', + ... backend='cv2') + >>> img = mmcv.imread(img_path, flag='color', channel_order='bgr', + ... backend='pillow') + >>> s3_img_path = 's3://bucket/img.jpg' + >>> # infer the file backend by the prefix s3 + >>> img = mmcv.imread(s3_img_path) + >>> # manually set the file backend petrel + >>> img = mmcv.imread(s3_img_path, file_client_args={ + ... 'backend': 'petrel'}) + >>> http_img_path = 'http://path/to/img.jpg' + >>> img = mmcv.imread(http_img_path) + >>> img = mmcv.imread(http_img_path, file_client_args={ + ... 'backend': 'http'}) + """ + + if isinstance(img_or_path, Path): + img_or_path = str(img_or_path) + + if isinstance(img_or_path, np.ndarray): + return img_or_path + elif is_str(img_or_path): + file_client = FileClient.infer_client(file_client_args, img_or_path) + img_bytes = file_client.get(img_or_path) + return imfrombytes(img_bytes, flag, channel_order, backend) + else: + raise TypeError('"img" must be a numpy array or a str or ' + 'a pathlib.Path object') + + +def imfrombytes(content, flag='color', channel_order='bgr', backend=None): + """Read an image from bytes. + + Args: + content (bytes): Image bytes got from files or other streams. + flag (str): Same as :func:`imread`. + channel_order (str): The channel order of the output, candidates + are 'bgr' and 'rgb'. Default to 'bgr'. + backend (str | None): The image decoding backend type. Options are + `cv2`, `pillow`, `turbojpeg`, `tifffile`, `None`. If backend is + None, the global imread_backend specified by ``mmcv.use_backend()`` + will be used. Default: None. + + Returns: + ndarray: Loaded image array. + + Examples: + >>> img_path = '/path/to/img.jpg' + >>> with open(img_path, 'rb') as f: + >>> img_buff = f.read() + >>> img = mmcv.imfrombytes(img_buff) + >>> img = mmcv.imfrombytes(img_buff, flag='color', channel_order='rgb') + >>> img = mmcv.imfrombytes(img_buff, backend='pillow') + >>> img = mmcv.imfrombytes(img_buff, backend='cv2') + """ + + if backend is None: + backend = imread_backend + if backend not in supported_backends: + raise ValueError( + f'backend: {backend} is not supported. Supported ' + "backends are 'cv2', 'turbojpeg', 'pillow', 'tifffile'") + if backend == 'turbojpeg': + img = jpeg.decode(content, _jpegflag(flag, channel_order)) + if img.shape[-1] == 1: + img = img[:, :, 0] + return img + elif backend == 'pillow': + with io.BytesIO(content) as buff: + img = Image.open(buff) + img = _pillow2array(img, flag, channel_order) + return img + elif backend == 'tifffile': + with io.BytesIO(content) as buff: + img = tifffile.imread(buff) + return img + else: + img_np = np.frombuffer(content, np.uint8) + flag = imread_flags[flag] if is_str(flag) else flag + img = cv2.imdecode(img_np, flag) + if flag == IMREAD_COLOR and channel_order == 'rgb': + cv2.cvtColor(img, cv2.COLOR_BGR2RGB, img) + return img + + +def imwrite(img, + file_path, + params=None, + auto_mkdir=None, + file_client_args=None): + """Write image to file. + + Note: + In v1.4.1 and later, add `file_client_args` parameters. + + Warning: + The parameter `auto_mkdir` will be deprecated in the future and every + file clients will make directory automatically. + + Args: + img (ndarray): Image array to be written. + file_path (str): Image file path. + params (None or list): Same as opencv :func:`imwrite` interface. + auto_mkdir (bool): If the parent folder of `file_path` does not exist, + whether to create it automatically. It will be deprecated. + file_client_args (dict | None): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + + Returns: + bool: Successful or not. + + Examples: + >>> # write to hard disk client + >>> ret = mmcv.imwrite(img, '/path/to/img.jpg') + >>> # infer the file backend by the prefix s3 + >>> ret = mmcv.imwrite(img, 's3://bucket/img.jpg') + >>> # manually set the file backend petrel + >>> ret = mmcv.imwrite(img, 's3://bucket/img.jpg', file_client_args={ + ... 'backend': 'petrel'}) + """ + assert is_filepath(file_path) + file_path = str(file_path) + if auto_mkdir is not None: + warnings.warn( + 'The parameter `auto_mkdir` will be deprecated in the future and ' + 'every file clients will make directory automatically.') + file_client = FileClient.infer_client(file_client_args, file_path) + img_ext = osp.splitext(file_path)[-1] + # Encode image according to image suffix. + # For example, if image path is '/path/your/img.jpg', the encode + # format is '.jpg'. + flag, img_buff = cv2.imencode(img_ext, img, params) + file_client.put(img_buff.tobytes(), file_path) + return flag diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/misc.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/misc.py new file mode 100644 index 0000000000000000000000000000000000000000..43934a689dd7ac6d35b772b7ce9921ff3b1fff50 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/misc.py @@ -0,0 +1,53 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import numpy as np + +import mmcv + +try: + import torch +except ImportError: + torch = None + + +def tensor2imgs(tensor, mean=None, std=None, to_rgb=True): + """Convert tensor to 3-channel images or 1-channel gray images. + + Args: + tensor (torch.Tensor): Tensor that contains multiple images, shape ( + N, C, H, W). :math:`C` can be either 3 or 1. + mean (tuple[float], optional): Mean of images. If None, + (0, 0, 0) will be used for tensor with 3-channel, + while (0, ) for tensor with 1-channel. Defaults to None. + std (tuple[float], optional): Standard deviation of images. If None, + (1, 1, 1) will be used for tensor with 3-channel, + while (1, ) for tensor with 1-channel. Defaults to None. + to_rgb (bool, optional): Whether the tensor was converted to RGB + format in the first place. If so, convert it back to BGR. + For the tensor with 1 channel, it must be False. Defaults to True. + + Returns: + list[np.ndarray]: A list that contains multiple images. + """ + + if torch is None: + raise RuntimeError('pytorch is not installed') + assert torch.is_tensor(tensor) and tensor.ndim == 4 + channels = tensor.size(1) + assert channels in [1, 3] + if mean is None: + mean = (0, ) * channels + if std is None: + std = (1, ) * channels + assert (channels == len(mean) == len(std) == 3) or \ + (channels == len(mean) == len(std) == 1 and not to_rgb) + + num_imgs = tensor.size(0) + mean = np.array(mean, dtype=np.float32) + std = np.array(std, dtype=np.float32) + imgs = [] + for img_id in range(num_imgs): + img = tensor[img_id, ...].cpu().numpy().transpose(1, 2, 0) + img = mmcv.imdenormalize( + img, mean, std, to_bgr=to_rgb).astype(np.uint8) + imgs.append(np.ascontiguousarray(img)) + return imgs diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/photometric.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/photometric.py new file mode 100644 index 0000000000000000000000000000000000000000..5085d012019c0cbf56f66f421a378278c1a058ae --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/image/photometric.py @@ -0,0 +1,428 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import cv2 +import numpy as np + +from ..utils import is_tuple_of +from .colorspace import bgr2gray, gray2bgr + + +def imnormalize(img, mean, std, to_rgb=True): + """Normalize an image with mean and std. + + Args: + img (ndarray): Image to be normalized. + mean (ndarray): The mean to be used for normalize. + std (ndarray): The std to be used for normalize. + to_rgb (bool): Whether to convert to rgb. + + Returns: + ndarray: The normalized image. + """ + img = img.copy().astype(np.float32) + return imnormalize_(img, mean, std, to_rgb) + + +def imnormalize_(img, mean, std, to_rgb=True): + """Inplace normalize an image with mean and std. + + Args: + img (ndarray): Image to be normalized. + mean (ndarray): The mean to be used for normalize. + std (ndarray): The std to be used for normalize. + to_rgb (bool): Whether to convert to rgb. + + Returns: + ndarray: The normalized image. + """ + # cv2 inplace normalization does not accept uint8 + assert img.dtype != np.uint8 + mean = np.float64(mean.reshape(1, -1)) + stdinv = 1 / np.float64(std.reshape(1, -1)) + if to_rgb: + cv2.cvtColor(img, cv2.COLOR_BGR2RGB, img) # inplace + cv2.subtract(img, mean, img) # inplace + cv2.multiply(img, stdinv, img) # inplace + return img + + +def imdenormalize(img, mean, std, to_bgr=True): + assert img.dtype != np.uint8 + mean = mean.reshape(1, -1).astype(np.float64) + std = std.reshape(1, -1).astype(np.float64) + img = cv2.multiply(img, std) # make a copy + cv2.add(img, mean, img) # inplace + if to_bgr: + cv2.cvtColor(img, cv2.COLOR_RGB2BGR, img) # inplace + return img + + +def iminvert(img): + """Invert (negate) an image. + + Args: + img (ndarray): Image to be inverted. + + Returns: + ndarray: The inverted image. + """ + return np.full_like(img, 255) - img + + +def solarize(img, thr=128): + """Solarize an image (invert all pixel values above a threshold) + + Args: + img (ndarray): Image to be solarized. + thr (int): Threshold for solarizing (0 - 255). + + Returns: + ndarray: The solarized image. + """ + img = np.where(img < thr, img, 255 - img) + return img + + +def posterize(img, bits): + """Posterize an image (reduce the number of bits for each color channel) + + Args: + img (ndarray): Image to be posterized. + bits (int): Number of bits (1 to 8) to use for posterizing. + + Returns: + ndarray: The posterized image. + """ + shift = 8 - bits + img = np.left_shift(np.right_shift(img, shift), shift) + return img + + +def adjust_color(img, alpha=1, beta=None, gamma=0): + r"""It blends the source image and its gray image: + + .. math:: + output = img * alpha + gray\_img * beta + gamma + + Args: + img (ndarray): The input source image. + alpha (int | float): Weight for the source image. Default 1. + beta (int | float): Weight for the converted gray image. + If None, it's assigned the value (1 - `alpha`). + gamma (int | float): Scalar added to each sum. + Same as :func:`cv2.addWeighted`. Default 0. + + Returns: + ndarray: Colored image which has the same size and dtype as input. + """ + gray_img = bgr2gray(img) + gray_img = np.tile(gray_img[..., None], [1, 1, 3]) + if beta is None: + beta = 1 - alpha + colored_img = cv2.addWeighted(img, alpha, gray_img, beta, gamma) + if not colored_img.dtype == np.uint8: + # Note when the dtype of `img` is not the default `np.uint8` + # (e.g. np.float32), the value in `colored_img` got from cv2 + # is not guaranteed to be in range [0, 255], so here clip + # is needed. + colored_img = np.clip(colored_img, 0, 255) + return colored_img + + +def imequalize(img): + """Equalize the image histogram. + + This function applies a non-linear mapping to the input image, + in order to create a uniform distribution of grayscale values + in the output image. + + Args: + img (ndarray): Image to be equalized. + + Returns: + ndarray: The equalized image. + """ + + def _scale_channel(im, c): + """Scale the data in the corresponding channel.""" + im = im[:, :, c] + # Compute the histogram of the image channel. + histo = np.histogram(im, 256, (0, 255))[0] + # For computing the step, filter out the nonzeros. + nonzero_histo = histo[histo > 0] + step = (np.sum(nonzero_histo) - nonzero_histo[-1]) // 255 + if not step: + lut = np.array(range(256)) + else: + # Compute the cumulative sum, shifted by step // 2 + # and then normalized by step. + lut = (np.cumsum(histo) + (step // 2)) // step + # Shift lut, prepending with 0. + lut = np.concatenate([[0], lut[:-1]], 0) + # handle potential integer overflow + lut[lut > 255] = 255 + # If step is zero, return the original image. + # Otherwise, index from lut. + return np.where(np.equal(step, 0), im, lut[im]) + + # Scales each channel independently and then stacks + # the result. + s1 = _scale_channel(img, 0) + s2 = _scale_channel(img, 1) + s3 = _scale_channel(img, 2) + equalized_img = np.stack([s1, s2, s3], axis=-1) + return equalized_img.astype(img.dtype) + + +def adjust_brightness(img, factor=1.): + """Adjust image brightness. + + This function controls the brightness of an image. An + enhancement factor of 0.0 gives a black image. + A factor of 1.0 gives the original image. This function + blends the source image and the degenerated black image: + + .. math:: + output = img * factor + degenerated * (1 - factor) + + Args: + img (ndarray): Image to be brightened. + factor (float): A value controls the enhancement. + Factor 1.0 returns the original image, lower + factors mean less color (brightness, contrast, + etc), and higher values more. Default 1. + + Returns: + ndarray: The brightened image. + """ + degenerated = np.zeros_like(img) + # Note manually convert the dtype to np.float32, to + # achieve as close results as PIL.ImageEnhance.Brightness. + # Set beta=1-factor, and gamma=0 + brightened_img = cv2.addWeighted( + img.astype(np.float32), factor, degenerated.astype(np.float32), + 1 - factor, 0) + brightened_img = np.clip(brightened_img, 0, 255) + return brightened_img.astype(img.dtype) + + +def adjust_contrast(img, factor=1.): + """Adjust image contrast. + + This function controls the contrast of an image. An + enhancement factor of 0.0 gives a solid grey + image. A factor of 1.0 gives the original image. It + blends the source image and the degenerated mean image: + + .. math:: + output = img * factor + degenerated * (1 - factor) + + Args: + img (ndarray): Image to be contrasted. BGR order. + factor (float): Same as :func:`mmcv.adjust_brightness`. + + Returns: + ndarray: The contrasted image. + """ + gray_img = bgr2gray(img) + hist = np.histogram(gray_img, 256, (0, 255))[0] + mean = round(np.sum(gray_img) / np.sum(hist)) + degenerated = (np.ones_like(img[..., 0]) * mean).astype(img.dtype) + degenerated = gray2bgr(degenerated) + contrasted_img = cv2.addWeighted( + img.astype(np.float32), factor, degenerated.astype(np.float32), + 1 - factor, 0) + contrasted_img = np.clip(contrasted_img, 0, 255) + return contrasted_img.astype(img.dtype) + + +def auto_contrast(img, cutoff=0): + """Auto adjust image contrast. + + This function maximize (normalize) image contrast by first removing cutoff + percent of the lightest and darkest pixels from the histogram and remapping + the image so that the darkest pixel becomes black (0), and the lightest + becomes white (255). + + Args: + img (ndarray): Image to be contrasted. BGR order. + cutoff (int | float | tuple): The cutoff percent of the lightest and + darkest pixels to be removed. If given as tuple, it shall be + (low, high). Otherwise, the single value will be used for both. + Defaults to 0. + + Returns: + ndarray: The contrasted image. + """ + + def _auto_contrast_channel(im, c, cutoff): + im = im[:, :, c] + # Compute the histogram of the image channel. + histo = np.histogram(im, 256, (0, 255))[0] + # Remove cut-off percent pixels from histo + histo_sum = np.cumsum(histo) + cut_low = histo_sum[-1] * cutoff[0] // 100 + cut_high = histo_sum[-1] - histo_sum[-1] * cutoff[1] // 100 + histo_sum = np.clip(histo_sum, cut_low, cut_high) - cut_low + histo = np.concatenate([[histo_sum[0]], np.diff(histo_sum)], 0) + + # Compute mapping + low, high = np.nonzero(histo)[0][0], np.nonzero(histo)[0][-1] + # If all the values have been cut off, return the origin img + if low >= high: + return im + scale = 255.0 / (high - low) + offset = -low * scale + lut = np.array(range(256)) + lut = lut * scale + offset + lut = np.clip(lut, 0, 255) + return lut[im] + + if isinstance(cutoff, (int, float)): + cutoff = (cutoff, cutoff) + else: + assert isinstance(cutoff, tuple), 'cutoff must be of type int, ' \ + f'float or tuple, but got {type(cutoff)} instead.' + # Auto adjusts contrast for each channel independently and then stacks + # the result. + s1 = _auto_contrast_channel(img, 0, cutoff) + s2 = _auto_contrast_channel(img, 1, cutoff) + s3 = _auto_contrast_channel(img, 2, cutoff) + contrasted_img = np.stack([s1, s2, s3], axis=-1) + return contrasted_img.astype(img.dtype) + + +def adjust_sharpness(img, factor=1., kernel=None): + """Adjust image sharpness. + + This function controls the sharpness of an image. An + enhancement factor of 0.0 gives a blurred image. A + factor of 1.0 gives the original image. And a factor + of 2.0 gives a sharpened image. It blends the source + image and the degenerated mean image: + + .. math:: + output = img * factor + degenerated * (1 - factor) + + Args: + img (ndarray): Image to be sharpened. BGR order. + factor (float): Same as :func:`mmcv.adjust_brightness`. + kernel (np.ndarray, optional): Filter kernel to be applied on the img + to obtain the degenerated img. Defaults to None. + + Note: + No value sanity check is enforced on the kernel set by users. So with + an inappropriate kernel, the ``adjust_sharpness`` may fail to perform + the function its name indicates but end up performing whatever + transform determined by the kernel. + + Returns: + ndarray: The sharpened image. + """ + + if kernel is None: + # adopted from PIL.ImageFilter.SMOOTH + kernel = np.array([[1., 1., 1.], [1., 5., 1.], [1., 1., 1.]]) / 13 + assert isinstance(kernel, np.ndarray), \ + f'kernel must be of type np.ndarray, but got {type(kernel)} instead.' + assert kernel.ndim == 2, \ + f'kernel must have a dimension of 2, but got {kernel.ndim} instead.' + + degenerated = cv2.filter2D(img, -1, kernel) + sharpened_img = cv2.addWeighted( + img.astype(np.float32), factor, degenerated.astype(np.float32), + 1 - factor, 0) + sharpened_img = np.clip(sharpened_img, 0, 255) + return sharpened_img.astype(img.dtype) + + +def adjust_lighting(img, eigval, eigvec, alphastd=0.1, to_rgb=True): + """AlexNet-style PCA jitter. + + This data augmentation is proposed in `ImageNet Classification with Deep + Convolutional Neural Networks + `_. + + Args: + img (ndarray): Image to be adjusted lighting. BGR order. + eigval (ndarray): the eigenvalue of the convariance matrix of pixel + values, respectively. + eigvec (ndarray): the eigenvector of the convariance matrix of pixel + values, respectively. + alphastd (float): The standard deviation for distribution of alpha. + Defaults to 0.1 + to_rgb (bool): Whether to convert img to rgb. + + Returns: + ndarray: The adjusted image. + """ + assert isinstance(eigval, np.ndarray) and isinstance(eigvec, np.ndarray), \ + f'eigval and eigvec should both be of type np.ndarray, got ' \ + f'{type(eigval)} and {type(eigvec)} instead.' + + assert eigval.ndim == 1 and eigvec.ndim == 2 + assert eigvec.shape == (3, eigval.shape[0]) + n_eigval = eigval.shape[0] + assert isinstance(alphastd, float), 'alphastd should be of type float, ' \ + f'got {type(alphastd)} instead.' + + img = img.copy().astype(np.float32) + if to_rgb: + cv2.cvtColor(img, cv2.COLOR_BGR2RGB, img) # inplace + + alpha = np.random.normal(0, alphastd, n_eigval) + alter = eigvec \ + * np.broadcast_to(alpha.reshape(1, n_eigval), (3, n_eigval)) \ + * np.broadcast_to(eigval.reshape(1, n_eigval), (3, n_eigval)) + alter = np.broadcast_to(alter.sum(axis=1).reshape(1, 1, 3), img.shape) + img_adjusted = img + alter + return img_adjusted + + +def lut_transform(img, lut_table): + """Transform array by look-up table. + + The function lut_transform fills the output array with values from the + look-up table. Indices of the entries are taken from the input array. + + Args: + img (ndarray): Image to be transformed. + lut_table (ndarray): look-up table of 256 elements; in case of + multi-channel input array, the table should either have a single + channel (in this case the same table is used for all channels) or + the same number of channels as in the input array. + + Returns: + ndarray: The transformed image. + """ + assert isinstance(img, np.ndarray) + assert 0 <= np.min(img) and np.max(img) <= 255 + assert isinstance(lut_table, np.ndarray) + assert lut_table.shape == (256, ) + + return cv2.LUT(np.array(img, dtype=np.uint8), lut_table) + + +def clahe(img, clip_limit=40.0, tile_grid_size=(8, 8)): + """Use CLAHE method to process the image. + + See `ZUIDERVELD,K. Contrast Limited Adaptive Histogram Equalization[J]. + Graphics Gems, 1994:474-485.` for more information. + + Args: + img (ndarray): Image to be processed. + clip_limit (float): Threshold for contrast limiting. Default: 40.0. + tile_grid_size (tuple[int]): Size of grid for histogram equalization. + Input image will be divided into equally sized rectangular tiles. + It defines the number of tiles in row and column. Default: (8, 8). + + Returns: + ndarray: The processed image. + """ + assert isinstance(img, np.ndarray) + assert img.ndim == 2 + assert isinstance(clip_limit, (float, int)) + assert is_tuple_of(tile_grid_size, int) + assert len(tile_grid_size) == 2 + + clahe = cv2.createCLAHE(clip_limit, tile_grid_size) + return clahe.apply(np.array(img, dtype=np.uint8)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/deprecated.json b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/deprecated.json new file mode 100644 index 0000000000000000000000000000000000000000..25cf6f28caecc22a77e3136fefa6b8dfc0e6cb5b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/deprecated.json @@ -0,0 +1,6 @@ +{ + "resnet50_caffe": "detectron/resnet50_caffe", + "resnet50_caffe_bgr": "detectron2/resnet50_caffe_bgr", + "resnet101_caffe": "detectron/resnet101_caffe", + "resnet101_caffe_bgr": "detectron2/resnet101_caffe_bgr" +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/mmcls.json b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/mmcls.json new file mode 100644 index 0000000000000000000000000000000000000000..c073a41d0aeb44ee0243f97ecc3558de538f9300 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/mmcls.json @@ -0,0 +1,59 @@ +{ + "vgg11": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg11_batch256_imagenet_20210208-4271cd6c.pth", + "vgg13": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg13_batch256_imagenet_20210208-4d1d6080.pth", + "vgg16": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg16_batch256_imagenet_20210208-db26f1a5.pth", + "vgg19": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg19_batch256_imagenet_20210208-e6920e4a.pth", + "vgg11_bn": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg11_bn_batch256_imagenet_20210207-f244902c.pth", + "vgg13_bn": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg13_bn_batch256_imagenet_20210207-1a8b7864.pth", + "vgg16_bn": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg16_bn_batch256_imagenet_20210208-7e55cd29.pth", + "vgg19_bn": "https://download.openmmlab.com/mmclassification/v0/vgg/vgg19_bn_batch256_imagenet_20210208-da620c4f.pth", + "resnet18": "https://download.openmmlab.com/mmclassification/v0/resnet/resnet18_8xb32_in1k_20210831-fbbb1da6.pth", + "resnet34": "https://download.openmmlab.com/mmclassification/v0/resnet/resnet34_8xb32_in1k_20210831-f257d4e6.pth", + "resnet50": "https://download.openmmlab.com/mmclassification/v0/resnet/resnet50_8xb32_in1k_20210831-ea4938fc.pth", + "resnet101": "https://download.openmmlab.com/mmclassification/v0/resnet/resnet101_8xb32_in1k_20210831-539c63f8.pth", + "resnet152": "https://download.openmmlab.com/mmclassification/v0/resnet/resnet152_8xb32_in1k_20210901-4d7582fa.pth", + "resnet50_v1d": "https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d50_b32x8_imagenet_20210531-db14775a.pth", + "resnet101_v1d": "https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d101_b32x8_imagenet_20210531-6e13bcd3.pth", + "resnet152_v1d": "https://download.openmmlab.com/mmclassification/v0/resnet/resnetv1d152_b32x8_imagenet_20210531-278cf22a.pth", + "resnext50_32x4d": "https://download.openmmlab.com/mmclassification/v0/resnext/resnext50_32x4d_b32x8_imagenet_20210429-56066e27.pth", + "resnext101_32x4d": "https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x4d_b32x8_imagenet_20210506-e0fa3dd5.pth", + "resnext101_32x8d": "https://download.openmmlab.com/mmclassification/v0/resnext/resnext101_32x8d_b32x8_imagenet_20210506-23a247d5.pth", + "resnext152_32x4d": "https://download.openmmlab.com/mmclassification/v0/resnext/resnext152_32x4d_b32x8_imagenet_20210524-927787be.pth", + "se-resnet50": "https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet50_batch256_imagenet_20200804-ae206104.pth", + "se-resnet101": "https://download.openmmlab.com/mmclassification/v0/se-resnet/se-resnet101_batch256_imagenet_20200804-ba5b51d4.pth", + "resnest50": "https://download.openmmlab.com/mmclassification/v0/resnest/resnest50_imagenet_converted-1ebf0afe.pth", + "resnest101": "https://download.openmmlab.com/mmclassification/v0/resnest/resnest101_imagenet_converted-032caa52.pth", + "resnest200": "https://download.openmmlab.com/mmclassification/v0/resnest/resnest200_imagenet_converted-581a60f2.pth", + "resnest269": "https://download.openmmlab.com/mmclassification/v0/resnest/resnest269_imagenet_converted-59930960.pth", + "shufflenet_v1": "https://download.openmmlab.com/mmclassification/v0/shufflenet_v1/shufflenet_v1_batch1024_imagenet_20200804-5d6cec73.pth", + "shufflenet_v2": "https://download.openmmlab.com/mmclassification/v0/shufflenet_v2/shufflenet_v2_batch1024_imagenet_20200812-5bf4721e.pth", + "mobilenet_v2": "https://download.openmmlab.com/mmclassification/v0/mobilenet_v2/mobilenet_v2_batch256_imagenet_20200708-3b2dc3af.pth", + "mobilenet_v3_small": "https://download.openmmlab.com/mmclassification/v0/mobilenet_v3/convert/mobilenet_v3_small-8427ecf0.pth", + "mobilenet_v3_large": "https://download.openmmlab.com/mmclassification/v0/mobilenet_v3/convert/mobilenet_v3_large-3ea3c186.pth", + "repvgg_A0": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A0_3rdparty_4xb64-coslr-120e_in1k_20210909-883ab98c.pth", + "repvgg_A1": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A1_3rdparty_4xb64-coslr-120e_in1k_20210909-24003a24.pth", + "repvgg_A2": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-A2_3rdparty_4xb64-coslr-120e_in1k_20210909-97d7695a.pth", + "repvgg_B0": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B0_3rdparty_4xb64-coslr-120e_in1k_20210909-446375f4.pth", + "repvgg_B1": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1_3rdparty_4xb64-coslr-120e_in1k_20210909-750cdf67.pth", + "repvgg_B1g2": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1g2_3rdparty_4xb64-coslr-120e_in1k_20210909-344f6422.pth", + "repvgg_B1g4": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B1g4_3rdparty_4xb64-coslr-120e_in1k_20210909-d4c1a642.pth", + "repvgg_B2": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B2_3rdparty_4xb64-coslr-120e_in1k_20210909-bd6b937c.pth", + "repvgg_B2g4": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B2g4_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-7b7955f0.pth", + "repvgg_B3": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B3_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-dda968bf.pth", + "repvgg_B3g4": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-B3g4_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-4e54846a.pth", + "repvgg_D2se": "https://download.openmmlab.com/mmclassification/v0/repvgg/repvgg-D2se_3rdparty_4xb64-autoaug-lbs-mixup-coslr-200e_in1k_20210909-cf3139b7.pth", + "res2net101_w26": "https://download.openmmlab.com/mmclassification/v0/res2net/res2net101-w26-s4_3rdparty_8xb32_in1k_20210927-870b6c36.pth", + "res2net50_w14": "https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w14-s8_3rdparty_8xb32_in1k_20210927-bc967bf1.pth", + "res2net50_w26": "https://download.openmmlab.com/mmclassification/v0/res2net/res2net50-w26-s8_3rdparty_8xb32_in1k_20210927-f547a94b.pth", + "swin_tiny": "https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925-66df6be6.pth", + "swin_small": "https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_small_224_b16x64_300e_imagenet_20210615_110219-7f9d988b.pth", + "swin_base": "https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_base_patch4_window7_224_22kto1k-f967f799.pth", + "swin_large": "https://download.openmmlab.com/mmclassification/v0/swin-transformer/convert/swin_large_patch4_window7_224_22kto1k-5f0996db.pth", + "t2t_vit_t_14": "https://download.openmmlab.com/mmclassification/v0/t2t-vit/t2t-vit-t-14_3rdparty_8xb64_in1k_20210928-b7c09b62.pth", + "t2t_vit_t_19": "https://download.openmmlab.com/mmclassification/v0/t2t-vit/t2t-vit-t-19_3rdparty_8xb64_in1k_20210928-7f1478d5.pth", + "t2t_vit_t_24": "https://download.openmmlab.com/mmclassification/v0/t2t-vit/t2t-vit-t-24_3rdparty_8xb64_in1k_20210928-fe95a61b.pth", + "tnt_small": "https://download.openmmlab.com/mmclassification/v0/tnt/tnt-small-p16_3rdparty_in1k_20210903-c56ee7df.pth", + "vit_base_p16": "https://download.openmmlab.com/mmclassification/v0/vit/finetune/vit-base-p16_in21k-pre-3rdparty_ft-64xb64_in1k-384_20210928-98e8652b.pth", + "vit_base_p32": "https://download.openmmlab.com/mmclassification/v0/vit/finetune/vit-base-p32_in21k-pre-3rdparty_ft-64xb64_in1k-384_20210928-9cea8599.pth", + "vit_large_p16": "https://download.openmmlab.com/mmclassification/v0/vit/finetune/vit-large-p16_in21k-pre-3rdparty_ft-64xb64_in1k-384_20210928-b20ba619.pth" +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/open_mmlab.json b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/open_mmlab.json new file mode 100644 index 0000000000000000000000000000000000000000..8311db4feef92faa0841c697d75efbee8430c3a0 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/open_mmlab.json @@ -0,0 +1,50 @@ +{ + "vgg16_caffe": "https://download.openmmlab.com/pretrain/third_party/vgg16_caffe-292e1171.pth", + "detectron/resnet50_caffe": "https://download.openmmlab.com/pretrain/third_party/resnet50_caffe-788b5fa3.pth", + "detectron2/resnet50_caffe": "https://download.openmmlab.com/pretrain/third_party/resnet50_msra-5891d200.pth", + "detectron/resnet101_caffe": "https://download.openmmlab.com/pretrain/third_party/resnet101_caffe-3ad79236.pth", + "detectron2/resnet101_caffe": "https://download.openmmlab.com/pretrain/third_party/resnet101_msra-6cc46731.pth", + "detectron2/resnext101_32x8d": "https://download.openmmlab.com/pretrain/third_party/resnext101_32x8d-1516f1aa.pth", + "resnext50_32x4d": "https://download.openmmlab.com/pretrain/third_party/resnext50-32x4d-0ab1a123.pth", + "resnext101_32x4d": "https://download.openmmlab.com/pretrain/third_party/resnext101_32x4d-a5af3160.pth", + "resnext101_64x4d": "https://download.openmmlab.com/pretrain/third_party/resnext101_64x4d-ee2c6f71.pth", + "contrib/resnet50_gn": "https://download.openmmlab.com/pretrain/third_party/resnet50_gn_thangvubk-ad1730dd.pth", + "detectron/resnet50_gn": "https://download.openmmlab.com/pretrain/third_party/resnet50_gn-9186a21c.pth", + "detectron/resnet101_gn": "https://download.openmmlab.com/pretrain/third_party/resnet101_gn-cac0ab98.pth", + "jhu/resnet50_gn_ws": "https://download.openmmlab.com/pretrain/third_party/resnet50_gn_ws-15beedd8.pth", + "jhu/resnet101_gn_ws": "https://download.openmmlab.com/pretrain/third_party/resnet101_gn_ws-3e3c308c.pth", + "jhu/resnext50_32x4d_gn_ws": "https://download.openmmlab.com/pretrain/third_party/resnext50_32x4d_gn_ws-0d87ac85.pth", + "jhu/resnext101_32x4d_gn_ws": "https://download.openmmlab.com/pretrain/third_party/resnext101_32x4d_gn_ws-34ac1a9e.pth", + "jhu/resnext50_32x4d_gn": "https://download.openmmlab.com/pretrain/third_party/resnext50_32x4d_gn-c7e8b754.pth", + "jhu/resnext101_32x4d_gn": "https://download.openmmlab.com/pretrain/third_party/resnext101_32x4d_gn-ac3bb84e.pth", + "msra/hrnetv2_w18_small": "https://download.openmmlab.com/pretrain/third_party/hrnetv2_w18_small-b5a04e21.pth", + "msra/hrnetv2_w18": "https://download.openmmlab.com/pretrain/third_party/hrnetv2_w18-00eb2006.pth", + "msra/hrnetv2_w32": "https://download.openmmlab.com/pretrain/third_party/hrnetv2_w32-dc9eeb4f.pth", + "msra/hrnetv2_w40": "https://download.openmmlab.com/pretrain/third_party/hrnetv2_w40-ed0b031c.pth", + "msra/hrnetv2_w48": "https://download.openmmlab.com/pretrain/third_party/hrnetv2_w48-d2186c55.pth", + "bninception_caffe": "https://download.openmmlab.com/pretrain/third_party/bn_inception_caffe-ed2e8665.pth", + "kin400/i3d_r50_f32s2_k400": "https://download.openmmlab.com/pretrain/third_party/i3d_r50_f32s2_k400-2c57e077.pth", + "kin400/nl3d_r50_f32s2_k400": "https://download.openmmlab.com/pretrain/third_party/nl3d_r50_f32s2_k400-fa7e7caa.pth", + "res2net101_v1d_26w_4s": "https://download.openmmlab.com/pretrain/third_party/res2net101_v1d_26w_4s_mmdetv2-f0a600f9.pth", + "regnetx_400mf": "https://download.openmmlab.com/pretrain/third_party/regnetx_400mf-a5b10d96.pth", + "regnetx_800mf": "https://download.openmmlab.com/pretrain/third_party/regnetx_800mf-1f4be4c7.pth", + "regnetx_1.6gf": "https://download.openmmlab.com/pretrain/third_party/regnetx_1.6gf-5791c176.pth", + "regnetx_3.2gf": "https://download.openmmlab.com/pretrain/third_party/regnetx_3.2gf-c2599b0f.pth", + "regnetx_4.0gf": "https://download.openmmlab.com/pretrain/third_party/regnetx_4.0gf-a88f671e.pth", + "regnetx_6.4gf": "https://download.openmmlab.com/pretrain/third_party/regnetx_6.4gf-006af45d.pth", + "regnetx_8.0gf": "https://download.openmmlab.com/pretrain/third_party/regnetx_8.0gf-3c68abe7.pth", + "regnetx_12gf": "https://download.openmmlab.com/pretrain/third_party/regnetx_12gf-4c2a3350.pth", + "resnet18_v1c": "https://download.openmmlab.com/pretrain/third_party/resnet18_v1c-b5776b93.pth", + "resnet50_v1c": "https://download.openmmlab.com/pretrain/third_party/resnet50_v1c-2cccc1ad.pth", + "resnet101_v1c": "https://download.openmmlab.com/pretrain/third_party/resnet101_v1c-e67eebb6.pth", + "mmedit/vgg16": "https://download.openmmlab.com/mmediting/third_party/vgg_state_dict.pth", + "mmedit/res34_en_nomixup": "https://download.openmmlab.com/mmediting/third_party/model_best_resnet34_En_nomixup.pth", + "mmedit/mobilenet_v2": "https://download.openmmlab.com/mmediting/third_party/mobilenet_v2.pth", + "contrib/mobilenet_v3_large": "https://download.openmmlab.com/pretrain/third_party/mobilenet_v3_large-bc2c3fd3.pth", + "contrib/mobilenet_v3_small": "https://download.openmmlab.com/pretrain/third_party/mobilenet_v3_small-47085aa1.pth", + "resnest50": "https://download.openmmlab.com/pretrain/third_party/resnest50_d2-7497a55b.pth", + "resnest101": "https://download.openmmlab.com/pretrain/third_party/resnest101_d2-f3b931b2.pth", + "resnest200": "https://download.openmmlab.com/pretrain/third_party/resnest200_d2-ca88e41f.pth", + "darknet53": "https://download.openmmlab.com/pretrain/third_party/darknet53-a628ea1b.pth", + "mmdet/mobilenet_v2": "https://download.openmmlab.com/mmdetection/v2.0/third_party/mobilenet_v2_batch256_imagenet-ff34753d.pth" +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/torchvision_0.12.json b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/torchvision_0.12.json new file mode 100644 index 0000000000000000000000000000000000000000..06defe67484dff91cf6f69109324cb1dd9d64bc3 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/model_zoo/torchvision_0.12.json @@ -0,0 +1,57 @@ +{ + "alexnet": "https://download.pytorch.org/models/alexnet-owt-7be5be79.pth", + "densenet121": "https://download.pytorch.org/models/densenet121-a639ec97.pth", + "densenet169": "https://download.pytorch.org/models/densenet169-b2777c0a.pth", + "densenet201": "https://download.pytorch.org/models/densenet201-c1103571.pth", + "densenet161": "https://download.pytorch.org/models/densenet161-8d451a50.pth", + "efficientnet_b0": "https://download.pytorch.org/models/efficientnet_b0_rwightman-3dd342df.pth", + "efficientnet_b1": "https://download.pytorch.org/models/efficientnet_b1_rwightman-533bc792.pth", + "efficientnet_b2": "https://download.pytorch.org/models/efficientnet_b2_rwightman-bcdf34b7.pth", + "efficientnet_b3": "https://download.pytorch.org/models/efficientnet_b3_rwightman-cf984f9c.pth", + "efficientnet_b4": "https://download.pytorch.org/models/efficientnet_b4_rwightman-7eb33cd5.pth", + "efficientnet_b5": "https://download.pytorch.org/models/efficientnet_b5_lukemelas-b6417697.pth", + "efficientnet_b6": "https://download.pytorch.org/models/efficientnet_b6_lukemelas-c76e70fd.pth", + "efficientnet_b7": "https://download.pytorch.org/models/efficientnet_b7_lukemelas-dcc49843.pth", + "googlenet": "https://download.pytorch.org/models/googlenet-1378be20.pth", + "inception_v3_google": "https://download.pytorch.org/models/inception_v3_google-0cc3c7bd.pth", + "mobilenet_v2": "https://download.pytorch.org/models/mobilenet_v2-b0353104.pth", + "mobilenet_v3_large": "https://download.pytorch.org/models/mobilenet_v3_large-8738ca79.pth", + "mobilenet_v3_small": "https://download.pytorch.org/models/mobilenet_v3_small-047dcff4.pth", + "regnet_y_400mf": "https://download.pytorch.org/models/regnet_y_400mf-c65dace8.pth", + "regnet_y_800mf": "https://download.pytorch.org/models/regnet_y_800mf-1b27b58c.pth", + "regnet_y_1_6gf": "https://download.pytorch.org/models/regnet_y_1_6gf-b11a554e.pth", + "regnet_y_3_2gf": "https://download.pytorch.org/models/regnet_y_3_2gf-b5a9779c.pth", + "regnet_y_8gf": "https://download.pytorch.org/models/regnet_y_8gf-d0d0e4a8.pth", + "regnet_y_16gf": "https://download.pytorch.org/models/regnet_y_16gf-9e6ed7dd.pth", + "regnet_y_32gf": "https://download.pytorch.org/models/regnet_y_32gf-4dee3f7a.pth", + "regnet_x_400mf": "https://download.pytorch.org/models/regnet_x_400mf-adf1edd5.pth", + "regnet_x_800mf": "https://download.pytorch.org/models/regnet_x_800mf-ad17e45c.pth", + "regnet_x_1_6gf": "https://download.pytorch.org/models/regnet_x_1_6gf-e3633e7f.pth", + "regnet_x_3_2gf": "https://download.pytorch.org/models/regnet_x_3_2gf-f342aeae.pth", + "regnet_x_8gf": "https://download.pytorch.org/models/regnet_x_8gf-03ceed89.pth", + "regnet_x_16gf": "https://download.pytorch.org/models/regnet_x_16gf-2007eb11.pth", + "regnet_x_32gf": "https://download.pytorch.org/models/regnet_x_32gf-9d47f8d0.pth", + "resnet18": "https://download.pytorch.org/models/resnet18-f37072fd.pth", + "resnet34": "https://download.pytorch.org/models/resnet34-b627a593.pth", + "resnet50": "https://download.pytorch.org/models/resnet50-0676ba61.pth", + "resnet101": "https://download.pytorch.org/models/resnet101-63fe2227.pth", + "resnet152": "https://download.pytorch.org/models/resnet152-394f9c45.pth", + "resnext50_32x4d": "https://download.pytorch.org/models/resnext50_32x4d-7cdf4587.pth", + "resnext101_32x8d": "https://download.pytorch.org/models/resnext101_32x8d-8ba56ff5.pth", + "wide_resnet50_2": "https://download.pytorch.org/models/wide_resnet50_2-95faca4d.pth", + "wide_resnet101_2": "https://download.pytorch.org/models/wide_resnet101_2-32ee1156.pth", + "shufflenetv2_x0.5": "https://download.pytorch.org/models/shufflenetv2_x0.5-f707e7126e.pth", + "shufflenetv2_x1.0": "https://download.pytorch.org/models/shufflenetv2_x1-5666bf0f80.pth", + "shufflenetv2_x1.5": null, + "shufflenetv2_x2.0": null, + "squeezenet1_0": "https://download.pytorch.org/models/squeezenet1_0-b66bff10.pth", + "squeezenet1_1": "https://download.pytorch.org/models/squeezenet1_1-b8a52dc0.pth", + "vgg11": "https://download.pytorch.org/models/vgg11-8a719046.pth", + "vgg13": "https://download.pytorch.org/models/vgg13-19584684.pth", + "vgg16": "https://download.pytorch.org/models/vgg16-397923af.pth", + "vgg19": "https://download.pytorch.org/models/vgg19-dcbb9e9d.pth", + "vgg11_bn": "https://download.pytorch.org/models/vgg11_bn-6002323d.pth", + "vgg13_bn": "https://download.pytorch.org/models/vgg13_bn-abd245e5.pth", + "vgg16_bn": "https://download.pytorch.org/models/vgg16_bn-6c64b313.pth", + "vgg19_bn": "https://download.pytorch.org/models/vgg19_bn-c79401a0.pth" +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..a12b79cefa7a248c06001bfeda648a4135ddde1d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/__init__.py @@ -0,0 +1,12 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .deprecated_wrappers import Conv2d_deprecated as Conv2d +from .deprecated_wrappers import ConvTranspose2d_deprecated as ConvTranspose2d +from .deprecated_wrappers import Linear_deprecated as Linear +from .deprecated_wrappers import MaxPool2d_deprecated as MaxPool2d +from .focal_loss import (SigmoidFocalLoss, SoftmaxFocalLoss, + sigmoid_focal_loss, softmax_focal_loss) +from .sync_bn import SyncBatchNorm +from .cc_attention import CrissCrossAttention +from .point_sample import * +from .psa_mask import PSAMask, PSAMaskFunction +from .info import * \ No newline at end of file diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/cc_attention.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/cc_attention.py new file mode 100644 index 0000000000000000000000000000000000000000..3fd83fcb9cc0692e0db79fe73938accac469a3d0 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/cc_attention.py @@ -0,0 +1,84 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +import torch.nn.functional as F + +from mmcv.cnn import PLUGIN_LAYERS, Scale + + +def NEG_INF_DIAG(n, device): + """Returns a diagonal matrix of size [n, n]. + + The diagonal are all "-inf". This is for avoiding calculating the + overlapped element in the Criss-Cross twice. + """ + return torch.diag(torch.tensor(float('-inf')).to(device).repeat(n), 0) + + +@PLUGIN_LAYERS.register_module() +class CrissCrossAttention(nn.Module): + """Criss-Cross Attention Module. + + .. note:: + Before v1.3.13, we use a CUDA op. Since v1.3.13, we switch + to a pure PyTorch and equivalent implementation. For more + details, please refer to https://github.com/open-mmlab/mmcv/pull/1201. + + Speed comparison for one forward pass + + - Input size: [2,512,97,97] + - Device: 1 NVIDIA GeForce RTX 2080 Ti + + +-----------------------+---------------+------------+---------------+ + | |PyTorch version|CUDA version|Relative speed | + +=======================+===============+============+===============+ + |with torch.no_grad() |0.00554402 s |0.0299619 s |5.4x | + +-----------------------+---------------+------------+---------------+ + |no with torch.no_grad()|0.00562803 s |0.0301349 s |5.4x | + +-----------------------+---------------+------------+---------------+ + + Args: + in_channels (int): Channels of the input feature map. + """ + + def __init__(self, in_channels): + super().__init__() + self.query_conv = nn.Conv2d(in_channels, in_channels // 8, 1) + self.key_conv = nn.Conv2d(in_channels, in_channels // 8, 1) + self.value_conv = nn.Conv2d(in_channels, in_channels, 1) + self.gamma = Scale(0.) + self.in_channels = in_channels + + def forward(self, x): + """forward function of Criss-Cross Attention. + + Args: + x (torch.Tensor): Input feature with the shape of + (batch_size, in_channels, height, width). + + Returns: + torch.Tensor: Output of the layer, with the shape of + (batch_size, in_channels, height, width) + """ + B, C, H, W = x.size() + query = self.query_conv(x) + key = self.key_conv(x) + value = self.value_conv(x) + energy_H = torch.einsum('bchw,bciw->bwhi', query, key) + NEG_INF_DIAG( + H, query.device) + energy_H = energy_H.transpose(1, 2) + energy_W = torch.einsum('bchw,bchj->bhwj', query, key) + attn = F.softmax( + torch.cat([energy_H, energy_W], dim=-1), dim=-1) # [B,H,W,(H+W)] + out = torch.einsum('bciw,bhwi->bchw', value, attn[..., :H]) + out += torch.einsum('bchj,bhwj->bchw', value, attn[..., H:]) + + out = self.gamma(out) + x + out = out.contiguous() + + return out + + def __repr__(self): + s = self.__class__.__name__ + s += f'(in_channels={self.in_channels})' + return s diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/README.md b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3bc02004017a0d607131b4de168b320c3beed23c --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/README.md @@ -0,0 +1,170 @@ +# Code Structure of CUDA operators + +This folder contains all non-python code for MMCV custom ops. Please follow the same architecture if you want to add new ops. + +## Directories Tree + +```folder +. +├── common +│ ├── box_iou_rotated_utils.hpp +│ ├── parrots_cpp_helper.hpp +│ ├── parrots_cuda_helper.hpp +│ ├── pytorch_cpp_helper.hpp +│ ├── pytorch_cuda_helper.hpp +│ ├── pytorch_device_registry.hpp +│   └── cuda +│   ├── common_cuda_helper.hpp +│   ├── parrots_cudawarpfunction.cuh +│   ├── ... +│   └── ops_cuda_kernel.cuh +├── onnxruntime +│   ├── onnxruntime_register.h +│   ├── onnxruntime_session_options_config_keys.h +│   ├── ort_mmcv_utils.h +│   ├── ... +│   ├── onnx_ops.h +│   └── cpu +│ ├── onnxruntime_register.cpp +│      ├── ... +│      └── onnx_ops_impl.cpp +├── parrots +│   ├── ... +│   ├── ops.cpp +│   ├── ops_parrots.cpp +│   └── ops_pytorch.h +├── pytorch +│   ├── info.cpp +│   ├── pybind.cpp +│   ├── ... +│   ├── ops.cpp +│   ├── cuda +│   │   ├── ... +│   │   └── ops_cuda.cu +│   └── cpu +│      ├── ... +│      └── ops.cpp +└── tensorrt + ├── trt_cuda_helper.cuh + ├── trt_plugin_helper.hpp + ├── trt_plugin.hpp + ├── trt_serialize.hpp + ├── ... + ├── trt_ops.hpp + └── plugins +    ├── trt_cuda_helper.cu +    ├── trt_plugin.cpp +    ├── ... +    ├── trt_ops.cpp +    └── trt_ops_kernel.cu +``` + +## Components + +- `common`: This directory contains all tools and shared codes. + - `cuda`: The cuda kernels which can be shared by all backends. **HIP** kernel is also here since they have similar syntax. +- `onnxruntime`: **ONNX Runtime** support for custom ops. + - `cpu`: CPU implementation of supported ops. +- `parrots`: **Parrots** is a deep learning frame for model training and inference. Parrots custom ops are placed in this directory. +- `pytorch`: **PyTorch** custom ops are supported by binding C++ to Python with **pybind11**. The ops implementation and binding codes are placed in this directory. + - `cuda`: This directory contains cuda kernel launchers, which feed memory pointers of tensor to the cuda kernel in `common/cuda`. The launchers provide c++ interface of cuda implementation of corresponding custom ops. + - `cpu`: This directory contain cpu implementations of corresponding custom ops. +- `tensorrt`: **TensorRT** support for custom ops. + - `plugins`: This directory contains the implementation of the supported custom ops. Some ops might also use shared cuda kernel in `common/cuda`. + +## How to add new PyTorch ops? + +1. (Optional) Add shared kernel in `common` to support special hardware platform. + + ```c++ + // src/common/cuda/new_ops_cuda_kernel.cuh + + template + __global__ void new_ops_forward_cuda_kernel(const T* input, T* output, ...) { + // forward here + } + + ``` + + Add cuda kernel launcher in `pytorch/cuda`. + + ```c++ + // src/pytorch/cuda + #include + + void NewOpsForwardCUDAKernelLauncher(Tensor input, Tensor output, ...){ + // initialize + at::cuda::CUDAGuard device_guard(input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + ... + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + input.scalar_type(), "new_ops_forward_cuda_kernel", ([&] { + new_ops_forward_cuda_kernel + <<>>( + input.data_ptr(), output.data_ptr(),...); + })); + AT_CUDA_CHECK(cudaGetLastError()); + } + ``` + +2. Register implementation for different devices. + + ```c++ + // src/pytorch/cuda/cudabind.cpp + ... + + Tensor new_ops_forward_cuda(Tensor input, Tensor output, ...){ + // implement cuda forward here + // use `NewOpsForwardCUDAKernelLauncher` here + } + // declare interface here. + Tensor new_ops_forward_impl(Tensor input, Tensor output, ...); + // register the implementation for given device (CUDA here). + REGISTER_DEVICE_IMPL(new_ops_forward_impl, CUDA, new_ops_forward_cuda); + ``` + +3. Add ops implementation in `pytorch` directory. Select different implementations according to device type. + + ```c++ + // src/pytorch/new_ops.cpp + Tensor new_ops_forward_impl(Tensor input, Tensor output, ...){ + // dispatch the implementation according to the device type of input. + DISPATCH_DEVICE_IMPL(new_ops_forward_impl, input, output, ...); + } + ... + + Tensor new_ops_forward(Tensor input, Tensor output, ...){ + return new_ops_forward_impl(input, output, ...); + } + ``` + +4. Binding the implementation in `pytorch/pybind.cpp` + + ```c++ + // src/pytorch/pybind.cpp + + ... + + Tensor new_ops_forward(Tensor input, Tensor output, ...); + + ... + + // bind with pybind11 + m.def("new_ops_forward", &new_ops_forward, "new_ops_forward", + py::arg("input"), py::arg("output"), ...); + + ... + + ``` + +5. Build MMCV again. Enjoy new ops in python + + ```python + from ..utils import ext_loader + ext_module = ext_loader.load_ext('_ext', ['new_ops_forward']) + + ... + + ext_module.new_ops_forward(input, output, ...) + + ``` diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/common_cuda_helper.hpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/common_cuda_helper.hpp new file mode 100644 index 0000000000000000000000000000000000000000..e18036bac56d401760be7e72f01c1abf241af1ef --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/common_cuda_helper.hpp @@ -0,0 +1,120 @@ +#ifndef COMMON_CUDA_HELPER +#define COMMON_CUDA_HELPER + +#include + +#define CUDA_1D_KERNEL_LOOP(i, n) \ + for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < (n); \ + i += blockDim.x * gridDim.x) + +#define CUDA_2D_KERNEL_LOOP(i, n, j, m) \ + for (size_t i = blockIdx.x * blockDim.x + threadIdx.x; i < (n); \ + i += blockDim.x * gridDim.x) \ + for (size_t j = blockIdx.y * blockDim.y + threadIdx.y; j < (m); \ + j += blockDim.y * gridDim.y) + +#define CUDA_2D_KERNEL_BLOCK_LOOP(i, n, j, m) \ + for (size_t i = blockIdx.x; i < (n); i += gridDim.x) \ + for (size_t j = blockIdx.y; j < (m); j += gridDim.y) + +#define THREADS_PER_BLOCK 512 + +inline int GET_BLOCKS(const int N, const int num_threads = THREADS_PER_BLOCK) { + int optimal_block_num = (N + num_threads - 1) / num_threads; + int max_block_num = 4096; + return std::min(optimal_block_num, max_block_num); +} + +template +__device__ T bilinear_interpolate(const T* input, const int height, + const int width, T y, T x, + const int index /* index for debug only*/) { + // deal with cases that inverse elements are out of feature map boundary + if (y < -1.0 || y > height || x < -1.0 || x > width) return 0; + + if (y <= 0) y = 0; + if (x <= 0) x = 0; + + int y_low = (int)y; + int x_low = (int)x; + int y_high; + int x_high; + + if (y_low >= height - 1) { + y_high = y_low = height - 1; + y = (T)y_low; + } else { + y_high = y_low + 1; + } + + if (x_low >= width - 1) { + x_high = x_low = width - 1; + x = (T)x_low; + } else { + x_high = x_low + 1; + } + + T ly = y - y_low; + T lx = x - x_low; + T hy = 1. - ly, hx = 1. - lx; + // do bilinear interpolation + T v1 = input[y_low * width + x_low]; + T v2 = input[y_low * width + x_high]; + T v3 = input[y_high * width + x_low]; + T v4 = input[y_high * width + x_high]; + T w1 = hy * hx, w2 = hy * lx, w3 = ly * hx, w4 = ly * lx; + + T val = (w1 * v1 + w2 * v2 + w3 * v3 + w4 * v4); + + return val; +} + +template +__device__ void bilinear_interpolate_gradient( + const int height, const int width, T y, T x, T& w1, T& w2, T& w3, T& w4, + int& x_low, int& x_high, int& y_low, int& y_high, + const int index /* index for debug only*/) { + // deal with cases that inverse elements are out of feature map boundary + if (y < -1.0 || y > height || x < -1.0 || x > width) { + // empty + w1 = w2 = w3 = w4 = 0.; + x_low = x_high = y_low = y_high = -1; + return; + } + + if (y <= 0) y = 0; + if (x <= 0) x = 0; + + y_low = (int)y; + x_low = (int)x; + + if (y_low >= height - 1) { + y_high = y_low = height - 1; + y = (T)y_low; + } else { + y_high = y_low + 1; + } + + if (x_low >= width - 1) { + x_high = x_low = width - 1; + x = (T)x_low; + } else { + x_high = x_low + 1; + } + + T ly = y - y_low; + T lx = x - x_low; + T hy = 1. - ly, hx = 1. - lx; + + // reference in forward + // T v1 = input[y_low * width + x_low]; + // T v2 = input[y_low * width + x_high]; + // T v3 = input[y_high * width + x_low]; + // T v4 = input[y_high * width + x_high]; + // T val = (w1 * v1 + w2 * v2 + w3 * v3 + w4 * v4); + + w1 = hy * hx, w2 = hy * lx, w3 = ly * hx, w4 = ly * lx; + + return; +} +#endif // COMMON_CUDA_HELPER diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/psamask_cuda_kernel.cuh b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/psamask_cuda_kernel.cuh new file mode 100644 index 0000000000000000000000000000000000000000..5d946686bdd5fdfbf8a27f6d040e15861202f471 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/psamask_cuda_kernel.cuh @@ -0,0 +1,141 @@ +// Copyright (c) OpenMMLab. All rights reserved +#ifndef PSAMASK_CUDA_KERNEL_CUH +#define PSAMASK_CUDA_KERNEL_CUH + +#ifdef MMCV_USE_PARROTS +#include "parrots_cuda_helper.hpp" +#else +#include "pytorch_cuda_helper.hpp" +#endif + +// CUDA: grid stride looping +#ifndef CUDA_KERNEL_LOOP +#define CUDA_KERNEL_LOOP(i, n) \ + for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < (n); \ + i += blockDim.x * gridDim.x) +#endif + +template +__global__ void psamask_collect_forward_cuda( + const int nthreads, const int h_feature, const int w_feature, + const int h_mask, const int w_mask, const int half_h_mask, + const int half_w_mask, const T* mask_data, T* buffer_data) { + CUDA_KERNEL_LOOP(index, nthreads) { + const int w = index % w_feature; + const int h = (index / w_feature) % h_feature; + const int n = index / w_feature / h_feature; + // effective mask region : [hstart, hend) x [wstart, wend) with mask-indexed + const int hstart = max(0, half_h_mask - h); + const int hend = min(h_mask, h_feature + half_h_mask - h); + const int wstart = max(0, half_w_mask - w); + const int wend = min(w_mask, w_feature + half_w_mask - w); + // (hidx, widx ) with mask-indexed + // (hidx + h - half_h_mask, widx + w - half_w_mask) with feature-indexed + for (int hidx = hstart; hidx < hend; hidx++) { + for (int widx = wstart; widx < wend; widx++) { + buffer_data[(n * h_feature * w_feature + + (hidx + h - half_h_mask) * w_feature + + (widx + w - half_w_mask)) * + h_feature * w_feature + + h * w_feature + w] = mask_data + [((n * h_mask * w_mask + hidx * w_mask + widx) * h_feature + h) * + w_feature + + w]; + } + } + } +} + +template +__global__ void psamask_distribute_forward_cuda( + const int nthreads, const int h_feature, const int w_feature, + const int h_mask, const int w_mask, const int half_h_mask, + const int half_w_mask, const T* mask_data, T* buffer_data) { + CUDA_KERNEL_LOOP(index, nthreads) { + const int w = index % w_feature; + const int h = (index / w_feature) % h_feature; + const int n = index / w_feature / h_feature; + // effective mask region : [hstart, hend) x [wstart, wend) with mask-indexed + const int hstart = max(0, half_h_mask - h); + const int hend = min(h_mask, h_feature + half_h_mask - h); + const int wstart = max(0, half_w_mask - w); + const int wend = min(w_mask, w_feature + half_w_mask - w); + // (hidx, widx ) with mask-indexed + // (hidx + h - half_h_mask, widx + w - half_w_mask) with feature-indexed + for (int hidx = hstart; hidx < hend; hidx++) { + for (int widx = wstart; widx < wend; widx++) { + buffer_data[(n * h_feature * w_feature + h * w_feature + w) * + h_feature * w_feature + + (hidx + h - half_h_mask) * w_feature + + (widx + w - half_w_mask)] = mask_data + [((n * h_mask * w_mask + hidx * w_mask + widx) * h_feature + h) * + w_feature + + w]; + } + } + } +} + +template +__global__ void psamask_collect_backward_cuda( + const int nthreads, const int h_feature, const int w_feature, + const int h_mask, const int w_mask, const int half_h_mask, + const int half_w_mask, const T* buffer_diff, T* mask_diff) { + CUDA_KERNEL_LOOP(index, nthreads) { + const int w = index % w_feature; + const int h = (index / w_feature) % h_feature; + const int n = index / w_feature / h_feature; + // effective mask region : [hstart, hend) x [wstart, wend) with mask-indexed + const int hstart = max(0, half_h_mask - h); + const int hend = min(h_mask, h_feature + half_h_mask - h); + const int wstart = max(0, half_w_mask - w); + const int wend = min(w_mask, w_feature + half_w_mask - w); + // (hidx, widx ) with mask-indexed + // (hidx + h - half_h_mask, widx + w - half_w_mask) with feature-indexed + for (int hidx = hstart; hidx < hend; hidx++) { + for (int widx = wstart; widx < wend; widx++) { + mask_diff[((n * h_mask * w_mask + hidx * w_mask + widx) * h_feature + + h) * + w_feature + + w] = buffer_diff[(n * h_feature * w_feature + + (hidx + h - half_h_mask) * w_feature + + (widx + w - half_w_mask)) * + h_feature * w_feature + + h * w_feature + w]; + } + } + } +} + +template +__global__ void psamask_distribute_backward_cuda( + const int nthreads, const int h_feature, const int w_feature, + const int h_mask, const int w_mask, const int half_h_mask, + const int half_w_mask, const T* buffer_diff, T* mask_diff) { + CUDA_KERNEL_LOOP(index, nthreads) { + const int w = index % w_feature; + const int h = (index / w_feature) % h_feature; + const int n = index / w_feature / h_feature; + // effective mask region : [hstart, hend) x [wstart, wend) with mask-indexed + const int hstart = max(0, half_h_mask - h); + const int hend = min(h_mask, h_feature + half_h_mask - h); + const int wstart = max(0, half_w_mask - w); + const int wend = min(w_mask, w_feature + half_w_mask - w); + // (hidx, widx ) with mask-indexed + // (hidx + h - half_h_mask, widx + w - half_w_mask) with feature-indexed + for (int hidx = hstart; hidx < hend; hidx++) { + for (int widx = wstart; widx < wend; widx++) { + mask_diff[((n * h_mask * w_mask + hidx * w_mask + widx) * h_feature + + h) * + w_feature + + w] = + buffer_diff[(n * h_feature * w_feature + h * w_feature + w) * + h_feature * w_feature + + (hidx + h - half_h_mask) * w_feature + + (widx + w - half_w_mask)]; + } + } + } +} + +#endif // PSAMASK_CUDA_KERNEL_CUH diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/sigmoid_focal_loss_cuda_kernel.cuh b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/sigmoid_focal_loss_cuda_kernel.cuh new file mode 100644 index 0000000000000000000000000000000000000000..1eb5f8fcccbaafdb62972652e3979803c0acd1ca --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/sigmoid_focal_loss_cuda_kernel.cuh @@ -0,0 +1,71 @@ +// Copyright (c) OpenMMLab. All rights reserved +#ifndef SIGMOID_FOCAL_LOSS_CUDA_KERNEL_CUH +#define SIGMOID_FOCAL_LOSS_CUDA_KERNEL_CUH + +#ifdef MMCV_USE_PARROTS +#include "parrots_cuda_helper.hpp" +#else +#include "pytorch_cuda_helper.hpp" +#endif + +template +__global__ void sigmoid_focal_loss_forward_cuda_kernel( + const int nthreads, const T* input, const int64_t* target, const T* weight, + T* output, const T gamma, const T alpha, const int num_classes) { + CUDA_1D_KERNEL_LOOP(index, nthreads) { + int n = index / num_classes; + int c = index % num_classes; + + int64_t t = target[n]; + T flag_p = (t == c); + T flag_n = (t != c); + + // p = sigmoid(x) = 1. / 1. + expf(-x) + T p = (T)1. / ((T)1. + expf(-input[index])); + + // (1 - p)**gamma * log(p) + T term_p = pow(((T)1. - p), gamma) * log(max(p, (T)FLT_MIN)); + // p**gamma * log(1 - p) + T term_n = pow(p, gamma) * log(max((T)1. - p, (T)FLT_MIN)); + + output[index] = (T)0.; + output[index] += -flag_p * alpha * term_p; + output[index] += -flag_n * ((T)1. - alpha) * term_n; + if (weight != NULL) { + output[index] *= weight[t]; + } + } +} + +template +__global__ void sigmoid_focal_loss_backward_cuda_kernel( + const int nthreads, const T* input, const int64_t* target, const T* weight, + T* grad_input, const T gamma, const T alpha, const int num_classes) { + CUDA_1D_KERNEL_LOOP(index, nthreads) { + int n = index / num_classes; + int c = index % num_classes; + + int64_t t = target[n]; + T flag_p = (t == c); + T flag_n = (t != c); + + // p = sigmoid(x) = 1. / 1. + expf(-x) + T p = (T)1. / ((T)1. + exp(-input[index])); + + // (1 - p)**gamma * (1 - p - gamma*p*log(p)) + T term_p = pow(((T)1. - p), gamma) * + ((T)1. - p - (gamma * p * log(max(p, (T)FLT_MIN)))); + // p**gamma * (gamma * (1 - p) * log(1 - p) - p) + T term_n = pow(p, gamma) * + (gamma * ((T)1. - p) * log(max((T)1. - p, (T)FLT_MIN)) - p); + + grad_input[index] = (T)0.; + grad_input[index] += -flag_p * alpha * term_p; + grad_input[index] += -flag_n * ((T)1. - alpha) * term_n; + if (weight != NULL) { + grad_input[index] *= weight[t]; + } + } +} + +#endif // SIGMOID_FOCAL_LOSS_CUDA_KERNEL_CUH diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/softmax_focal_loss_cuda_kernel.cuh b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/softmax_focal_loss_cuda_kernel.cuh new file mode 100644 index 0000000000000000000000000000000000000000..631b2c6175412a9503f6c385ee6597d9527d754f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/softmax_focal_loss_cuda_kernel.cuh @@ -0,0 +1,72 @@ +// Copyright (c) OpenMMLab. All rights reserved +#ifndef SOFTMAX_FOCAL_LOSS_CUDA_KERNEL_CUH +#define SOFTMAX_FOCAL_LOSS_CUDA_KERNEL_CUH + +#ifdef MMCV_USE_PARROTS +#include "parrots_cuda_helper.hpp" +#else +#include "pytorch_cuda_helper.hpp" +#endif + +template +__global__ void softmax_focal_loss_forward_cuda_kernel( + const int nthreads, const T* softmax, const int64_t* target, + const T* weight, T* output, const T gamma, const T alpha, + const int num_classes) { + CUDA_1D_KERNEL_LOOP(index, nthreads) { + int64_t label = target[index]; + T pred = softmax[index * num_classes + label]; + + if (label >= 0) { + output[index] = + -alpha * pow((T)1. - pred, gamma) * log(max(pred, (T)FLT_MIN)); + } else { + output[index] = 0; + } + if (weight != NULL) { + output[index] *= weight[label]; + } + } +} + +template +__global__ void softmax_focal_loss_backward_cuda1_kernel( + const int nthreads, const T* softmax, const int64_t* target, + const T* weight, T* buff, const T gamma, const T alpha, + const int num_classes) { + CUDA_1D_KERNEL_LOOP(index, nthreads) { + int64_t label = target[index]; + T pred = softmax[index * num_classes + label]; + + if (label >= 0) { + buff[index] = alpha * (-pow((T)1. - pred, gamma) + + gamma * pow((T)1. - pred, gamma - 1) * pred * + log(max(pred, (T)FLT_MIN))); + } else { + buff[index] = 0; + } + if (weight != NULL) { + buff[index] *= weight[label]; + } + } +} + +template +__global__ void softmax_focal_loss_backward_cuda2_kernel( + const int nthreads, const T* softmax, const int64_t* target, const T* buff, + T* grad_input, const int num_classes) { + CUDA_1D_KERNEL_LOOP(index, nthreads) { + int n = index / num_classes; + int c = index % num_classes; + int64_t label = target[n]; + + if (label >= 0) { + T flag = (label == c ? (T)1. : (T)0.); + grad_input[index] = buff[n] * (flag - softmax[index]); + } else { + grad_input[index] = 0; + } + } +} + +#endif // SOFTMAX_FOCAL_LOSS_CUDA_KERNEL_CUH diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/sync_bn_cuda_kernel.cuh b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/sync_bn_cuda_kernel.cuh new file mode 100644 index 0000000000000000000000000000000000000000..4ec6a466886832d38c72da6e3a3574e72d53cec8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/cuda/sync_bn_cuda_kernel.cuh @@ -0,0 +1,331 @@ +// Copyright (c) OpenMMLab. All rights reserved +#ifndef SYNCBN_CUDA_KERNEL_CUH +#define SYNCBN_CUDA_KERNEL_CUH + +#ifdef MMCV_USE_PARROTS +#include "parrots_cuda_helper.hpp" +#else +#include "pytorch_cuda_helper.hpp" +#endif + +template +__global__ void sync_bn_forward_mean_cuda_kernel(const T *input, float *mean, + int num, int channels, + int spatial) { + __shared__ float buffer[THREADS_PER_BLOCK]; + int tid = threadIdx.x; + int c = blockIdx.x; + buffer[tid] = 0; + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = (i / spatial) * channels * spatial + c * spatial + i % spatial; + buffer[tid] += input[index]; + } + __syncthreads(); + + for (int s = blockDim.x / 2; s > 0; s >>= 1) { + if (tid < s) { + buffer[tid] += buffer[tid + s]; + } + __syncthreads(); + } + int total = num * spatial; + if (tid == 0) { + mean[c] = buffer[0] / total; + } +} + +template <> +__global__ void sync_bn_forward_mean_cuda_kernel(const phalf *input, + float *mean, int num, + int channels, int spatial) { + __shared__ float buffer[THREADS_PER_BLOCK]; + int tid = threadIdx.x; + int c = blockIdx.x; + buffer[tid] = 0; + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = (i / spatial) * channels * spatial + c * spatial + i % spatial; + buffer[tid] += static_cast(input[index]); + } + __syncthreads(); + + for (int s = blockDim.x / 2; s > 0; s >>= 1) { + if (tid < s) { + buffer[tid] += buffer[tid + s]; + } + __syncthreads(); + } + int total = num * spatial; + if (tid == 0) { + mean[c] = buffer[0] / total; + } +} + +template +__global__ void sync_bn_forward_var_cuda_kernel(const T *input, + const float *mean, float *var, + int num, int channels, + int spatial) { + __shared__ float buffer[THREADS_PER_BLOCK]; + int tid = threadIdx.x; + int c = blockIdx.x; + buffer[tid] = 0; + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = (i / spatial) * channels * spatial + c * spatial + i % spatial; + float td = input[index] - mean[c]; + buffer[tid] += td * td; + } + __syncthreads(); + for (int s = blockDim.x / 2; s > 0; s >>= 1) { + if (tid < s) { + buffer[tid] += buffer[tid + s]; + } + __syncthreads(); + } + int total = num * spatial; + if (tid == 0) { + var[c] = buffer[0] / total; + } +} + +template <> +__global__ void sync_bn_forward_var_cuda_kernel(const phalf *input, + const float *mean, float *var, + int num, int channels, + int spatial) { + __shared__ float buffer[THREADS_PER_BLOCK]; + int tid = threadIdx.x; + int c = blockIdx.x; + buffer[tid] = 0; + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = (i / spatial) * channels * spatial + c * spatial + i % spatial; + float td = static_cast(input[index]) - mean[c]; + buffer[tid] += td * td; + } + __syncthreads(); + for (int s = blockDim.x / 2; s > 0; s >>= 1) { + if (tid < s) { + buffer[tid] += buffer[tid + s]; + } + __syncthreads(); + } + int total = num * spatial; + if (tid == 0) { + var[c] = buffer[0] / total; + } +} + +template +__global__ void sync_bn_forward_output_cuda_kernel( + const T *input, const float *mean, const float *var, float *running_mean, + float *running_var, const float *weight, const float *bias, float *norm, + float *std, T *output, int num, int channels, int spatial, float eps, + float momentum, int group_size) { + int tid = threadIdx.x; + int c = blockIdx.x; + float mean_value = mean[c]; + float std_value = sqrt(var[c] + eps); + + if (weight != nullptr) { + float weight_value = weight[c]; + float bias_value = bias[c]; + if (norm != nullptr) { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + norm[index] = (input[index] - mean_value) / std_value; + output[index] = norm[index] * weight_value + bias_value; + } + } else { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + output[index] = + (input[index] - mean_value) / std_value * weight_value + bias_value; + } + } + } else { + if (norm != nullptr) { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + output[index] = norm[index] = (input[index] - mean_value) / std_value; + } + } else { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + output[index] = (input[index] - mean_value) / std_value; + } + } + } + if (tid == 0) { + if (std != nullptr) std[c] = std_value; + if (running_mean != nullptr) { + running_mean[c] = + momentum * mean_value + (1 - momentum) * running_mean[c]; + int count = num * spatial * group_size; + float var_unbias = count > 1 ? var[c] * count / (count - 1) : var[c]; + running_var[c] = momentum * var_unbias + (1 - momentum) * running_var[c]; + } + } +} + +template <> +__global__ void sync_bn_forward_output_cuda_kernel( + const phalf *input, const float *mean, const float *var, + float *running_mean, float *running_var, const float *weight, + const float *bias, float *norm, float *std, phalf *output, int num, + int channels, int spatial, float eps, float momentum, int group_size) { + int tid = threadIdx.x; + int c = blockIdx.x; + float mean_value = mean[c]; + float std_value = sqrt(var[c] + eps); + if (weight != nullptr) { + float weight_value = weight[c]; + float bias_value = bias[c]; + if (norm != nullptr) { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + norm[index] = + (static_cast(input[index]) - mean_value) / std_value; + output[index] = + static_cast(norm[index] * weight_value + bias_value); + } + } else { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + output[index] = + static_cast((static_cast(input[index]) - mean_value) / + std_value * weight_value + + bias_value); + } + } + } else { + if (norm != nullptr) { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + norm[index] = + (static_cast(input[index]) - mean_value) / std_value; + output[index] = static_cast(norm[index]); + } + } else { + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = + (i / spatial) * channels * spatial + c * spatial + i % spatial; + output[index] = static_cast( + (static_cast(input[index]) - mean_value) / std_value); + } + } + } + if (tid == 0) { + if (std != nullptr) std[c] = std_value; + if (running_mean != nullptr) { + running_mean[c] = + momentum * mean_value + (1 - momentum) * running_mean[c]; + int count = num * spatial * group_size; + float var_unbias = count > 1 ? var[c] * count / (count - 1) : var[c]; + running_var[c] = momentum * var_unbias + (1 - momentum) * running_var[c]; + } + } +} + +template +__global__ void sync_bn_backward_param_cuda_kernel(const T *grad_output, + const float *norm, + float *grad_weight, + float *grad_bias, int num, + int channels, int spatial) { + __shared__ float buffer1[THREADS_PER_BLOCK]; + __shared__ float buffer2[THREADS_PER_BLOCK]; + + int tid = threadIdx.x; + int c = blockIdx.x; + buffer1[tid] = buffer2[tid] = 0; + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = (i / spatial) * channels * spatial + c * spatial + i % spatial; + buffer1[tid] += grad_output[index] * norm[index]; + buffer2[tid] += grad_output[index]; + } + __syncthreads(); + + for (int s = blockDim.x / 2; s > 0; s >>= 1) { + if (tid < s) { + buffer1[tid] += buffer1[tid + s]; + buffer2[tid] += buffer2[tid + s]; + } + __syncthreads(); + } + if (tid == 0) { + grad_weight[c] = buffer1[0]; + grad_bias[c] = buffer2[0]; + } +} + +template <> +__global__ void sync_bn_backward_param_cuda_kernel(const phalf *grad_output, + const float *norm, + float *grad_weight, + float *grad_bias, int num, + int channels, int spatial) { + __shared__ float buffer1[THREADS_PER_BLOCK]; + __shared__ float buffer2[THREADS_PER_BLOCK]; + + int tid = threadIdx.x; + int c = blockIdx.x; + buffer1[tid] = buffer2[tid] = 0; + for (int i = tid; i < num * spatial; i += blockDim.x) { + int index = (i / spatial) * channels * spatial + c * spatial + i % spatial; + buffer1[tid] += static_cast(grad_output[index]) * norm[index]; + buffer2[tid] += static_cast(grad_output[index]); + } + __syncthreads(); + + for (int s = blockDim.x / 2; s > 0; s >>= 1) { + if (tid < s) { + buffer1[tid] += buffer1[tid + s]; + buffer2[tid] += buffer2[tid + s]; + } + __syncthreads(); + } + if (tid == 0) { + grad_weight[c] = buffer1[0]; + grad_bias[c] = buffer2[0]; + } +} + +template +__global__ void sync_bn_backward_data_cuda_kernel( + int output_size, const T *grad_output, const float *weight, + const float *grad_weight, const float *grad_bias, const float *norm, + const float *std, T *grad_input, int num, int channels, int spatial) { + int factor = num * spatial; + CUDA_1D_KERNEL_LOOP(index, output_size) { + int c = (index / spatial) % channels; + grad_input[index] = + weight[c] * + (grad_output[index] - + (grad_weight[c] * norm[index] + grad_bias[c]) / factor) / + std[c]; + } +} + +template <> +__global__ void sync_bn_backward_data_cuda_kernel( + int output_size, const phalf *grad_output, const float *weight, + const float *grad_weight, const float *grad_bias, const float *norm, + const float *std, phalf *grad_input, int num, int channels, int spatial) { + int factor = num * spatial; + CUDA_1D_KERNEL_LOOP(index, output_size) { + int c = (index / spatial) % channels; + grad_input[index] = static_cast( + weight[c] * + (static_cast(grad_output[index]) - + (grad_weight[c] * norm[index] + grad_bias[c]) / factor) / + std[c]); + } +} + +#endif // SYNCBN_CUDA_KERNEL_CUH diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_cpp_helper.hpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_cpp_helper.hpp new file mode 100644 index 0000000000000000000000000000000000000000..f68e8740561ef833c09e1ba9f999922f5d04bce5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_cpp_helper.hpp @@ -0,0 +1,27 @@ +#ifndef PYTORCH_CPP_HELPER +#define PYTORCH_CPP_HELPER +#include + +#include + +using namespace at; + +#define CHECK_CUDA(x) \ + TORCH_CHECK(x.device().is_cuda(), #x " must be a CUDA tensor") +#define CHECK_MLU(x) \ + TORCH_CHECK(x.device().type() == at::kMLU, #x " must be a MLU tensor") +#define CHECK_CPU(x) \ + TORCH_CHECK(x.device().type() == at::kCPU, #x " must be a CPU tensor") +#define CHECK_CONTIGUOUS(x) \ + TORCH_CHECK(x.is_contiguous(), #x " must be contiguous") +#define CHECK_CUDA_INPUT(x) \ + CHECK_CUDA(x); \ + CHECK_CONTIGUOUS(x) +#define CHECK_MLU_INPUT(x) \ + CHECK_MLU(x); \ + CHECK_CONTIGUOUS(x) +#define CHECK_CPU_INPUT(x) \ + CHECK_CPU(x); \ + CHECK_CONTIGUOUS(x) + +#endif // PYTORCH_CPP_HELPER diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_cuda_helper.hpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_cuda_helper.hpp new file mode 100644 index 0000000000000000000000000000000000000000..9869b535f8a1de758b0c35612dbd4ac2a1701ad9 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_cuda_helper.hpp @@ -0,0 +1,19 @@ +#ifndef PYTORCH_CUDA_HELPER +#define PYTORCH_CUDA_HELPER + +#include +#include +#include + +#include +#include + +#include "common_cuda_helper.hpp" + +using at::Half; +using at::Tensor; +using phalf = at::Half; + +#define __PHALF(x) (x) + +#endif // PYTORCH_CUDA_HELPER diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_device_registry.hpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_device_registry.hpp new file mode 100644 index 0000000000000000000000000000000000000000..2a32b7270c3521f960394af7d18cbbd03ba50df1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/common/pytorch_device_registry.hpp @@ -0,0 +1,141 @@ +#ifndef PYTORCH_DEVICE_REGISTRY_H +#define PYTORCH_DEVICE_REGISTRY_H + +// Using is recommended in the official documentation in +// https://pytorch.org/tutorials/advanced/cpp_extension.html#writing-the-c-op. +// However, we use for compatibility with CUDA 9.0 +// Read https://github.com/pytorch/extension-cpp/issues/35 for more details. +#include + +#include +#include +#include +#include + +inline std::string GetDeviceStr(const at::Device& device) { + std::string str = DeviceTypeName(device.type(), true); + if (device.has_index()) { + str.push_back(':'); + str.append(std::to_string(device.index())); + } + return str; +} + +// Registry +template +class DeviceRegistry; + +template +class DeviceRegistry { + public: + using FunctionType = Ret (*)(Args...); + static const int MAX_DEVICE_TYPES = + int8_t(at::DeviceType::COMPILE_TIME_MAX_DEVICE_TYPES); + + void Register(at::DeviceType device, FunctionType function) { + funcs_[int8_t(device)] = function; + } + + FunctionType Find(at::DeviceType device) const { + return funcs_[int8_t(device)]; + } + + static DeviceRegistry& instance() { + static DeviceRegistry inst; + return inst; + } + + private: + DeviceRegistry() { + for (size_t i = 0; i < MAX_DEVICE_TYPES; ++i) { + funcs_[i] = nullptr; + } + }; + FunctionType funcs_[MAX_DEVICE_TYPES]; +}; + +// get device of first tensor param + +template , at::Tensor>::value, + bool> = true> +at::Device GetFirstTensorDevice(T&& t, Args&&... args) { + return std::forward(t).device(); +} +template , at::Tensor>::value, + bool> = true> +at::Device GetFirstTensorDevice(T&& t, Args&&... args) { + return GetFirstTensorDevice(std::forward(args)...); +} + +// check device consistency + +inline std::pair CheckDeviceConsistency( + const at::Device& device, int index) { + return {index, device}; +} + +template , at::Tensor>::value, + bool> = true> +std::pair CheckDeviceConsistency(const at::Device& device, + int index, T&& t, + Args&&... args); + +template , at::Tensor>::value, + bool> = true> +std::pair CheckDeviceConsistency(const at::Device& device, + int index, T&& t, + Args&&... args) { + auto new_device = std::forward(t).device(); + if (new_device.type() != device.type() || + new_device.index() != device.index()) { + return {index, new_device}; + } + return CheckDeviceConsistency(device, index + 1, std::forward(args)...); +} + +template < + typename T, typename... Args, + std::enable_if_t, at::Tensor>::value, bool>> +std::pair CheckDeviceConsistency(const at::Device& device, + int index, T&& t, + Args&&... args) { + return CheckDeviceConsistency(device, index + 1, std::forward(args)...); +} + +// dispatch + +template +auto Dispatch(const R& registry, const char* name, Args&&... args) { + auto device = GetFirstTensorDevice(std::forward(args)...); + auto inconsist = + CheckDeviceConsistency(device, 0, std::forward(args)...); + TORCH_CHECK(inconsist.first >= int(sizeof...(Args)), name, ": at param ", + inconsist.first, + ", inconsistent device: ", GetDeviceStr(inconsist.second).c_str(), + " vs ", GetDeviceStr(device).c_str(), "\n") + auto f_ptr = registry.Find(device.type()); + TORCH_CHECK(f_ptr != nullptr, name, ": implementation for device ", + GetDeviceStr(device).c_str(), " not found.\n") + return f_ptr(std::forward(args)...); +} + +// helper macro + +#define DEVICE_REGISTRY(key) DeviceRegistry::instance() + +#define REGISTER_DEVICE_IMPL(key, device, value) \ + struct key##_##device##_registerer { \ + key##_##device##_registerer() { \ + DEVICE_REGISTRY(key).Register(at::k##device, value); \ + } \ + }; \ + static key##_##device##_registerer _##key##_##device##_registerer; + +#define DISPATCH_DEVICE_IMPL(key, ...) \ + Dispatch(DEVICE_REGISTRY(key), #key, __VA_ARGS__) + +#endif // PYTORCH_DEVICE_REGISTRY diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/cudabind.cpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/cudabind.cpp new file mode 100644 index 0000000000000000000000000000000000000000..dc376bb51554be508e9b06b774bdb60bdc139f7f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/cudabind.cpp @@ -0,0 +1,210 @@ +#include "pytorch_cpp_helper.hpp" +#include "pytorch_device_registry.hpp" + +void SigmoidFocalLossForwardCUDAKernelLauncher(Tensor input, Tensor target, + Tensor weight, Tensor output, + const float gamma, + const float alpha); + +void SigmoidFocalLossBackwardCUDAKernelLauncher(Tensor input, Tensor target, + Tensor weight, + Tensor grad_input, + const float gamma, + const float alpha); + +void SoftmaxFocalLossForwardCUDAKernelLauncher(Tensor softmax, Tensor target, + Tensor weight, Tensor output, + const float gamma, + const float alpha); + +void SoftmaxFocalLossBackwardCUDAKernelLauncher(Tensor softmax, Tensor target, + Tensor weight, Tensor buff, + Tensor grad_input, + const float gamma, + const float alpha); + +void sigmoid_focal_loss_forward_cuda(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha) { + SigmoidFocalLossForwardCUDAKernelLauncher(input, target, weight, output, + gamma, alpha); +} + +void sigmoid_focal_loss_backward_cuda(Tensor input, Tensor target, + Tensor weight, Tensor grad_input, + float gamma, float alpha) { + SigmoidFocalLossBackwardCUDAKernelLauncher(input, target, weight, grad_input, + gamma, alpha); +} + +void softmax_focal_loss_forward_cuda(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha) { + SoftmaxFocalLossForwardCUDAKernelLauncher(input, target, weight, output, + gamma, alpha); +} + +void softmax_focal_loss_backward_cuda(Tensor input, Tensor target, + Tensor weight, Tensor buff, + Tensor grad_input, float gamma, + float alpha) { + SoftmaxFocalLossBackwardCUDAKernelLauncher(input, target, weight, buff, + grad_input, gamma, alpha); +} + +void sigmoid_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha); + +void sigmoid_focal_loss_backward_impl(Tensor input, Tensor target, + Tensor weight, Tensor grad_input, + float gamma, float alpha); + +void softmax_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha); + +void softmax_focal_loss_backward_impl(Tensor input, Tensor target, + Tensor weight, Tensor buff, + Tensor grad_input, float gamma, + float alpha); + +REGISTER_DEVICE_IMPL(sigmoid_focal_loss_forward_impl, CUDA, + sigmoid_focal_loss_forward_cuda); +REGISTER_DEVICE_IMPL(sigmoid_focal_loss_backward_impl, CUDA, + sigmoid_focal_loss_backward_cuda); +REGISTER_DEVICE_IMPL(softmax_focal_loss_forward_impl, CUDA, + softmax_focal_loss_forward_cuda); +REGISTER_DEVICE_IMPL(softmax_focal_loss_backward_impl, CUDA, + softmax_focal_loss_backward_cuda); + +void SyncBNForwardMeanCUDAKernelLauncher(const Tensor input, Tensor mean); + +void SyncBNForwardVarCUDAKernelLauncher(const Tensor input, const Tensor mean, + Tensor var); + +void SyncBNForwardOutputCUDAKernelLauncher( + const Tensor input, const Tensor mean, const Tensor var, + Tensor running_mean, Tensor running_var, const Tensor weight, + const Tensor bias, Tensor norm, Tensor std, Tensor output, float eps, + float momentum, int group_size); + +void SyncBNBackwardParamCUDAKernelLauncher(const Tensor grad_output, + const Tensor norm, + Tensor grad_weight, + Tensor grad_bias); + +void SyncBNBackwardDataCUDAKernelLauncher(const Tensor grad_output, + const Tensor weight, + const Tensor grad_weight, + const Tensor grad_bias, + const Tensor norm, const Tensor std, + Tensor grad_input); + +void sync_bn_forward_mean_cuda(const Tensor input, Tensor mean) { + SyncBNForwardMeanCUDAKernelLauncher(input, mean); +} + +void sync_bn_forward_var_cuda(const Tensor input, const Tensor mean, + Tensor var) { + SyncBNForwardVarCUDAKernelLauncher(input, mean, var); +} + +void sync_bn_forward_output_cuda(const Tensor input, const Tensor mean, + const Tensor var, Tensor running_mean, + Tensor running_var, const Tensor weight, + const Tensor bias, Tensor norm, Tensor std, + Tensor output, float eps, float momentum, + int group_size) { + SyncBNForwardOutputCUDAKernelLauncher(input, mean, var, running_mean, + running_var, weight, bias, norm, std, + output, eps, momentum, group_size); +} + +void sync_bn_backward_param_cuda(const Tensor grad_output, const Tensor norm, + Tensor grad_weight, Tensor grad_bias) { + SyncBNBackwardParamCUDAKernelLauncher(grad_output, norm, grad_weight, + grad_bias); +} + +void sync_bn_backward_data_cuda(const Tensor grad_output, const Tensor weight, + const Tensor grad_weight, + const Tensor grad_bias, const Tensor norm, + const Tensor std, Tensor grad_input) { + SyncBNBackwardDataCUDAKernelLauncher(grad_output, weight, grad_weight, + grad_bias, norm, std, grad_input); +} + +void sync_bn_forward_mean_impl(const Tensor input, Tensor mean); + +void sync_bn_forward_var_impl(const Tensor input, const Tensor mean, + Tensor var); + +void sync_bn_forward_output_impl(const Tensor input, const Tensor mean, + const Tensor var, Tensor running_mean, + Tensor running_var, const Tensor weight, + const Tensor bias, Tensor norm, Tensor std, + Tensor output, float eps, float momentum, + int group_size); + +void sync_bn_backward_param_impl(const Tensor grad_output, const Tensor norm, + Tensor grad_weight, Tensor grad_bias); + +void sync_bn_backward_data_impl(const Tensor grad_output, const Tensor weight, + const Tensor grad_weight, + const Tensor grad_bias, const Tensor norm, + const Tensor std, Tensor grad_input); + +REGISTER_DEVICE_IMPL(sync_bn_forward_mean_impl, CUDA, + sync_bn_forward_mean_cuda); +REGISTER_DEVICE_IMPL(sync_bn_forward_var_impl, CUDA, sync_bn_forward_var_cuda); +REGISTER_DEVICE_IMPL(sync_bn_forward_output_impl, CUDA, + sync_bn_forward_output_cuda); +REGISTER_DEVICE_IMPL(sync_bn_backward_param_impl, CUDA, + sync_bn_backward_param_cuda); +REGISTER_DEVICE_IMPL(sync_bn_backward_data_impl, CUDA, + sync_bn_backward_data_cuda); + + + +void PSAMaskForwardCUDAKernelLauncher(const int psa_type, const Tensor input, + Tensor output, const int num_, + const int h_feature, const int w_feature, + const int h_mask, const int w_mask, + const int half_h_mask, + const int half_w_mask); + +void PSAMaskBackwardCUDAKernelLauncher( + const int psa_type, const Tensor grad_output, Tensor grad_input, + const int num_, const int h_feature, const int w_feature, const int h_mask, + const int w_mask, const int half_h_mask, const int half_w_mask); + +void psamask_forward_cuda(const int psa_type, const Tensor input, Tensor output, + const int num_, const int h_feature, + const int w_feature, const int h_mask, + const int w_mask, const int half_h_mask, + const int half_w_mask) { + PSAMaskForwardCUDAKernelLauncher(psa_type, input, output, num_, h_feature, + w_feature, h_mask, w_mask, half_h_mask, + half_w_mask); +} + +void psamask_backward_cuda(const int psa_type, const Tensor grad_output, + Tensor grad_input, const int num_, + const int h_feature, const int w_feature, + const int h_mask, const int w_mask, + const int half_h_mask, const int half_w_mask) { + PSAMaskBackwardCUDAKernelLauncher(psa_type, grad_output, grad_input, num_, + h_feature, w_feature, h_mask, w_mask, + half_h_mask, half_w_mask); +} + +void psamask_forward_impl(const int psa_type, const Tensor input, Tensor output, + const int num_, const int h_feature, + const int w_feature, const int h_mask, + const int w_mask, const int half_h_mask, + const int half_w_mask); + +void psamask_backward_impl(const int psa_type, const Tensor grad_output, + Tensor grad_input, const int num_, + const int h_feature, const int w_feature, + const int h_mask, const int w_mask, + const int half_h_mask, const int half_w_mask); +REGISTER_DEVICE_IMPL(psamask_forward_impl, CUDA, psamask_forward_cuda); +REGISTER_DEVICE_IMPL(psamask_backward_impl, CUDA, psamask_backward_cuda); diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/focal_loss_cuda.cu b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/focal_loss_cuda.cu new file mode 100644 index 0000000000000000000000000000000000000000..cb899f954fd969e57a23d5723bf2f9c49b35a853 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/focal_loss_cuda.cu @@ -0,0 +1,111 @@ +// Copyright (c) OpenMMLab. All rights reserved +#include "pytorch_cuda_helper.hpp" +#include "sigmoid_focal_loss_cuda_kernel.cuh" +#include "softmax_focal_loss_cuda_kernel.cuh" + +void SigmoidFocalLossForwardCUDAKernelLauncher(Tensor input, Tensor target, + Tensor weight, Tensor output, + const float gamma, + const float alpha) { + int output_size = output.numel(); + int num_classes = input.size(1); + AT_ASSERTM(target.max().item() <= (int64_t)num_classes, + "target label should smaller or equal than num classes"); + at::cuda::CUDAGuard device_guard(input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + input.scalar_type(), "sigmoid_focal_loss_forward_cuda_kernel", [&] { + sigmoid_focal_loss_forward_cuda_kernel + <<>>( + output_size, input.data_ptr(), + target.data_ptr(), weight.data_ptr(), + output.data_ptr(), gamma, alpha, num_classes); + }); + + AT_CUDA_CHECK(cudaGetLastError()); +} + +void SigmoidFocalLossBackwardCUDAKernelLauncher(Tensor input, Tensor target, + Tensor weight, + Tensor grad_input, + const float gamma, + const float alpha) { + int output_size = grad_input.numel(); + int num_classes = input.size(1); + + at::cuda::CUDAGuard device_guard(grad_input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + input.scalar_type(), "sigmoid_focal_loss_backward_cuda_kernel", [&] { + sigmoid_focal_loss_backward_cuda_kernel + <<>>( + output_size, input.data_ptr(), + target.data_ptr(), weight.data_ptr(), + grad_input.data_ptr(), gamma, alpha, num_classes); + }); + + AT_CUDA_CHECK(cudaGetLastError()); +} + +void SoftmaxFocalLossForwardCUDAKernelLauncher(Tensor softmax, Tensor target, + Tensor weight, Tensor output, + const float gamma, + const float alpha) { + int output_size = output.numel(); + int num_classes = softmax.size(1); + + AT_ASSERTM(target.max().item() <= (int64_t)num_classes, + "target label should smaller or equal than num classes"); + at::cuda::CUDAGuard device_guard(softmax.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + softmax.scalar_type(), "softmax_focal_loss_forward_cuda_kernel", [&] { + softmax_focal_loss_forward_cuda_kernel + <<>>( + output_size, softmax.data_ptr(), + target.data_ptr(), weight.data_ptr(), + output.data_ptr(), gamma, alpha, num_classes); + }); + + AT_CUDA_CHECK(cudaGetLastError()); +} + +void SoftmaxFocalLossBackwardCUDAKernelLauncher(Tensor softmax, Tensor target, + Tensor weight, Tensor buff, + Tensor grad_input, + const float gamma, + const float alpha) { + int num_classes = softmax.size(1); + + int output_size = buff.numel(); + at::cuda::CUDAGuard device_guard(grad_input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + grad_input.scalar_type(), + "softmax_focal_loss_backward_cuda1_" + "kernel", + [&] { + softmax_focal_loss_backward_cuda1_kernel + <<>>( + output_size, softmax.data_ptr(), + target.data_ptr(), weight.data_ptr(), + buff.data_ptr(), gamma, alpha, num_classes); + }); + + AT_CUDA_CHECK(cudaGetLastError()); + + output_size = grad_input.numel(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + grad_input.scalar_type(), + "softmax_focal_loss_backward_cuda2_" + "kernel", + [&] { + softmax_focal_loss_backward_cuda2_kernel + <<>>( + output_size, softmax.data_ptr(), + target.data_ptr(), buff.data_ptr(), + grad_input.data_ptr(), num_classes); + }); + + AT_CUDA_CHECK(cudaGetLastError()); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/psamask_cuda.cu b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/psamask_cuda.cu new file mode 100644 index 0000000000000000000000000000000000000000..a0bdfa60c2d3ba75d089d0bfa44648821aaf4fed --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/psamask_cuda.cu @@ -0,0 +1,60 @@ +// Copyright (c) OpenMMLab. All rights reserved +// Modified from +// https://github.com/hszhao/semseg/blob/master/lib/psa/src + +#include + +#include "psamask_cuda_kernel.cuh" +#include "pytorch_cuda_helper.hpp" + +void PSAMaskForwardCUDAKernelLauncher(const int psa_type, const Tensor input, + Tensor output, const int num_, + const int h_feature, const int w_feature, + const int h_mask, const int w_mask, + const int half_h_mask, + const int half_w_mask) { + int nthreads = num_ * h_feature * w_feature; + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + if (psa_type == 0) + AT_DISPATCH_FLOATING_TYPES( + input.scalar_type(), "psamask_collect_forward_cuda", [&] { + psamask_collect_forward_cuda<<>>( + nthreads, h_feature, w_feature, h_mask, w_mask, half_h_mask, + half_w_mask, input.data_ptr(), + output.data_ptr()); + }); + else + AT_DISPATCH_FLOATING_TYPES( + input.scalar_type(), "psamask_distribute_forward_cuda", [&] { + psamask_distribute_forward_cuda + <<>>( + nthreads, h_feature, w_feature, h_mask, w_mask, half_h_mask, + half_w_mask, input.data_ptr(), + output.data_ptr()); + }); +} + +void PSAMaskBackwardCUDAKernelLauncher( + const int psa_type, const Tensor grad_output, Tensor grad_input, + const int num_, const int h_feature, const int w_feature, const int h_mask, + const int w_mask, const int half_h_mask, const int half_w_mask) { + int nthreads = num_ * h_feature * w_feature; + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + if (psa_type == 0) + AT_DISPATCH_FLOATING_TYPES( + grad_input.scalar_type(), "psamask_collect_backward_cuda", [&] { + psamask_collect_backward_cuda<<>>( + nthreads, h_feature, w_feature, h_mask, w_mask, half_h_mask, + half_w_mask, grad_output.data_ptr(), + grad_input.data_ptr()); + }); + else + AT_DISPATCH_FLOATING_TYPES( + grad_input.scalar_type(), "psamask_distribute_backward_cuda", [&] { + psamask_distribute_backward_cuda + <<>>( + nthreads, h_feature, w_feature, h_mask, w_mask, half_h_mask, + half_w_mask, grad_output.data_ptr(), + grad_input.data_ptr()); + }); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/sync_bn_cuda.cu b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/sync_bn_cuda.cu new file mode 100644 index 0000000000000000000000000000000000000000..657c81701b7c114af700c4f8cf37094c705b9a94 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/cuda/sync_bn_cuda.cu @@ -0,0 +1,110 @@ +// Copyright (c) OpenMMLab. All rights reserved +#include "pytorch_cuda_helper.hpp" +#include "sync_bn_cuda_kernel.cuh" + +void SyncBNForwardMeanCUDAKernelLauncher(const Tensor input, Tensor mean) { + int num = input.size(0); + int channels = input.size(1); + int spatial = input.size(2); + + at::cuda::CUDAGuard device_guard(input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + input.scalar_type(), "sync_bn_forward_mean_cuda_kernel", [&] { + sync_bn_forward_mean_cuda_kernel + <<>>( + input.data_ptr(), mean.data_ptr(), num, + channels, spatial); + }); + AT_CUDA_CHECK(cudaGetLastError()); +} + +void SyncBNForwardVarCUDAKernelLauncher(const Tensor input, const Tensor mean, + Tensor var) { + int num = input.size(0); + int channels = input.size(1); + int spatial = input.size(2); + + at::cuda::CUDAGuard device_guard(input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + input.scalar_type(), "sync_bn_forward_mean_cuda_kernel", [&] { + sync_bn_forward_var_cuda_kernel + <<>>( + input.data_ptr(), mean.data_ptr(), + var.data_ptr(), num, channels, spatial); + }); + AT_CUDA_CHECK(cudaGetLastError()); +} + +void SyncBNForwardOutputCUDAKernelLauncher( + const Tensor input, const Tensor mean, const Tensor var, + Tensor running_mean, Tensor running_var, const Tensor weight, + const Tensor bias, Tensor norm, Tensor std, Tensor output, float eps, + float momentum, int group_size) { + int num = input.size(0); + int channels = input.size(1); + int spatial = input.size(2); + + at::cuda::CUDAGuard device_guard(input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + input.scalar_type(), "sync_bn_forward_mean_cuda_kernel", [&] { + sync_bn_forward_output_cuda_kernel + <<>>( + input.data_ptr(), mean.data_ptr(), + var.data_ptr(), running_mean.data_ptr(), + running_var.data_ptr(), weight.data_ptr(), + bias.data_ptr(), norm.data_ptr(), + std.data_ptr(), output.data_ptr(), num, + channels, spatial, eps, momentum, group_size); + }); + AT_CUDA_CHECK(cudaGetLastError()); +} + +void SyncBNBackwardParamCUDAKernelLauncher(const Tensor grad_output, + const Tensor norm, + Tensor grad_weight, + Tensor grad_bias) { + int num = grad_output.size(0); + int channels = grad_output.size(1); + int spatial = grad_output.size(2); + + at::cuda::CUDAGuard device_guard(grad_output.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + grad_output.scalar_type(), "sync_bn_backward_param_cuda_kernel", [&] { + sync_bn_backward_param_cuda_kernel + <<>>( + grad_output.data_ptr(), norm.data_ptr(), + grad_weight.data_ptr(), grad_bias.data_ptr(), num, + channels, spatial); + }); + AT_CUDA_CHECK(cudaGetLastError()); +} + +void SyncBNBackwardDataCUDAKernelLauncher(const Tensor grad_output, + const Tensor weight, + const Tensor grad_weight, + const Tensor grad_bias, + const Tensor norm, const Tensor std, + Tensor grad_input) { + int output_size = grad_input.numel(); + int num = grad_input.size(0); + int channels = grad_input.size(1); + int spatial = grad_input.size(2); + + at::cuda::CUDAGuard device_guard(grad_input.device()); + cudaStream_t stream = at::cuda::getCurrentCUDAStream(); + AT_DISPATCH_FLOATING_TYPES_AND_HALF( + grad_output.scalar_type(), "sync_bn_backward_data_cuda_kernel", [&] { + sync_bn_backward_data_cuda_kernel + <<>>( + output_size, grad_output.data_ptr(), + weight.data_ptr(), grad_weight.data_ptr(), + grad_bias.data_ptr(), norm.data_ptr(), + std.data_ptr(), grad_input.data_ptr(), num, + channels, spatial); + }); + AT_CUDA_CHECK(cudaGetLastError()); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/focal_loss.cpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/focal_loss.cpp new file mode 100644 index 0000000000000000000000000000000000000000..ed0e2186532d9d6d909f76d653283bbdc29eac11 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/focal_loss.cpp @@ -0,0 +1,53 @@ +// Copyright (c) OpenMMLab. All rights reserved +#include "pytorch_cpp_helper.hpp" +#include "pytorch_device_registry.hpp" + +void sigmoid_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha) { + DISPATCH_DEVICE_IMPL(sigmoid_focal_loss_forward_impl, input, target, weight, + output, gamma, alpha); +} + +void sigmoid_focal_loss_backward_impl(Tensor input, Tensor target, + Tensor weight, Tensor grad_input, + float gamma, float alpha) { + DISPATCH_DEVICE_IMPL(sigmoid_focal_loss_backward_impl, input, target, weight, + grad_input, gamma, alpha); +} + +void softmax_focal_loss_forward_impl(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha) { + DISPATCH_DEVICE_IMPL(softmax_focal_loss_forward_impl, input, target, weight, + output, gamma, alpha); +} + +void softmax_focal_loss_backward_impl(Tensor input, Tensor target, + Tensor weight, Tensor buff, + Tensor grad_input, float gamma, + float alpha) { + DISPATCH_DEVICE_IMPL(softmax_focal_loss_backward_impl, input, target, weight, + buff, grad_input, gamma, alpha); +} + +void sigmoid_focal_loss_forward(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha) { + sigmoid_focal_loss_forward_impl(input, target, weight, output, gamma, alpha); +} + +void sigmoid_focal_loss_backward(Tensor input, Tensor target, Tensor weight, + Tensor grad_input, float gamma, float alpha) { + sigmoid_focal_loss_backward_impl(input, target, weight, grad_input, gamma, + alpha); +} + +void softmax_focal_loss_forward(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha) { + softmax_focal_loss_forward_impl(input, target, weight, output, gamma, alpha); +} + +void softmax_focal_loss_backward(Tensor input, Tensor target, Tensor weight, + Tensor buff, Tensor grad_input, float gamma, + float alpha) { + softmax_focal_loss_backward_impl(input, target, weight, buff, grad_input, + gamma, alpha); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/info.cpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/info.cpp new file mode 100644 index 0000000000000000000000000000000000000000..a08d227d4c6e94f0dabd8cebab7bf2d77b9df4b9 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/info.cpp @@ -0,0 +1,56 @@ +// Copyright (c) OpenMMLab. All rights reserved +// modified from +// https://github.com/facebookresearch/detectron2/blob/master/detectron2/layers/csrc/vision.cpp +#include "pytorch_cpp_helper.hpp" + +#ifdef MMCV_WITH_CUDA +#ifndef HIP_DIFF +#include +int get_cudart_version() { return CUDART_VERSION; } +#endif +#endif + +std::string get_compiling_cuda_version() { +#ifdef MMCV_WITH_CUDA +#ifndef HIP_DIFF + std::ostringstream oss; + // copied from + // https://github.com/pytorch/pytorch/blob/master/aten/src/ATen/cuda/detail/CUDAHooks.cpp#L231 + auto printCudaStyleVersion = [&](int v) { + oss << (v / 1000) << "." << (v / 10 % 100); + if (v % 10 != 0) { + oss << "." << (v % 10); + } + }; + printCudaStyleVersion(get_cudart_version()); + return oss.str(); +#else + return std::string("rocm not available"); +#endif +#else + return std::string("not available"); +#endif +} + +// similar to +// https://github.com/pytorch/pytorch/blob/master/aten/src/ATen/Version.cpp +std::string get_compiler_version() { + std::ostringstream ss; +#if defined(__GNUC__) +#ifndef __clang__ + { ss << "GCC " << __GNUC__ << "." << __GNUC_MINOR__; } +#endif +#endif + +#if defined(__clang_major__) + { + ss << "clang " << __clang_major__ << "." << __clang_minor__ << "." + << __clang_patchlevel__; + } +#endif + +#if defined(_MSC_VER) + { ss << "MSVC " << _MSC_FULL_VER; } +#endif + return ss.str(); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/psamask.cpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/psamask.cpp new file mode 100644 index 0000000000000000000000000000000000000000..6064c9ba5fd7ec9bcfef22b3abcc65ef50106d67 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/psamask.cpp @@ -0,0 +1,41 @@ +// Copyright (c) OpenMMLab. All rights reserved +// Modified from +// https://github.com/hszhao/semseg/blob/master/lib/psa/src +#include "pytorch_cpp_helper.hpp" +#include "pytorch_device_registry.hpp" + +void psamask_forward_impl(const int psa_type, const Tensor input, Tensor output, + const int num_, const int h_feature, + const int w_feature, const int h_mask, + const int w_mask, const int half_h_mask, + const int half_w_mask) { + DISPATCH_DEVICE_IMPL(psamask_forward_impl, psa_type, input, output, num_, + h_feature, w_feature, h_mask, w_mask, half_h_mask, + half_w_mask); +} + +void psamask_backward_impl(const int psa_type, const Tensor grad_output, + Tensor grad_input, const int num_, + const int h_feature, const int w_feature, + const int h_mask, const int w_mask, + const int half_h_mask, const int half_w_mask) { + DISPATCH_DEVICE_IMPL(psamask_backward_impl, psa_type, grad_output, grad_input, + num_, h_feature, w_feature, h_mask, w_mask, half_h_mask, + half_w_mask); +} + +void psamask_forward(const Tensor input, Tensor output, const int psa_type, + const int num_, const int h_feature, const int w_feature, + const int h_mask, const int w_mask, const int half_h_mask, + const int half_w_mask) { + psamask_forward_impl(psa_type, input, output, num_, h_feature, w_feature, + h_mask, w_mask, half_h_mask, half_w_mask); +} + +void psamask_backward(Tensor grad_output, const Tensor grad_input, + const int psa_type, const int num_, const int h_feature, + const int w_feature, const int h_mask, const int w_mask, + const int half_h_mask, const int half_w_mask) { + psamask_backward_impl(psa_type, grad_output, grad_input, num_, h_feature, + w_feature, h_mask, w_mask, half_h_mask, half_w_mask); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/pybind.cpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/pybind.cpp new file mode 100644 index 0000000000000000000000000000000000000000..1e99174d804fcc84375a516f58f998e395091917 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/pybind.cpp @@ -0,0 +1,106 @@ +// Copyright (c) OpenMMLab. All rights reserved +#include + +#include "pytorch_cpp_helper.hpp" + +std::string get_compiler_version(); +std::string get_compiling_cuda_version(); + +void sigmoid_focal_loss_forward(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha); + +void sigmoid_focal_loss_backward(Tensor input, Tensor target, Tensor weight, + Tensor grad_input, float gamma, float alpha); + +void softmax_focal_loss_forward(Tensor input, Tensor target, Tensor weight, + Tensor output, float gamma, float alpha); + +void softmax_focal_loss_backward(Tensor input, Tensor target, Tensor weight, + Tensor buff, Tensor grad_input, float gamma, + float alpha); + +void sync_bn_forward_mean(const Tensor input, Tensor mean); + +void sync_bn_forward_var(const Tensor input, const Tensor mean, Tensor var); + +void sync_bn_forward_output(const Tensor input, const Tensor mean, + const Tensor var, const Tensor weight, + const Tensor bias, Tensor running_mean, + Tensor running_var, Tensor norm, Tensor std, + Tensor output, float eps, float momentum, + int group_size); + +void sync_bn_backward_param(const Tensor grad_output, const Tensor norm, + Tensor grad_weight, Tensor grad_bias); + +void sync_bn_backward_data(const Tensor grad_output, const Tensor weight, + const Tensor grad_weight, const Tensor grad_bias, + const Tensor norm, const Tensor std, + Tensor grad_input); + + +void psamask_forward(const Tensor input, Tensor output, const int psa_type, + const int num_, const int h_feature, const int w_feature, + const int h_mask, const int w_mask, const int half_h_mask, + const int half_w_mask); + +void psamask_backward(Tensor grad_output, const Tensor grad_input, + const int psa_type, const int num_, const int h_feature, + const int w_feature, const int h_mask, const int w_mask, + const int half_h_mask, const int half_w_mask); + + +PYBIND11_MODULE(TORCH_EXTENSION_NAME, m) { + m.def("get_compiler_version", &get_compiler_version, "get_compiler_version"); + m.def("get_compiling_cuda_version", &get_compiling_cuda_version, + "get_compiling_cuda_version"); + + + m.def("sigmoid_focal_loss_forward", &sigmoid_focal_loss_forward, + "sigmoid_focal_loss_forward ", py::arg("input"), py::arg("target"), + py::arg("weight"), py::arg("output"), py::arg("gamma"), + py::arg("alpha")); + m.def("sigmoid_focal_loss_backward", &sigmoid_focal_loss_backward, + "sigmoid_focal_loss_backward", py::arg("input"), py::arg("target"), + py::arg("weight"), py::arg("grad_input"), py::arg("gamma"), + py::arg("alpha")); + m.def("softmax_focal_loss_forward", &softmax_focal_loss_forward, + "softmax_focal_loss_forward", py::arg("input"), py::arg("target"), + py::arg("weight"), py::arg("output"), py::arg("gamma"), + py::arg("alpha")); + m.def("softmax_focal_loss_backward", &softmax_focal_loss_backward, + "softmax_focal_loss_backward", py::arg("input"), py::arg("target"), + py::arg("weight"), py::arg("buff"), py::arg("grad_input"), + py::arg("gamma"), py::arg("alpha")); + + // SyncBN + m.def("sync_bn_forward_mean", &sync_bn_forward_mean, "sync_bn forward_mean", + py::arg("input"), py::arg("mean")); + m.def("sync_bn_forward_var", &sync_bn_forward_var, "sync_bn forward_var", + py::arg("input"), py::arg("mean"), py::arg("var")); + m.def("sync_bn_forward_output", &sync_bn_forward_output, + "sync_bn forward_output", py::arg("input"), py::arg("mean"), + py::arg("var"), py::arg("weight"), py::arg("bias"), + py::arg("running_mean"), py::arg("running_var"), py::arg("norm"), + py::arg("std"), py::arg("output"), py::arg("eps"), py::arg("momentum"), + py::arg("group_size")); + m.def("sync_bn_backward_param", &sync_bn_backward_param, + "sync_bn backward_param", py::arg("grad_output"), py::arg("norm"), + py::arg("grad_weight"), py::arg("grad_bias")); + m.def("sync_bn_backward_data", &sync_bn_backward_data, + "sync_bn backward_data", py::arg("grad_output"), py::arg("weight"), + py::arg("grad_weight"), py::arg("grad_bias"), py::arg("norm"), + py::arg("std"), py::arg("grad_input")); + + // PASMask + m.def("psamask_forward", &psamask_forward, "PSAMASK forward (CPU/CUDA)", + py::arg("input"), py::arg("output"), py::arg("psa_type"), + py::arg("num_"), py::arg("h_feature"), py::arg("w_feature"), + py::arg("h_mask"), py::arg("w_mask"), py::arg("half_h_mask"), + py::arg("half_w_mask")); + m.def("psamask_backward", &psamask_backward, "PSAMASK backward (CPU/CUDA)", + py::arg("grad_output"), py::arg("grad_input"), py::arg("psa_type"), + py::arg("num_"), py::arg("h_feature"), py::arg("w_feature"), + py::arg("h_mask"), py::arg("w_mask"), py::arg("half_h_mask"), + py::arg("half_w_mask")); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/sync_bn.cpp b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/sync_bn.cpp new file mode 100644 index 0000000000000000000000000000000000000000..fd5a513273a7bbce2cf41c790706fe4801f4c414 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/csrc/pytorch/sync_bn.cpp @@ -0,0 +1,69 @@ +// Copyright (c) OpenMMLab. All rights reserved +#include "pytorch_cpp_helper.hpp" +#include "pytorch_device_registry.hpp" + +void sync_bn_forward_mean_impl(const Tensor input, Tensor mean) { + DISPATCH_DEVICE_IMPL(sync_bn_forward_mean_impl, input, mean); +} + +void sync_bn_forward_var_impl(const Tensor input, const Tensor mean, + Tensor var) { + DISPATCH_DEVICE_IMPL(sync_bn_forward_var_impl, input, mean, var); +} + +void sync_bn_forward_output_impl(const Tensor input, const Tensor mean, + const Tensor var, Tensor running_mean, + Tensor running_var, const Tensor weight, + const Tensor bias, Tensor norm, Tensor std, + Tensor output, float eps, float momentum, + int group_size) { + DISPATCH_DEVICE_IMPL(sync_bn_forward_output_impl, input, mean, var, + running_mean, running_var, weight, bias, norm, std, + output, eps, momentum, group_size); +} + +void sync_bn_backward_param_impl(const Tensor grad_output, const Tensor norm, + Tensor grad_weight, Tensor grad_bias) { + DISPATCH_DEVICE_IMPL(sync_bn_backward_param_impl, grad_output, norm, + grad_weight, grad_bias); +} + +void sync_bn_backward_data_impl(const Tensor grad_output, const Tensor weight, + const Tensor grad_weight, + const Tensor grad_bias, const Tensor norm, + const Tensor std, Tensor grad_input) { + DISPATCH_DEVICE_IMPL(sync_bn_backward_data_impl, grad_output, weight, + grad_weight, grad_bias, norm, std, grad_input); +} + +void sync_bn_forward_mean(const Tensor input, Tensor mean) { + sync_bn_forward_mean_impl(input, mean); +} + +void sync_bn_forward_var(const Tensor input, const Tensor mean, Tensor var) { + sync_bn_forward_var_impl(input, mean, var); +} + +void sync_bn_forward_output(const Tensor input, const Tensor mean, + const Tensor var, const Tensor weight, + const Tensor bias, Tensor running_mean, + Tensor running_var, Tensor norm, Tensor std, + Tensor output, float eps, float momentum, + int group_size) { + sync_bn_forward_output_impl(input, mean, var, running_mean, running_var, + weight, bias, norm, std, output, eps, momentum, + group_size); +} + +void sync_bn_backward_param(const Tensor grad_output, const Tensor norm, + Tensor grad_weight, Tensor grad_bias) { + sync_bn_backward_param_impl(grad_output, norm, grad_weight, grad_bias); +} + +void sync_bn_backward_data(const Tensor grad_output, const Tensor weight, + const Tensor grad_weight, const Tensor grad_bias, + const Tensor norm, const Tensor std, + Tensor grad_input) { + sync_bn_backward_data_impl(grad_output, weight, grad_weight, grad_bias, norm, + std, grad_input); +} diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/deprecated_wrappers.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/deprecated_wrappers.py new file mode 100644 index 0000000000000000000000000000000000000000..629a8033ff56be221b71a475ffd650ab7164f114 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/deprecated_wrappers.py @@ -0,0 +1,46 @@ +# Copyright (c) OpenMMLab. All rights reserved. +# This file is for backward compatibility. +# Module wrappers for empty tensor have been moved to mmcv.cnn.bricks. +import warnings + +from ..cnn.bricks.wrappers import Conv2d, ConvTranspose2d, Linear, MaxPool2d + + +class Conv2d_deprecated(Conv2d): + + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + warnings.warn( + 'Importing Conv2d wrapper from "mmcv.ops" will be deprecated in' + ' the future. Please import them from "mmcv.cnn" instead', + DeprecationWarning) + + +class ConvTranspose2d_deprecated(ConvTranspose2d): + + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + warnings.warn( + 'Importing ConvTranspose2d wrapper from "mmcv.ops" will be ' + 'deprecated in the future. Please import them from "mmcv.cnn" ' + 'instead', DeprecationWarning) + + +class MaxPool2d_deprecated(MaxPool2d): + + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + warnings.warn( + 'Importing MaxPool2d wrapper from "mmcv.ops" will be deprecated in' + ' the future. Please import them from "mmcv.cnn" instead', + DeprecationWarning) + + +class Linear_deprecated(Linear): + + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + warnings.warn( + 'Importing Linear wrapper from "mmcv.ops" will be deprecated in' + ' the future. Please import them from "mmcv.cnn" instead', + DeprecationWarning) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/focal_loss.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/focal_loss.py new file mode 100644 index 0000000000000000000000000000000000000000..805860516182715b174184fc862aff6286bbc5b3 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/focal_loss.py @@ -0,0 +1,213 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +from torch.autograd import Function +from torch.autograd.function import once_differentiable + +from ..utils import ext_loader + +ext_module = ext_loader.load_ext('_ext', [ + 'sigmoid_focal_loss_forward', 'sigmoid_focal_loss_backward', + 'softmax_focal_loss_forward', 'softmax_focal_loss_backward' +]) + + +class SigmoidFocalLossFunction(Function): + + @staticmethod + def symbolic(g, input, target, gamma, alpha, weight, reduction): + return g.op( + 'mmcv::MMCVSigmoidFocalLoss', + input, + target, + gamma_f=gamma, + alpha_f=alpha, + weight_f=weight, + reduction_s=reduction) + + @staticmethod + def forward(ctx, + input, + target, + gamma=2.0, + alpha=0.25, + weight=None, + reduction='mean'): + + assert isinstance( + target, (torch.Tensor, torch.LongTensor, torch.cuda.LongTensor)) + assert input.dim() == 2 + assert target.dim() == 1 + assert input.size(0) == target.size(0) + if weight is None: + weight = input.new_empty(0) + else: + assert weight.dim() == 1 + assert input.size(1) == weight.size(0) + ctx.reduction_dict = {'none': 0, 'mean': 1, 'sum': 2} + assert reduction in ctx.reduction_dict.keys() + + ctx.gamma = float(gamma) + ctx.alpha = float(alpha) + ctx.reduction = ctx.reduction_dict[reduction] + + output = input.new_zeros(input.size()) + + ext_module.sigmoid_focal_loss_forward( + input, target, weight, output, gamma=ctx.gamma, alpha=ctx.alpha) + if ctx.reduction == ctx.reduction_dict['mean']: + output = output.sum() / input.size(0) + elif ctx.reduction == ctx.reduction_dict['sum']: + output = output.sum() + ctx.save_for_backward(input, target, weight) + return output + + @staticmethod + @once_differentiable + def backward(ctx, grad_output): + input, target, weight = ctx.saved_tensors + + grad_input = input.new_zeros(input.size()) + + ext_module.sigmoid_focal_loss_backward( + input, + target, + weight, + grad_input, + gamma=ctx.gamma, + alpha=ctx.alpha) + + grad_input *= grad_output + if ctx.reduction == ctx.reduction_dict['mean']: + grad_input /= input.size(0) + return grad_input, None, None, None, None, None + + +sigmoid_focal_loss = SigmoidFocalLossFunction.apply + + +class SigmoidFocalLoss(nn.Module): + + def __init__(self, gamma, alpha, weight=None, reduction='mean'): + super(SigmoidFocalLoss, self).__init__() + self.gamma = gamma + self.alpha = alpha + self.register_buffer('weight', weight) + self.reduction = reduction + + def forward(self, input, target): + return sigmoid_focal_loss(input, target, self.gamma, self.alpha, + self.weight, self.reduction) + + def __repr__(self): + s = self.__class__.__name__ + s += f'(gamma={self.gamma}, ' + s += f'alpha={self.alpha}, ' + s += f'reduction={self.reduction})' + return s + + +class SoftmaxFocalLossFunction(Function): + + @staticmethod + def symbolic(g, input, target, gamma, alpha, weight, reduction): + return g.op( + 'mmcv::MMCVSoftmaxFocalLoss', + input, + target, + gamma_f=gamma, + alpha_f=alpha, + weight_f=weight, + reduction_s=reduction) + + @staticmethod + def forward(ctx, + input, + target, + gamma=2.0, + alpha=0.25, + weight=None, + reduction='mean'): + + assert isinstance(target, (torch.LongTensor, torch.cuda.LongTensor)) + assert input.dim() == 2 + assert target.dim() == 1 + assert input.size(0) == target.size(0) + if weight is None: + weight = input.new_empty(0) + else: + assert weight.dim() == 1 + assert input.size(1) == weight.size(0) + ctx.reduction_dict = {'none': 0, 'mean': 1, 'sum': 2} + assert reduction in ctx.reduction_dict.keys() + + ctx.gamma = float(gamma) + ctx.alpha = float(alpha) + ctx.reduction = ctx.reduction_dict[reduction] + + channel_stats, _ = torch.max(input, dim=1) + input_softmax = input - channel_stats.unsqueeze(1).expand_as(input) + input_softmax.exp_() + + channel_stats = input_softmax.sum(dim=1) + input_softmax /= channel_stats.unsqueeze(1).expand_as(input) + + output = input.new_zeros(input.size(0)) + ext_module.softmax_focal_loss_forward( + input_softmax, + target, + weight, + output, + gamma=ctx.gamma, + alpha=ctx.alpha) + + if ctx.reduction == ctx.reduction_dict['mean']: + output = output.sum() / input.size(0) + elif ctx.reduction == ctx.reduction_dict['sum']: + output = output.sum() + ctx.save_for_backward(input_softmax, target, weight) + return output + + @staticmethod + def backward(ctx, grad_output): + input_softmax, target, weight = ctx.saved_tensors + buff = input_softmax.new_zeros(input_softmax.size(0)) + grad_input = input_softmax.new_zeros(input_softmax.size()) + + ext_module.softmax_focal_loss_backward( + input_softmax, + target, + weight, + buff, + grad_input, + gamma=ctx.gamma, + alpha=ctx.alpha) + + grad_input *= grad_output + if ctx.reduction == ctx.reduction_dict['mean']: + grad_input /= input_softmax.size(0) + return grad_input, None, None, None, None, None + + +softmax_focal_loss = SoftmaxFocalLossFunction.apply + + +class SoftmaxFocalLoss(nn.Module): + + def __init__(self, gamma, alpha, weight=None, reduction='mean'): + super(SoftmaxFocalLoss, self).__init__() + self.gamma = gamma + self.alpha = alpha + self.register_buffer('weight', weight) + self.reduction = reduction + + def forward(self, input, target): + return softmax_focal_loss(input, target, self.gamma, self.alpha, + self.weight, self.reduction) + + def __repr__(self): + s = self.__class__.__name__ + s += f'(gamma={self.gamma}, ' + s += f'alpha={self.alpha}, ' + s += f'reduction={self.reduction})' + return s diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/info.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/info.py new file mode 100644 index 0000000000000000000000000000000000000000..29f2e5598ae2bb5866ccd15a7d3b4de33c0cd14d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/info.py @@ -0,0 +1,36 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import glob +import os + +import torch + +if torch.__version__ == 'parrots': + import parrots + + def get_compiler_version(): + return 'GCC ' + parrots.version.compiler + + def get_compiling_cuda_version(): + return parrots.version.cuda +else: + from ..utils import ext_loader + ext_module = ext_loader.load_ext( + '_ext', ['get_compiler_version', 'get_compiling_cuda_version']) + + def get_compiler_version(): + return ext_module.get_compiler_version() + + def get_compiling_cuda_version(): + return ext_module.get_compiling_cuda_version() + + +def get_onnxruntime_op_path(): + wildcard = os.path.join( + os.path.abspath(os.path.dirname(os.path.dirname(__file__))), + '_ext_ort.*.so') + + paths = glob.glob(wildcard) + if len(paths) > 0: + return paths[0] + else: + return '' diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/point_sample.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/point_sample.py new file mode 100644 index 0000000000000000000000000000000000000000..9a70b28e20458e1c7915054e4c29df105201dfdf --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/point_sample.py @@ -0,0 +1,346 @@ +# Modified from https://github.com/facebookresearch/detectron2/tree/master/projects/PointRend # noqa + +from os import path as osp + +import torch +import torch.nn as nn +import torch.nn.functional as F +from torch.nn.modules.utils import _pair +from torch.onnx.operators import shape_as_tensor + + +def bilinear_grid_sample(im, grid, align_corners=False): + """Given an input and a flow-field grid, computes the output using input + values and pixel locations from grid. Supported only bilinear interpolation + method to sample the input pixels. + + Args: + im (torch.Tensor): Input feature map, shape (N, C, H, W) + grid (torch.Tensor): Point coordinates, shape (N, Hg, Wg, 2) + align_corners {bool}: If set to True, the extrema (-1 and 1) are + considered as referring to the center points of the input’s + corner pixels. If set to False, they are instead considered as + referring to the corner points of the input’s corner pixels, + making the sampling more resolution agnostic. + + Returns: + torch.Tensor: A tensor with sampled points, shape (N, C, Hg, Wg) + """ + n, c, h, w = im.shape + gn, gh, gw, _ = grid.shape + assert n == gn + + x = grid[:, :, :, 0] + y = grid[:, :, :, 1] + + if align_corners: + x = ((x + 1) / 2) * (w - 1) + y = ((y + 1) / 2) * (h - 1) + else: + x = ((x + 1) * w - 1) / 2 + y = ((y + 1) * h - 1) / 2 + + x = x.view(n, -1) + y = y.view(n, -1) + + x0 = torch.floor(x).long() + y0 = torch.floor(y).long() + x1 = x0 + 1 + y1 = y0 + 1 + + wa = ((x1 - x) * (y1 - y)).unsqueeze(1) + wb = ((x1 - x) * (y - y0)).unsqueeze(1) + wc = ((x - x0) * (y1 - y)).unsqueeze(1) + wd = ((x - x0) * (y - y0)).unsqueeze(1) + + # Apply default for grid_sample function zero padding + im_padded = F.pad(im, pad=[1, 1, 1, 1], mode='constant', value=0) + padded_h = h + 2 + padded_w = w + 2 + # save points positions after padding + x0, x1, y0, y1 = x0 + 1, x1 + 1, y0 + 1, y1 + 1 + + # Clip coordinates to padded image size + x0 = torch.where(x0 < 0, torch.tensor(0), x0) + x0 = torch.where(x0 > padded_w - 1, torch.tensor(padded_w - 1), x0) + x1 = torch.where(x1 < 0, torch.tensor(0), x1) + x1 = torch.where(x1 > padded_w - 1, torch.tensor(padded_w - 1), x1) + y0 = torch.where(y0 < 0, torch.tensor(0), y0) + y0 = torch.where(y0 > padded_h - 1, torch.tensor(padded_h - 1), y0) + y1 = torch.where(y1 < 0, torch.tensor(0), y1) + y1 = torch.where(y1 > padded_h - 1, torch.tensor(padded_h - 1), y1) + + im_padded = im_padded.view(n, c, -1) + + x0_y0 = (x0 + y0 * padded_w).unsqueeze(1).expand(-1, c, -1) + x0_y1 = (x0 + y1 * padded_w).unsqueeze(1).expand(-1, c, -1) + x1_y0 = (x1 + y0 * padded_w).unsqueeze(1).expand(-1, c, -1) + x1_y1 = (x1 + y1 * padded_w).unsqueeze(1).expand(-1, c, -1) + + Ia = torch.gather(im_padded, 2, x0_y0) + Ib = torch.gather(im_padded, 2, x0_y1) + Ic = torch.gather(im_padded, 2, x1_y0) + Id = torch.gather(im_padded, 2, x1_y1) + + return (Ia * wa + Ib * wb + Ic * wc + Id * wd).reshape(n, c, gh, gw) + + +def is_in_onnx_export_without_custom_ops(): + from mmcv.ops import get_onnxruntime_op_path + ort_custom_op_path = get_onnxruntime_op_path() + return torch.onnx.is_in_onnx_export( + ) and not osp.exists(ort_custom_op_path) + + +def normalize(grid): + """Normalize input grid from [-1, 1] to [0, 1] + + Args: + grid (torch.Tensor): The grid to be normalize, range [-1, 1]. + + Returns: + torch.Tensor: Normalized grid, range [0, 1]. + """ + + return (grid + 1.0) / 2.0 + + +def denormalize(grid): + """Denormalize input grid from range [0, 1] to [-1, 1] + + Args: + grid (torch.Tensor): The grid to be denormalize, range [0, 1]. + + Returns: + torch.Tensor: Denormalized grid, range [-1, 1]. + """ + + return grid * 2.0 - 1.0 + + +def generate_grid(num_grid, size, device): + """Generate regular square grid of points in [0, 1] x [0, 1] coordinate + space. + + Args: + num_grid (int): The number of grids to sample, one for each region. + size (tuple[int, int]): The side size of the regular grid. + device (torch.device): Desired device of returned tensor. + + Returns: + torch.Tensor: A tensor of shape (num_grid, size[0]*size[1], 2) that + contains coordinates for the regular grids. + """ + + affine_trans = torch.tensor([[[1., 0., 0.], [0., 1., 0.]]], device=device) + grid = F.affine_grid( + affine_trans, torch.Size((1, 1, *size)), align_corners=False) + grid = normalize(grid) + return grid.view(1, -1, 2).expand(num_grid, -1, -1) + + +def rel_roi_point_to_abs_img_point(rois, rel_roi_points): + """Convert roi based relative point coordinates to image based absolute + point coordinates. + + Args: + rois (torch.Tensor): RoIs or BBoxes, shape (N, 4) or (N, 5) + rel_roi_points (torch.Tensor): Point coordinates inside RoI, relative + to RoI, location, range (0, 1), shape (N, P, 2) + Returns: + torch.Tensor: Image based absolute point coordinates, shape (N, P, 2) + """ + + with torch.no_grad(): + assert rel_roi_points.size(0) == rois.size(0) + assert rois.dim() == 2 + assert rel_roi_points.dim() == 3 + assert rel_roi_points.size(2) == 2 + # remove batch idx + if rois.size(1) == 5: + rois = rois[:, 1:] + abs_img_points = rel_roi_points.clone() + # To avoid an error during exporting to onnx use independent + # variables instead inplace computation + xs = abs_img_points[:, :, 0] * (rois[:, None, 2] - rois[:, None, 0]) + ys = abs_img_points[:, :, 1] * (rois[:, None, 3] - rois[:, None, 1]) + xs += rois[:, None, 0] + ys += rois[:, None, 1] + abs_img_points = torch.stack([xs, ys], dim=2) + return abs_img_points + + +def get_shape_from_feature_map(x): + """Get spatial resolution of input feature map considering exporting to + onnx mode. + + Args: + x (torch.Tensor): Input tensor, shape (N, C, H, W) + + Returns: + torch.Tensor: Spatial resolution (width, height), shape (1, 1, 2) + """ + if torch.onnx.is_in_onnx_export(): + img_shape = shape_as_tensor(x)[2:].flip(0).view(1, 1, 2).to( + x.device).float() + else: + img_shape = torch.tensor(x.shape[2:]).flip(0).view(1, 1, 2).to( + x.device).float() + return img_shape + + +def abs_img_point_to_rel_img_point(abs_img_points, img, spatial_scale=1.): + """Convert image based absolute point coordinates to image based relative + coordinates for sampling. + + Args: + abs_img_points (torch.Tensor): Image based absolute point coordinates, + shape (N, P, 2) + img (tuple or torch.Tensor): (height, width) of image or feature map. + spatial_scale (float, optional): Scale points by this factor. + Default: 1. + + Returns: + Tensor: Image based relative point coordinates for sampling, shape + (N, P, 2). + """ + + assert (isinstance(img, tuple) and len(img) == 2) or \ + (isinstance(img, torch.Tensor) and len(img.shape) == 4) + + if isinstance(img, tuple): + h, w = img + scale = torch.tensor([w, h], + dtype=torch.float, + device=abs_img_points.device) + scale = scale.view(1, 1, 2) + else: + scale = get_shape_from_feature_map(img) + + return abs_img_points / scale * spatial_scale + + +def rel_roi_point_to_rel_img_point(rois, + rel_roi_points, + img, + spatial_scale=1.): + """Convert roi based relative point coordinates to image based absolute + point coordinates. + + Args: + rois (torch.Tensor): RoIs or BBoxes, shape (N, 4) or (N, 5) + rel_roi_points (torch.Tensor): Point coordinates inside RoI, relative + to RoI, location, range (0, 1), shape (N, P, 2) + img (tuple or torch.Tensor): (height, width) of image or feature map. + spatial_scale (float, optional): Scale points by this factor. + Default: 1. + + Returns: + torch.Tensor: Image based relative point coordinates for sampling, + shape (N, P, 2). + """ + + abs_img_point = rel_roi_point_to_abs_img_point(rois, rel_roi_points) + rel_img_point = abs_img_point_to_rel_img_point(abs_img_point, img, + spatial_scale) + + return rel_img_point + + +def point_sample(input, points, align_corners=False, **kwargs): + """A wrapper around :func:`grid_sample` to support 3D point_coords tensors + Unlike :func:`torch.nn.functional.grid_sample` it assumes point_coords to + lie inside ``[0, 1] x [0, 1]`` square. + + Args: + input (torch.Tensor): Feature map, shape (N, C, H, W). + points (torch.Tensor): Image based absolute point coordinates + (normalized), range [0, 1] x [0, 1], shape (N, P, 2) or + (N, Hgrid, Wgrid, 2). + align_corners (bool, optional): Whether align_corners. + Default: False + + Returns: + torch.Tensor: Features of `point` on `input`, shape (N, C, P) or + (N, C, Hgrid, Wgrid). + """ + + add_dim = False + if points.dim() == 3: + add_dim = True + points = points.unsqueeze(2) + if is_in_onnx_export_without_custom_ops(): + # If custom ops for onnx runtime not compiled use python + # implementation of grid_sample function to make onnx graph + # with supported nodes + output = bilinear_grid_sample( + input, denormalize(points), align_corners=align_corners) + else: + output = F.grid_sample( + input, denormalize(points), align_corners=align_corners, **kwargs) + if add_dim: + output = output.squeeze(3) + return output + + +class SimpleRoIAlign(nn.Module): + + def __init__(self, output_size, spatial_scale, aligned=True): + """Simple RoI align in PointRend, faster than standard RoIAlign. + + Args: + output_size (tuple[int]): h, w + spatial_scale (float): scale the input boxes by this number + aligned (bool): if False, use the legacy implementation in + MMDetection, align_corners=True will be used in F.grid_sample. + If True, align the results more perfectly. + """ + + super(SimpleRoIAlign, self).__init__() + self.output_size = _pair(output_size) + self.spatial_scale = float(spatial_scale) + # to be consistent with other RoI ops + self.use_torchvision = False + self.aligned = aligned + + def forward(self, features, rois): + num_imgs = features.size(0) + num_rois = rois.size(0) + rel_roi_points = generate_grid( + num_rois, self.output_size, device=rois.device) + + if torch.onnx.is_in_onnx_export(): + rel_img_points = rel_roi_point_to_rel_img_point( + rois, rel_roi_points, features, self.spatial_scale) + rel_img_points = rel_img_points.reshape(num_imgs, -1, + *rel_img_points.shape[1:]) + point_feats = point_sample( + features, rel_img_points, align_corners=not self.aligned) + point_feats = point_feats.transpose(1, 2) + else: + point_feats = [] + for batch_ind in range(num_imgs): + # unravel batch dim + feat = features[batch_ind].unsqueeze(0) + inds = (rois[:, 0].long() == batch_ind) + if inds.any(): + rel_img_points = rel_roi_point_to_rel_img_point( + rois[inds], rel_roi_points[inds], feat, + self.spatial_scale).unsqueeze(0) + point_feat = point_sample( + feat, rel_img_points, align_corners=not self.aligned) + point_feat = point_feat.squeeze(0).transpose(0, 1) + point_feats.append(point_feat) + + point_feats = torch.cat(point_feats, dim=0) + + channels = features.size(1) + roi_feats = point_feats.reshape(num_rois, channels, *self.output_size) + + return roi_feats + + def __repr__(self): + format_str = self.__class__.__name__ + format_str += '(output_size={}, spatial_scale={}'.format( + self.output_size, self.spatial_scale) + return format_str diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/psa_mask.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/psa_mask.py new file mode 100644 index 0000000000000000000000000000000000000000..cdf14e62b50e8d4dd6856c94333c703bcc4c9ab6 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/psa_mask.py @@ -0,0 +1,92 @@ +# Modified from https://github.com/hszhao/semseg/blob/master/lib/psa +from torch import nn +from torch.autograd import Function +from torch.nn.modules.utils import _pair + +from ..utils import ext_loader + +ext_module = ext_loader.load_ext('_ext', + ['psamask_forward', 'psamask_backward']) + + +class PSAMaskFunction(Function): + + @staticmethod + def symbolic(g, input, psa_type, mask_size): + return g.op( + 'mmcv::MMCVPSAMask', + input, + psa_type_i=psa_type, + mask_size_i=mask_size) + + @staticmethod + def forward(ctx, input, psa_type, mask_size): + ctx.psa_type = psa_type + ctx.mask_size = _pair(mask_size) + ctx.save_for_backward(input) + + h_mask, w_mask = ctx.mask_size + batch_size, channels, h_feature, w_feature = input.size() + assert channels == h_mask * w_mask + output = input.new_zeros( + (batch_size, h_feature * w_feature, h_feature, w_feature)) + + ext_module.psamask_forward( + input, + output, + psa_type=psa_type, + num_=batch_size, + h_feature=h_feature, + w_feature=w_feature, + h_mask=h_mask, + w_mask=w_mask, + half_h_mask=(h_mask - 1) // 2, + half_w_mask=(w_mask - 1) // 2) + return output + + @staticmethod + def backward(ctx, grad_output): + input = ctx.saved_tensors[0] + psa_type = ctx.psa_type + h_mask, w_mask = ctx.mask_size + batch_size, channels, h_feature, w_feature = input.size() + grad_input = grad_output.new_zeros( + (batch_size, channels, h_feature, w_feature)) + ext_module.psamask_backward( + grad_output, + grad_input, + psa_type=psa_type, + num_=batch_size, + h_feature=h_feature, + w_feature=w_feature, + h_mask=h_mask, + w_mask=w_mask, + half_h_mask=(h_mask - 1) // 2, + half_w_mask=(w_mask - 1) // 2) + return grad_input, None, None, None + + +psa_mask = PSAMaskFunction.apply + + +class PSAMask(nn.Module): + + def __init__(self, psa_type, mask_size=None): + super(PSAMask, self).__init__() + assert psa_type in ['collect', 'distribute'] + if psa_type == 'collect': + psa_type_enum = 0 + else: + psa_type_enum = 1 + self.psa_type_enum = psa_type_enum + self.mask_size = mask_size + self.psa_type = psa_type + + def forward(self, input): + return psa_mask(input, self.psa_type_enum, self.mask_size) + + def __repr__(self): + s = self.__class__.__name__ + s += f'(psa_type={self.psa_type}, ' + s += f'mask_size={self.mask_size})' + return s diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/sync_bn.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/sync_bn.py new file mode 100644 index 0000000000000000000000000000000000000000..04302f03131785c99430868142a6ba5bb8600b1d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/ops/sync_bn.py @@ -0,0 +1,279 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.distributed as dist +import torch.nn.functional as F +from torch.autograd import Function +from torch.autograd.function import once_differentiable +from torch.nn.modules.module import Module +from torch.nn.parameter import Parameter + +from mmcv.cnn import NORM_LAYERS +from ..utils import ext_loader + +ext_module = ext_loader.load_ext('_ext', [ + 'sync_bn_forward_mean', 'sync_bn_forward_var', 'sync_bn_forward_output', + 'sync_bn_backward_param', 'sync_bn_backward_data' +]) + + +class SyncBatchNormFunction(Function): + + @staticmethod + def symbolic(g, input, running_mean, running_var, weight, bias, momentum, + eps, group, group_size, stats_mode): + return g.op( + 'mmcv::MMCVSyncBatchNorm', + input, + running_mean, + running_var, + weight, + bias, + momentum_f=momentum, + eps_f=eps, + group_i=group, + group_size_i=group_size, + stats_mode=stats_mode) + + @staticmethod + def forward(self, input, running_mean, running_var, weight, bias, momentum, + eps, group, group_size, stats_mode): + self.momentum = momentum + self.eps = eps + self.group = group + self.group_size = group_size + self.stats_mode = stats_mode + + assert isinstance( + input, (torch.HalfTensor, torch.FloatTensor, + torch.cuda.HalfTensor, torch.cuda.FloatTensor)), \ + f'only support Half or Float Tensor, but {input.type()}' + output = torch.zeros_like(input) + input3d = input.flatten(start_dim=2) + output3d = output.view_as(input3d) + num_channels = input3d.size(1) + + # ensure mean/var/norm/std are initialized as zeros + # ``torch.empty()`` does not guarantee that + mean = torch.zeros( + num_channels, dtype=torch.float, device=input3d.device) + var = torch.zeros( + num_channels, dtype=torch.float, device=input3d.device) + norm = torch.zeros_like( + input3d, dtype=torch.float, device=input3d.device) + std = torch.zeros( + num_channels, dtype=torch.float, device=input3d.device) + + batch_size = input3d.size(0) + if batch_size > 0: + ext_module.sync_bn_forward_mean(input3d, mean) + batch_flag = torch.ones([1], device=mean.device, dtype=mean.dtype) + else: + # skip updating mean and leave it as zeros when the input is empty + batch_flag = torch.zeros([1], device=mean.device, dtype=mean.dtype) + + # synchronize mean and the batch flag + vec = torch.cat([mean, batch_flag]) + if self.stats_mode == 'N': + vec *= batch_size + if self.group_size > 1: + dist.all_reduce(vec, group=self.group) + total_batch = vec[-1].detach() + mean = vec[:num_channels] + + if self.stats_mode == 'default': + mean = mean / self.group_size + elif self.stats_mode == 'N': + mean = mean / total_batch.clamp(min=1) + else: + raise NotImplementedError + + # leave var as zeros when the input is empty + if batch_size > 0: + ext_module.sync_bn_forward_var(input3d, mean, var) + + if self.stats_mode == 'N': + var *= batch_size + if self.group_size > 1: + dist.all_reduce(var, group=self.group) + + if self.stats_mode == 'default': + var /= self.group_size + elif self.stats_mode == 'N': + var /= total_batch.clamp(min=1) + else: + raise NotImplementedError + + # if the total batch size over all the ranks is zero, + # we should not update the statistics in the current batch + update_flag = total_batch.clamp(max=1) + momentum = update_flag * self.momentum + ext_module.sync_bn_forward_output( + input3d, + mean, + var, + weight, + bias, + running_mean, + running_var, + norm, + std, + output3d, + eps=self.eps, + momentum=momentum, + group_size=self.group_size) + self.save_for_backward(norm, std, weight) + return output + + @staticmethod + @once_differentiable + def backward(self, grad_output): + norm, std, weight = self.saved_tensors + grad_weight = torch.zeros_like(weight) + grad_bias = torch.zeros_like(weight) + grad_input = torch.zeros_like(grad_output) + grad_output3d = grad_output.flatten(start_dim=2) + grad_input3d = grad_input.view_as(grad_output3d) + + batch_size = grad_input3d.size(0) + if batch_size > 0: + ext_module.sync_bn_backward_param(grad_output3d, norm, grad_weight, + grad_bias) + + # all reduce + if self.group_size > 1: + dist.all_reduce(grad_weight, group=self.group) + dist.all_reduce(grad_bias, group=self.group) + grad_weight /= self.group_size + grad_bias /= self.group_size + + if batch_size > 0: + ext_module.sync_bn_backward_data(grad_output3d, weight, + grad_weight, grad_bias, norm, std, + grad_input3d) + + return grad_input, None, None, grad_weight, grad_bias, \ + None, None, None, None, None + + +@NORM_LAYERS.register_module(name='MMSyncBN') +class SyncBatchNorm(Module): + """Synchronized Batch Normalization. + + Args: + num_features (int): number of features/chennels in input tensor + eps (float, optional): a value added to the denominator for numerical + stability. Defaults to 1e-5. + momentum (float, optional): the value used for the running_mean and + running_var computation. Defaults to 0.1. + affine (bool, optional): whether to use learnable affine parameters. + Defaults to True. + track_running_stats (bool, optional): whether to track the running + mean and variance during training. When set to False, this + module does not track such statistics, and initializes statistics + buffers ``running_mean`` and ``running_var`` as ``None``. When + these buffers are ``None``, this module always uses batch + statistics in both training and eval modes. Defaults to True. + group (int, optional): synchronization of stats happen within + each process group individually. By default it is synchronization + across the whole world. Defaults to None. + stats_mode (str, optional): The statistical mode. Available options + includes ``'default'`` and ``'N'``. Defaults to 'default'. + When ``stats_mode=='default'``, it computes the overall statistics + using those from each worker with equal weight, i.e., the + statistics are synchronized and simply divied by ``group``. This + mode will produce inaccurate statistics when empty tensors occur. + When ``stats_mode=='N'``, it compute the overall statistics using + the total number of batches in each worker ignoring the number of + group, i.e., the statistics are synchronized and then divied by + the total batch ``N``. This mode is beneficial when empty tensors + occur during training, as it average the total mean by the real + number of batch. + """ + + def __init__(self, + num_features, + eps=1e-5, + momentum=0.1, + affine=True, + track_running_stats=True, + group=None, + stats_mode='default'): + super(SyncBatchNorm, self).__init__() + self.num_features = num_features + self.eps = eps + self.momentum = momentum + self.affine = affine + self.track_running_stats = track_running_stats + group = dist.group.WORLD if group is None else group + self.group = group + self.group_size = dist.get_world_size(group) + assert stats_mode in ['default', 'N'], \ + f'"stats_mode" only accepts "default" and "N", got "{stats_mode}"' + self.stats_mode = stats_mode + if self.affine: + self.weight = Parameter(torch.Tensor(num_features)) + self.bias = Parameter(torch.Tensor(num_features)) + else: + self.register_parameter('weight', None) + self.register_parameter('bias', None) + if self.track_running_stats: + self.register_buffer('running_mean', torch.zeros(num_features)) + self.register_buffer('running_var', torch.ones(num_features)) + self.register_buffer('num_batches_tracked', + torch.tensor(0, dtype=torch.long)) + else: + self.register_buffer('running_mean', None) + self.register_buffer('running_var', None) + self.register_buffer('num_batches_tracked', None) + self.reset_parameters() + + def reset_running_stats(self): + if self.track_running_stats: + self.running_mean.zero_() + self.running_var.fill_(1) + self.num_batches_tracked.zero_() + + def reset_parameters(self): + self.reset_running_stats() + if self.affine: + self.weight.data.uniform_() # pytorch use ones_() + self.bias.data.zero_() + + def forward(self, input): + if input.dim() < 2: + raise ValueError( + f'expected at least 2D input, got {input.dim()}D input') + if self.momentum is None: + exponential_average_factor = 0.0 + else: + exponential_average_factor = self.momentum + + if self.training and self.track_running_stats: + if self.num_batches_tracked is not None: + self.num_batches_tracked += 1 + if self.momentum is None: # use cumulative moving average + exponential_average_factor = 1.0 / float( + self.num_batches_tracked) + else: # use exponential moving average + exponential_average_factor = self.momentum + + if self.training or not self.track_running_stats: + return SyncBatchNormFunction.apply( + input, self.running_mean, self.running_var, self.weight, + self.bias, exponential_average_factor, self.eps, self.group, + self.group_size, self.stats_mode) + else: + return F.batch_norm(input, self.running_mean, self.running_var, + self.weight, self.bias, False, + exponential_average_factor, self.eps) + + def __repr__(self): + s = self.__class__.__name__ + s += f'({self.num_features}, ' + s += f'eps={self.eps}, ' + s += f'momentum={self.momentum}, ' + s += f'affine={self.affine}, ' + s += f'track_running_stats={self.track_running_stats}, ' + s += f'group_size={self.group_size},' + s += f'stats_mode={self.stats_mode})' + return s diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..2ed2c17ad357742e423beeaf4d35db03fe9af469 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/__init__.py @@ -0,0 +1,13 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .collate import collate +from .data_container import DataContainer +from .data_parallel import MMDataParallel +from .distributed import MMDistributedDataParallel +from .registry import MODULE_WRAPPERS +from .scatter_gather import scatter, scatter_kwargs +from .utils import is_module_wrapper + +__all__ = [ + 'collate', 'DataContainer', 'MMDataParallel', 'MMDistributedDataParallel', + 'scatter', 'scatter_kwargs', 'is_module_wrapper', 'MODULE_WRAPPERS' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/_functions.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/_functions.py new file mode 100644 index 0000000000000000000000000000000000000000..95c58bf1a89a0b9ad8305dbd09e3f4648fe26467 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/_functions.py @@ -0,0 +1,76 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +from torch.nn.parallel._functions import _get_stream + + +def scatter(input, devices, streams=None): + """Scatters tensor across multiple GPUs.""" + if streams is None: + streams = [None] * len(devices) + + if isinstance(input, list): + chunk_size = (len(input) - 1) // len(devices) + 1 + outputs = [ + scatter(input[i], [devices[i // chunk_size]], + [streams[i // chunk_size]]) for i in range(len(input)) + ] + return outputs + elif isinstance(input, torch.Tensor): + output = input.contiguous() + # TODO: copy to a pinned buffer first (if copying from CPU) + stream = streams[0] if output.numel() > 0 else None + if devices != [-1]: + with torch.cuda.device(devices[0]), torch.cuda.stream(stream): + output = output.cuda(devices[0], non_blocking=True) + + return output + else: + raise Exception(f'Unknown type {type(input)}.') + + +def synchronize_stream(output, devices, streams): + if isinstance(output, list): + chunk_size = len(output) // len(devices) + for i in range(len(devices)): + for j in range(chunk_size): + synchronize_stream(output[i * chunk_size + j], [devices[i]], + [streams[i]]) + elif isinstance(output, torch.Tensor): + if output.numel() != 0: + with torch.cuda.device(devices[0]): + main_stream = torch.cuda.current_stream() + main_stream.wait_stream(streams[0]) + output.record_stream(main_stream) + else: + raise Exception(f'Unknown type {type(output)}.') + + +def get_input_device(input): + if isinstance(input, list): + for item in input: + input_device = get_input_device(item) + if input_device != -1: + return input_device + return -1 + elif isinstance(input, torch.Tensor): + return input.get_device() if input.is_cuda else -1 + else: + raise Exception(f'Unknown type {type(input)}.') + + +class Scatter: + + @staticmethod + def forward(target_gpus, input): + input_device = get_input_device(input) + streams = None + if input_device == -1 and target_gpus != [-1]: + # Perform CPU to GPU copies in a background stream + streams = [_get_stream(device) for device in target_gpus] + + outputs = scatter(input, target_gpus, streams) + # Synchronize with the copy stream + if streams is not None: + synchronize_stream(outputs, target_gpus, streams) + + return tuple(outputs) if isinstance(outputs, list) else (outputs, ) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/collate.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/collate.py new file mode 100644 index 0000000000000000000000000000000000000000..ad749197df21b0d74297548be5f66a696adebf7f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/collate.py @@ -0,0 +1,84 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from collections.abc import Mapping, Sequence + +import torch +import torch.nn.functional as F +from torch.utils.data.dataloader import default_collate + +from .data_container import DataContainer + + +def collate(batch, samples_per_gpu=1): + """Puts each data field into a tensor/DataContainer with outer dimension + batch size. + + Extend default_collate to add support for + :type:`~mmcv.parallel.DataContainer`. There are 3 cases. + + 1. cpu_only = True, e.g., meta data + 2. cpu_only = False, stack = True, e.g., images tensors + 3. cpu_only = False, stack = False, e.g., gt bboxes + """ + + if not isinstance(batch, Sequence): + raise TypeError(f'{batch.dtype} is not supported.') + + if isinstance(batch[0], DataContainer): + stacked = [] + if batch[0].cpu_only: + for i in range(0, len(batch), samples_per_gpu): + stacked.append( + [sample.data for sample in batch[i:i + samples_per_gpu]]) + return DataContainer( + stacked, batch[0].stack, batch[0].padding_value, cpu_only=True) + elif batch[0].stack: + for i in range(0, len(batch), samples_per_gpu): + assert isinstance(batch[i].data, torch.Tensor) + + if batch[i].pad_dims is not None: + ndim = batch[i].dim() + assert ndim > batch[i].pad_dims + max_shape = [0 for _ in range(batch[i].pad_dims)] + for dim in range(1, batch[i].pad_dims + 1): + max_shape[dim - 1] = batch[i].size(-dim) + for sample in batch[i:i + samples_per_gpu]: + for dim in range(0, ndim - batch[i].pad_dims): + assert batch[i].size(dim) == sample.size(dim) + for dim in range(1, batch[i].pad_dims + 1): + max_shape[dim - 1] = max(max_shape[dim - 1], + sample.size(-dim)) + padded_samples = [] + for sample in batch[i:i + samples_per_gpu]: + pad = [0 for _ in range(batch[i].pad_dims * 2)] + for dim in range(1, batch[i].pad_dims + 1): + pad[2 * dim - + 1] = max_shape[dim - 1] - sample.size(-dim) + padded_samples.append( + F.pad( + sample.data, pad, value=sample.padding_value)) + stacked.append(default_collate(padded_samples)) + elif batch[i].pad_dims is None: + stacked.append( + default_collate([ + sample.data + for sample in batch[i:i + samples_per_gpu] + ])) + else: + raise ValueError( + 'pad_dims should be either None or integers (1-3)') + + else: + for i in range(0, len(batch), samples_per_gpu): + stacked.append( + [sample.data for sample in batch[i:i + samples_per_gpu]]) + return DataContainer(stacked, batch[0].stack, batch[0].padding_value) + elif isinstance(batch[0], Sequence): + transposed = zip(*batch) + return [collate(samples, samples_per_gpu) for samples in transposed] + elif isinstance(batch[0], Mapping): + return { + key: collate([d[key] for d in batch], samples_per_gpu) + for key in batch[0] + } + else: + return default_collate(batch) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/data_container.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/data_container.py new file mode 100644 index 0000000000000000000000000000000000000000..cedb0d32a51a1f575a622b38de2cee3ab4757821 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/data_container.py @@ -0,0 +1,89 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import functools + +import torch + + +def assert_tensor_type(func): + + @functools.wraps(func) + def wrapper(*args, **kwargs): + if not isinstance(args[0].data, torch.Tensor): + raise AttributeError( + f'{args[0].__class__.__name__} has no attribute ' + f'{func.__name__} for type {args[0].datatype}') + return func(*args, **kwargs) + + return wrapper + + +class DataContainer: + """A container for any type of objects. + + Typically tensors will be stacked in the collate function and sliced along + some dimension in the scatter function. This behavior has some limitations. + 1. All tensors have to be the same size. + 2. Types are limited (numpy array or Tensor). + + We design `DataContainer` and `MMDataParallel` to overcome these + limitations. The behavior can be either of the following. + + - copy to GPU, pad all tensors to the same size and stack them + - copy to GPU without stacking + - leave the objects as is and pass it to the model + - pad_dims specifies the number of last few dimensions to do padding + """ + + def __init__(self, + data, + stack=False, + padding_value=0, + cpu_only=False, + pad_dims=2): + self._data = data + self._cpu_only = cpu_only + self._stack = stack + self._padding_value = padding_value + assert pad_dims in [None, 1, 2, 3] + self._pad_dims = pad_dims + + def __repr__(self): + return f'{self.__class__.__name__}({repr(self.data)})' + + def __len__(self): + return len(self._data) + + @property + def data(self): + return self._data + + @property + def datatype(self): + if isinstance(self.data, torch.Tensor): + return self.data.type() + else: + return type(self.data) + + @property + def cpu_only(self): + return self._cpu_only + + @property + def stack(self): + return self._stack + + @property + def padding_value(self): + return self._padding_value + + @property + def pad_dims(self): + return self._pad_dims + + @assert_tensor_type + def size(self, *args, **kwargs): + return self.data.size(*args, **kwargs) + + @assert_tensor_type + def dim(self): + return self.data.dim() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/data_parallel.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/data_parallel.py new file mode 100644 index 0000000000000000000000000000000000000000..7a5abeb6ed4e7159634a43a299b8747937242e7b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/data_parallel.py @@ -0,0 +1,97 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from itertools import chain + +from torch.nn.parallel import DataParallel + +from .scatter_gather import scatter_kwargs + + +class MMDataParallel(DataParallel): + """The DataParallel module that supports DataContainer. + + MMDataParallel has two main differences with PyTorch DataParallel: + + - It supports a custom type :class:`DataContainer` which allows more + flexible control of input data during both GPU and CPU inference. + - It implement two more APIs ``train_step()`` and ``val_step()``. + + .. warning:: + MMDataParallel only supports single GPU training, if you need to + train with multiple GPUs, please use MMDistributedDataParallel + instead. If you have multiple GPUs and you just want to use + MMDataParallel, you can set the environment variable + ``CUDA_VISIBLE_DEVICES=0`` or instantiate ``MMDataParallel`` with + ``device_ids=[0]``. + + Args: + module (:class:`nn.Module`): Module to be encapsulated. + device_ids (list[int]): Device IDS of modules to be scattered to. + Defaults to None when GPU is not available. + output_device (str | int): Device ID for output. Defaults to None. + dim (int): Dimension used to scatter the data. Defaults to 0. + """ + + def __init__(self, *args, dim=0, **kwargs): + super(MMDataParallel, self).__init__(*args, dim=dim, **kwargs) + self.dim = dim + + def forward(self, *inputs, **kwargs): + """Override the original forward function. + + The main difference lies in the CPU inference where the data in + :class:`DataContainers` will still be gathered. + """ + if not self.device_ids: + # We add the following line thus the module could gather and + # convert data containers as those in GPU inference + inputs, kwargs = self.scatter(inputs, kwargs, [-1]) + return self.module(*inputs[0], **kwargs[0]) + else: + return super().forward(*inputs, **kwargs) + + def scatter(self, inputs, kwargs, device_ids): + return scatter_kwargs(inputs, kwargs, device_ids, dim=self.dim) + + def train_step(self, *inputs, **kwargs): + if not self.device_ids: + # We add the following line thus the module could gather and + # convert data containers as those in GPU inference + inputs, kwargs = self.scatter(inputs, kwargs, [-1]) + return self.module.train_step(*inputs[0], **kwargs[0]) + + assert len(self.device_ids) == 1, \ + ('MMDataParallel only supports single GPU training, if you need to' + ' train with multiple GPUs, please use MMDistributedDataParallel' + ' instead.') + + for t in chain(self.module.parameters(), self.module.buffers()): + if t.device != self.src_device_obj: + raise RuntimeError( + 'module must have its parameters and buffers ' + f'on device {self.src_device_obj} (device_ids[0]) but ' + f'found one of them on device: {t.device}') + + inputs, kwargs = self.scatter(inputs, kwargs, self.device_ids) + return self.module.train_step(*inputs[0], **kwargs[0]) + + def val_step(self, *inputs, **kwargs): + if not self.device_ids: + # We add the following line thus the module could gather and + # convert data containers as those in GPU inference + inputs, kwargs = self.scatter(inputs, kwargs, [-1]) + return self.module.val_step(*inputs[0], **kwargs[0]) + + assert len(self.device_ids) == 1, \ + ('MMDataParallel only supports single GPU training, if you need to' + ' train with multiple GPUs, please use MMDistributedDataParallel' + ' instead.') + + for t in chain(self.module.parameters(), self.module.buffers()): + if t.device != self.src_device_obj: + raise RuntimeError( + 'module must have its parameters and buffers ' + f'on device {self.src_device_obj} (device_ids[0]) but ' + f'found one of them on device: {t.device}') + + inputs, kwargs = self.scatter(inputs, kwargs, self.device_ids) + return self.module.val_step(*inputs[0], **kwargs[0]) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/distributed.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/distributed.py new file mode 100644 index 0000000000000000000000000000000000000000..0188ca4ab26010b6f62f14fbefdb8b862d4eaac7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/distributed.py @@ -0,0 +1,138 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +from torch.nn.parallel.distributed import (DistributedDataParallel, + _find_tensors) + +from mmcv import print_log +from mmcv.utils import TORCH_VERSION, digit_version +from .scatter_gather import scatter_kwargs + + +class MMDistributedDataParallel(DistributedDataParallel): + """The DDP module that supports DataContainer. + + MMDDP has two main differences with PyTorch DDP: + + - It supports a custom type :class:`DataContainer` which allows more + flexible control of input data. + - It implement two APIs ``train_step()`` and ``val_step()``. + """ + + def to_kwargs(self, inputs, kwargs, device_id): + # Use `self.to_kwargs` instead of `self.scatter` in pytorch1.8 + # to move all tensors to device_id + return scatter_kwargs(inputs, kwargs, [device_id], dim=self.dim) + + def scatter(self, inputs, kwargs, device_ids): + return scatter_kwargs(inputs, kwargs, device_ids, dim=self.dim) + + def train_step(self, *inputs, **kwargs): + """train_step() API for module wrapped by DistributedDataParallel. + + This method is basically the same as + ``DistributedDataParallel.forward()``, while replacing + ``self.module.forward()`` with ``self.module.train_step()``. + It is compatible with PyTorch 1.1 - 1.5. + """ + + # In PyTorch >= 1.7, ``reducer._rebuild_buckets()`` is moved from the + # end of backward to the beginning of forward. + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) >= digit_version('1.7') + and self.reducer._rebuild_buckets()): + print_log( + 'Reducer buckets have been rebuilt in this iteration.', + logger='mmcv') + + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) >= digit_version('1.11.0a0')): + if self._check_sync_bufs_pre_fwd(): + self._sync_buffers() + else: + if (getattr(self, 'require_forward_param_sync', False) + and self.require_forward_param_sync): + self._sync_params() + + if self.device_ids: + inputs, kwargs = self.scatter(inputs, kwargs, self.device_ids) + if len(self.device_ids) == 1: + output = self.module.train_step(*inputs[0], **kwargs[0]) + else: + outputs = self.parallel_apply( + self._module_copies[:len(inputs)], inputs, kwargs) + output = self.gather(outputs, self.output_device) + else: + output = self.module.train_step(*inputs, **kwargs) + + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) >= digit_version('1.11.0a0')): + if self._check_sync_bufs_post_fwd(): + self._sync_buffers() + + if (torch.is_grad_enabled() + and getattr(self, 'require_backward_grad_sync', False) + and self.require_backward_grad_sync): + if self.find_unused_parameters: + self.reducer.prepare_for_backward(list(_find_tensors(output))) + else: + self.reducer.prepare_for_backward([]) + else: + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) > digit_version('1.2')): + self.require_forward_param_sync = False + return output + + def val_step(self, *inputs, **kwargs): + """val_step() API for module wrapped by DistributedDataParallel. + + This method is basically the same as + ``DistributedDataParallel.forward()``, while replacing + ``self.module.forward()`` with ``self.module.val_step()``. + It is compatible with PyTorch 1.1 - 1.5. + """ + # In PyTorch >= 1.7, ``reducer._rebuild_buckets()`` is moved from the + # end of backward to the beginning of forward. + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) >= digit_version('1.7') + and self.reducer._rebuild_buckets()): + print_log( + 'Reducer buckets have been rebuilt in this iteration.', + logger='mmcv') + + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) >= digit_version('1.11.0a0')): + if self._check_sync_bufs_pre_fwd(): + self._sync_buffers() + else: + if (getattr(self, 'require_forward_param_sync', False) + and self.require_forward_param_sync): + self._sync_params() + + if self.device_ids: + inputs, kwargs = self.scatter(inputs, kwargs, self.device_ids) + if len(self.device_ids) == 1: + output = self.module.val_step(*inputs[0], **kwargs[0]) + else: + outputs = self.parallel_apply( + self._module_copies[:len(inputs)], inputs, kwargs) + output = self.gather(outputs, self.output_device) + else: + output = self.module.val_step(*inputs, **kwargs) + + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) >= digit_version('1.11.0a0')): + if self._check_sync_bufs_post_fwd(): + self._sync_buffers() + + if (torch.is_grad_enabled() + and getattr(self, 'require_backward_grad_sync', False) + and self.require_backward_grad_sync): + if self.find_unused_parameters: + self.reducer.prepare_for_backward(list(_find_tensors(output))) + else: + self.reducer.prepare_for_backward([]) + else: + if ('parrots' not in TORCH_VERSION + and digit_version(TORCH_VERSION) > digit_version('1.2')): + self.require_forward_param_sync = False + return output diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/distributed_deprecated.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/distributed_deprecated.py new file mode 100644 index 0000000000000000000000000000000000000000..b593d4a9e04a53adb9cc5850cd7e30321e1528ce --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/distributed_deprecated.py @@ -0,0 +1,70 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.distributed as dist +import torch.nn as nn +from torch._utils import (_flatten_dense_tensors, _take_tensors, + _unflatten_dense_tensors) + +from mmcv.utils import TORCH_VERSION, digit_version +from .registry import MODULE_WRAPPERS +from .scatter_gather import scatter_kwargs + + +@MODULE_WRAPPERS.register_module() +class MMDistributedDataParallel(nn.Module): + + def __init__(self, + module, + dim=0, + broadcast_buffers=True, + bucket_cap_mb=25): + super(MMDistributedDataParallel, self).__init__() + self.module = module + self.dim = dim + self.broadcast_buffers = broadcast_buffers + + self.broadcast_bucket_size = bucket_cap_mb * 1024 * 1024 + self._sync_params() + + def _dist_broadcast_coalesced(self, tensors, buffer_size): + for tensors in _take_tensors(tensors, buffer_size): + flat_tensors = _flatten_dense_tensors(tensors) + dist.broadcast(flat_tensors, 0) + for tensor, synced in zip( + tensors, _unflatten_dense_tensors(flat_tensors, tensors)): + tensor.copy_(synced) + + def _sync_params(self): + module_states = list(self.module.state_dict().values()) + if len(module_states) > 0: + self._dist_broadcast_coalesced(module_states, + self.broadcast_bucket_size) + if self.broadcast_buffers: + if (TORCH_VERSION != 'parrots' + and digit_version(TORCH_VERSION) < digit_version('1.0')): + buffers = [b.data for b in self.module._all_buffers()] + else: + buffers = [b.data for b in self.module.buffers()] + if len(buffers) > 0: + self._dist_broadcast_coalesced(buffers, + self.broadcast_bucket_size) + + def scatter(self, inputs, kwargs, device_ids): + return scatter_kwargs(inputs, kwargs, device_ids, dim=self.dim) + + def forward(self, *inputs, **kwargs): + inputs, kwargs = self.scatter(inputs, kwargs, + [torch.cuda.current_device()]) + return self.module(*inputs[0], **kwargs[0]) + + def train_step(self, *inputs, **kwargs): + inputs, kwargs = self.scatter(inputs, kwargs, + [torch.cuda.current_device()]) + output = self.module.train_step(*inputs[0], **kwargs[0]) + return output + + def val_step(self, *inputs, **kwargs): + inputs, kwargs = self.scatter(inputs, kwargs, + [torch.cuda.current_device()]) + output = self.module.val_step(*inputs[0], **kwargs[0]) + return output diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/registry.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/registry.py new file mode 100644 index 0000000000000000000000000000000000000000..144f9fb168a45bfe3dd0abde9886be41174121a1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/registry.py @@ -0,0 +1,8 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from torch.nn.parallel import DataParallel, DistributedDataParallel + +from mmcv.utils import Registry + +MODULE_WRAPPERS = Registry('module wrapper') +MODULE_WRAPPERS.register_module(module=DataParallel) +MODULE_WRAPPERS.register_module(module=DistributedDataParallel) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/scatter_gather.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/scatter_gather.py new file mode 100644 index 0000000000000000000000000000000000000000..900ff88566f8f14830590459dc4fd16d4b382e47 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/scatter_gather.py @@ -0,0 +1,59 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +from torch.nn.parallel._functions import Scatter as OrigScatter + +from ._functions import Scatter +from .data_container import DataContainer + + +def scatter(inputs, target_gpus, dim=0): + """Scatter inputs to target gpus. + + The only difference from original :func:`scatter` is to add support for + :type:`~mmcv.parallel.DataContainer`. + """ + + def scatter_map(obj): + if isinstance(obj, torch.Tensor): + if target_gpus != [-1]: + return OrigScatter.apply(target_gpus, None, dim, obj) + else: + # for CPU inference we use self-implemented scatter + return Scatter.forward(target_gpus, obj) + if isinstance(obj, DataContainer): + if obj.cpu_only: + return obj.data + else: + return Scatter.forward(target_gpus, obj.data) + if isinstance(obj, tuple) and len(obj) > 0: + return list(zip(*map(scatter_map, obj))) + if isinstance(obj, list) and len(obj) > 0: + out = list(map(list, zip(*map(scatter_map, obj)))) + return out + if isinstance(obj, dict) and len(obj) > 0: + out = list(map(type(obj), zip(*map(scatter_map, obj.items())))) + return out + return [obj for targets in target_gpus] + + # After scatter_map is called, a scatter_map cell will exist. This cell + # has a reference to the actual function scatter_map, which has references + # to a closure that has a reference to the scatter_map cell (because the + # fn is recursive). To avoid this reference cycle, we set the function to + # None, clearing the cell + try: + return scatter_map(inputs) + finally: + scatter_map = None + + +def scatter_kwargs(inputs, kwargs, target_gpus, dim=0): + """Scatter with support for kwargs dictionary.""" + inputs = scatter(inputs, target_gpus, dim) if inputs else [] + kwargs = scatter(kwargs, target_gpus, dim) if kwargs else [] + if len(inputs) < len(kwargs): + inputs.extend([() for _ in range(len(kwargs) - len(inputs))]) + elif len(kwargs) < len(inputs): + kwargs.extend([{} for _ in range(len(inputs) - len(kwargs))]) + inputs = tuple(inputs) + kwargs = tuple(kwargs) + return inputs, kwargs diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/utils.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/utils.py new file mode 100644 index 0000000000000000000000000000000000000000..0f5712cb42c38a2e8563bf563efb6681383cab9b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/parallel/utils.py @@ -0,0 +1,20 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .registry import MODULE_WRAPPERS + + +def is_module_wrapper(module): + """Check if a module is a module wrapper. + + The following 3 modules in MMCV (and their subclasses) are regarded as + module wrappers: DataParallel, DistributedDataParallel, + MMDistributedDataParallel (the deprecated version). You may add you own + module wrapper by registering it to mmcv.parallel.MODULE_WRAPPERS. + + Args: + module (nn.Module): The module to be checked. + + Returns: + bool: True if the input module is a module wrapper. + """ + module_wrappers = tuple(MODULE_WRAPPERS.module_dict.values()) + return isinstance(module, module_wrappers) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..183d536727e4a217936378d74b3a64ef4a6377e8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/__init__.py @@ -0,0 +1,73 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .base_module import BaseModule, ModuleDict, ModuleList, Sequential +from .base_runner import BaseRunner +from .builder import RUNNERS, build_runner +from .checkpoint import (CheckpointLoader, _load_checkpoint, + _load_checkpoint_with_prefix, load_checkpoint, + load_state_dict, save_checkpoint, weights_to_cpu) +from .default_constructor import DefaultRunnerConstructor +from .dist_utils import (allreduce_grads, allreduce_params, get_dist_info, + init_dist, master_only) +from .epoch_based_runner import EpochBasedRunner, Runner +from .fp16_utils import LossScaler, auto_fp16, force_fp32, wrap_fp16_model +from .hooks import (HOOKS, CheckpointHook, ClearMLLoggerHook, ClosureHook, + DistEvalHook, DistSamplerSeedHook, DvcliveLoggerHook, + EMAHook, EvalHook, Fp16OptimizerHook, + GradientCumulativeFp16OptimizerHook, + GradientCumulativeOptimizerHook, Hook, IterTimerHook, + LoggerHook, MlflowLoggerHook, NeptuneLoggerHook, + OptimizerHook, PaviLoggerHook, SegmindLoggerHook, + SyncBuffersHook, TensorboardLoggerHook, TextLoggerHook, + WandbLoggerHook) +from .hooks.lr_updater import StepLrUpdaterHook # noqa +from .hooks.lr_updater import (CosineAnnealingLrUpdaterHook, + CosineRestartLrUpdaterHook, CyclicLrUpdaterHook, + ExpLrUpdaterHook, FixedLrUpdaterHook, + FlatCosineAnnealingLrUpdaterHook, + InvLrUpdaterHook, LinearAnnealingLrUpdaterHook, + LrUpdaterHook, OneCycleLrUpdaterHook, + PolyLrUpdaterHook) +from .hooks.momentum_updater import (CosineAnnealingMomentumUpdaterHook, + CyclicMomentumUpdaterHook, + LinearAnnealingMomentumUpdaterHook, + MomentumUpdaterHook, + OneCycleMomentumUpdaterHook, + StepMomentumUpdaterHook) +from .iter_based_runner import IterBasedRunner, IterLoader +from .log_buffer import LogBuffer +from .optimizer import (OPTIMIZER_BUILDERS, OPTIMIZERS, + DefaultOptimizerConstructor, build_optimizer, + build_optimizer_constructor) +from .priority import Priority, get_priority +from .utils import get_host_info, get_time_str, obj_from_dict, set_random_seed + +# initialize ipu to registor ipu runner to RUNNERS +from mmcv.device import ipu # isort:skip # noqa + +__all__ = [ + 'BaseRunner', 'Runner', 'EpochBasedRunner', 'IterBasedRunner', 'LogBuffer', + 'HOOKS', 'Hook', 'CheckpointHook', 'ClosureHook', 'LrUpdaterHook', + 'FixedLrUpdaterHook', 'StepLrUpdaterHook', 'ExpLrUpdaterHook', + 'PolyLrUpdaterHook', 'InvLrUpdaterHook', 'CosineAnnealingLrUpdaterHook', + 'FlatCosineAnnealingLrUpdaterHook', 'CosineRestartLrUpdaterHook', + 'CyclicLrUpdaterHook', 'OneCycleLrUpdaterHook', 'MomentumUpdaterHook', + 'StepMomentumUpdaterHook', 'CosineAnnealingMomentumUpdaterHook', + 'CyclicMomentumUpdaterHook', 'OneCycleMomentumUpdaterHook', + 'OptimizerHook', 'IterTimerHook', 'DistSamplerSeedHook', 'LoggerHook', + 'PaviLoggerHook', 'TextLoggerHook', 'TensorboardLoggerHook', + 'NeptuneLoggerHook', 'WandbLoggerHook', 'MlflowLoggerHook', + 'DvcliveLoggerHook', '_load_checkpoint', 'load_state_dict', + 'load_checkpoint', 'weights_to_cpu', 'save_checkpoint', 'Priority', + 'get_priority', 'get_host_info', 'get_time_str', 'obj_from_dict', + 'init_dist', 'get_dist_info', 'master_only', 'OPTIMIZER_BUILDERS', + 'OPTIMIZERS', 'DefaultOptimizerConstructor', 'build_optimizer', + 'build_optimizer_constructor', 'IterLoader', 'set_random_seed', + 'auto_fp16', 'force_fp32', 'wrap_fp16_model', 'Fp16OptimizerHook', + 'SyncBuffersHook', 'EMAHook', 'build_runner', 'RUNNERS', 'allreduce_grads', + 'allreduce_params', 'LossScaler', 'CheckpointLoader', 'BaseModule', + '_load_checkpoint_with_prefix', 'EvalHook', 'DistEvalHook', 'Sequential', + 'ModuleDict', 'ModuleList', 'GradientCumulativeOptimizerHook', + 'GradientCumulativeFp16OptimizerHook', 'DefaultRunnerConstructor', + 'SegmindLoggerHook', 'LinearAnnealingMomentumUpdaterHook', + 'LinearAnnealingLrUpdaterHook', 'ClearMLLoggerHook' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/base_module.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/base_module.py new file mode 100644 index 0000000000000000000000000000000000000000..7937eca379024f5d6062dc3ba8f4450e8cfc40fb --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/base_module.py @@ -0,0 +1,208 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import warnings +from abc import ABCMeta +from collections import defaultdict +from logging import FileHandler + +import torch.nn as nn + +from mmcv.runner.dist_utils import master_only +from mmcv.utils.logging import get_logger, logger_initialized, print_log + + +class BaseModule(nn.Module, metaclass=ABCMeta): + """Base module for all modules in openmmlab. + + ``BaseModule`` is a wrapper of ``torch.nn.Module`` with additional + functionality of parameter initialization. Compared with + ``torch.nn.Module``, ``BaseModule`` mainly adds three attributes. + + - ``init_cfg``: the config to control the initialization. + - ``init_weights``: The function of parameter initialization and recording + initialization information. + - ``_params_init_info``: Used to track the parameter initialization + information. This attribute only exists during executing the + ``init_weights``. + + Args: + init_cfg (dict, optional): Initialization config dict. + """ + + def __init__(self, init_cfg=None): + """Initialize BaseModule, inherited from `torch.nn.Module`""" + + # NOTE init_cfg can be defined in different levels, but init_cfg + # in low levels has a higher priority. + + super(BaseModule, self).__init__() + # define default value of init_cfg instead of hard code + # in init_weights() function + self._is_init = False + + self.init_cfg = copy.deepcopy(init_cfg) + + # Backward compatibility in derived classes + # if pretrained is not None: + # warnings.warn('DeprecationWarning: pretrained is a deprecated \ + # key, please consider using init_cfg') + # self.init_cfg = dict(type='Pretrained', checkpoint=pretrained) + + @property + def is_init(self): + return self._is_init + + def init_weights(self): + """Initialize the weights.""" + + is_top_level_module = False + # check if it is top-level module + if not hasattr(self, '_params_init_info'): + # The `_params_init_info` is used to record the initialization + # information of the parameters + # the key should be the obj:`nn.Parameter` of model and the value + # should be a dict containing + # - init_info (str): The string that describes the initialization. + # - tmp_mean_value (FloatTensor): The mean of the parameter, + # which indicates whether the parameter has been modified. + # this attribute would be deleted after all parameters + # is initialized. + self._params_init_info = defaultdict(dict) + is_top_level_module = True + + # Initialize the `_params_init_info`, + # When detecting the `tmp_mean_value` of + # the corresponding parameter is changed, update related + # initialization information + for name, param in self.named_parameters(): + self._params_init_info[param][ + 'init_info'] = f'The value is the same before and ' \ + f'after calling `init_weights` ' \ + f'of {self.__class__.__name__} ' + self._params_init_info[param][ + 'tmp_mean_value'] = param.data.mean() + + # pass `params_init_info` to all submodules + # All submodules share the same `params_init_info`, + # so it will be updated when parameters are + # modified at any level of the model. + for sub_module in self.modules(): + sub_module._params_init_info = self._params_init_info + + # Get the initialized logger, if not exist, + # create a logger named `mmcv` + logger_names = list(logger_initialized.keys()) + logger_name = logger_names[0] if logger_names else 'mmcv' + + from ..cnn import initialize + from ..cnn.utils.weight_init import update_init_info + module_name = self.__class__.__name__ + if not self._is_init: + if self.init_cfg: + print_log( + f'initialize {module_name} with init_cfg {self.init_cfg}', + logger=logger_name) + initialize(self, self.init_cfg) + if isinstance(self.init_cfg, dict): + # prevent the parameters of + # the pre-trained model + # from being overwritten by + # the `init_weights` + if self.init_cfg['type'] == 'Pretrained': + return + + for m in self.children(): + if hasattr(m, 'init_weights'): + m.init_weights() + # users may overload the `init_weights` + update_init_info( + m, + init_info=f'Initialized by ' + f'user-defined `init_weights`' + f' in {m.__class__.__name__} ') + + self._is_init = True + else: + warnings.warn(f'init_weights of {self.__class__.__name__} has ' + f'been called more than once.') + + if is_top_level_module: + self._dump_init_info(logger_name) + + for sub_module in self.modules(): + del sub_module._params_init_info + + @master_only + def _dump_init_info(self, logger_name): + """Dump the initialization information to a file named + `initialization.log.json` in workdir. + + Args: + logger_name (str): The name of logger. + """ + + logger = get_logger(logger_name) + + with_file_handler = False + # dump the information to the logger file if there is a `FileHandler` + for handler in logger.handlers: + if isinstance(handler, FileHandler): + handler.stream.write( + 'Name of parameter - Initialization information\n') + for name, param in self.named_parameters(): + handler.stream.write( + f'\n{name} - {param.shape}: ' + f"\n{self._params_init_info[param]['init_info']} \n") + handler.stream.flush() + with_file_handler = True + if not with_file_handler: + for name, param in self.named_parameters(): + print_log( + f'\n{name} - {param.shape}: ' + f"\n{self._params_init_info[param]['init_info']} \n ", + logger=logger_name) + + def __repr__(self): + s = super().__repr__() + if self.init_cfg: + s += f'\ninit_cfg={self.init_cfg}' + return s + + +class Sequential(BaseModule, nn.Sequential): + """Sequential module in openmmlab. + + Args: + init_cfg (dict, optional): Initialization config dict. + """ + + def __init__(self, *args, init_cfg=None): + BaseModule.__init__(self, init_cfg) + nn.Sequential.__init__(self, *args) + + +class ModuleList(BaseModule, nn.ModuleList): + """ModuleList in openmmlab. + + Args: + modules (iterable, optional): an iterable of modules to add. + init_cfg (dict, optional): Initialization config dict. + """ + + def __init__(self, modules=None, init_cfg=None): + BaseModule.__init__(self, init_cfg) + nn.ModuleList.__init__(self, modules) + + +class ModuleDict(BaseModule, nn.ModuleDict): + """ModuleDict in openmmlab. + + Args: + modules (dict, optional): a mapping (dictionary) of (string: module) + or an iterable of key-value pairs of type (string, module). + init_cfg (dict, optional): Initialization config dict. + """ + + def __init__(self, modules=None, init_cfg=None): + BaseModule.__init__(self, init_cfg) + nn.ModuleDict.__init__(self, modules) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/base_runner.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/base_runner.py new file mode 100644 index 0000000000000000000000000000000000000000..12a0025f89b9879d143f0fda590b2997f4607bc2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/base_runner.py @@ -0,0 +1,544 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import logging +import os.path as osp +import warnings +from abc import ABCMeta, abstractmethod + +import torch +from torch.optim import Optimizer + +import mmcv +from ..parallel import is_module_wrapper +from .checkpoint import load_checkpoint +from .dist_utils import get_dist_info +from .hooks import HOOKS, Hook +from .log_buffer import LogBuffer +from .priority import Priority, get_priority +from .utils import get_time_str + + +class BaseRunner(metaclass=ABCMeta): + """The base class of Runner, a training helper for PyTorch. + + All subclasses should implement the following APIs: + + - ``run()`` + - ``train()`` + - ``val()`` + - ``save_checkpoint()`` + + Args: + model (:obj:`torch.nn.Module`): The model to be run. + batch_processor (callable): A callable method that process a data + batch. The interface of this method should be + `batch_processor(model, data, train_mode) -> dict` + optimizer (dict or :obj:`torch.optim.Optimizer`): It can be either an + optimizer (in most cases) or a dict of optimizers (in models that + requires more than one optimizer, e.g., GAN). + work_dir (str, optional): The working directory to save checkpoints + and logs. Defaults to None. + logger (:obj:`logging.Logger`): Logger used during training. + Defaults to None. (The default value is just for backward + compatibility) + meta (dict | None): A dict records some import information such as + environment info and seed, which will be logged in logger hook. + Defaults to None. + max_epochs (int, optional): Total training epochs. + max_iters (int, optional): Total training iterations. + """ + + def __init__(self, + model, + batch_processor=None, + optimizer=None, + work_dir=None, + logger=None, + meta=None, + max_iters=None, + max_epochs=None): + if batch_processor is not None: + if not callable(batch_processor): + raise TypeError('batch_processor must be callable, ' + f'but got {type(batch_processor)}') + warnings.warn( + 'batch_processor is deprecated, please implement ' + 'train_step() and val_step() in the model instead.', + DeprecationWarning) + # raise an error is `batch_processor` is not None and + # `model.train_step()` exists. + if is_module_wrapper(model): + _model = model.module + else: + _model = model + if hasattr(_model, 'train_step') or hasattr(_model, 'val_step'): + raise RuntimeError( + 'batch_processor and model.train_step()/model.val_step() ' + 'cannot be both available.') + else: + assert hasattr(model, 'train_step') + + # check the type of `optimizer` + if isinstance(optimizer, dict): + for name, optim in optimizer.items(): + if not isinstance(optim, Optimizer): + raise TypeError( + f'optimizer must be a dict of torch.optim.Optimizers, ' + f'but optimizer["{name}"] is a {type(optim)}') + elif not isinstance(optimizer, Optimizer) and optimizer is not None: + raise TypeError( + f'optimizer must be a torch.optim.Optimizer object ' + f'or dict or None, but got {type(optimizer)}') + + # check the type of `logger` + if not isinstance(logger, logging.Logger): + raise TypeError(f'logger must be a logging.Logger object, ' + f'but got {type(logger)}') + + # check the type of `meta` + if meta is not None and not isinstance(meta, dict): + raise TypeError( + f'meta must be a dict or None, but got {type(meta)}') + + self.model = model + self.batch_processor = batch_processor + self.optimizer = optimizer + self.logger = logger + self.meta = meta + # create work_dir + if mmcv.is_str(work_dir): + self.work_dir = osp.abspath(work_dir) + mmcv.mkdir_or_exist(self.work_dir) + elif work_dir is None: + self.work_dir = None + else: + raise TypeError('"work_dir" must be a str or None') + + # get model name from the model class + if hasattr(self.model, 'module'): + self._model_name = self.model.module.__class__.__name__ + else: + self._model_name = self.model.__class__.__name__ + + self._rank, self._world_size = get_dist_info() + self.timestamp = get_time_str() + self.mode = None + self._hooks = [] + self._epoch = 0 + self._iter = 0 + self._inner_iter = 0 + + if max_epochs is not None and max_iters is not None: + raise ValueError( + 'Only one of `max_epochs` or `max_iters` can be set.') + + self._max_epochs = max_epochs + self._max_iters = max_iters + # TODO: Redesign LogBuffer, it is not flexible and elegant enough + self.log_buffer = LogBuffer() + + @property + def model_name(self): + """str: Name of the model, usually the module class name.""" + return self._model_name + + @property + def rank(self): + """int: Rank of current process. (distributed training)""" + return self._rank + + @property + def world_size(self): + """int: Number of processes participating in the job. + (distributed training)""" + return self._world_size + + @property + def hooks(self): + """list[:obj:`Hook`]: A list of registered hooks.""" + return self._hooks + + @property + def epoch(self): + """int: Current epoch.""" + return self._epoch + + @property + def iter(self): + """int: Current iteration.""" + return self._iter + + @property + def inner_iter(self): + """int: Iteration in an epoch.""" + return self._inner_iter + + @property + def max_epochs(self): + """int: Maximum training epochs.""" + return self._max_epochs + + @property + def max_iters(self): + """int: Maximum training iterations.""" + return self._max_iters + + @abstractmethod + def train(self): + pass + + @abstractmethod + def val(self): + pass + + @abstractmethod + def run(self, data_loaders, workflow, **kwargs): + pass + + @abstractmethod + def save_checkpoint(self, + out_dir, + filename_tmpl, + save_optimizer=True, + meta=None, + create_symlink=True): + pass + + def current_lr(self): + """Get current learning rates. + + Returns: + list[float] | dict[str, list[float]]: Current learning rates of all + param groups. If the runner has a dict of optimizers, this method + will return a dict. + """ + if isinstance(self.optimizer, torch.optim.Optimizer): + lr = [group['lr'] for group in self.optimizer.param_groups] + elif isinstance(self.optimizer, dict): + lr = dict() + for name, optim in self.optimizer.items(): + lr[name] = [group['lr'] for group in optim.param_groups] + else: + raise RuntimeError( + 'lr is not applicable because optimizer does not exist.') + return lr + + def current_momentum(self): + """Get current momentums. + + Returns: + list[float] | dict[str, list[float]]: Current momentums of all + param groups. If the runner has a dict of optimizers, this method + will return a dict. + """ + + def _get_momentum(optimizer): + momentums = [] + for group in optimizer.param_groups: + if 'momentum' in group.keys(): + momentums.append(group['momentum']) + elif 'betas' in group.keys(): + momentums.append(group['betas'][0]) + else: + momentums.append(0) + return momentums + + if self.optimizer is None: + raise RuntimeError( + 'momentum is not applicable because optimizer does not exist.') + elif isinstance(self.optimizer, torch.optim.Optimizer): + momentums = _get_momentum(self.optimizer) + elif isinstance(self.optimizer, dict): + momentums = dict() + for name, optim in self.optimizer.items(): + momentums[name] = _get_momentum(optim) + return momentums + + def register_hook(self, hook, priority='NORMAL'): + """Register a hook into the hook list. + + The hook will be inserted into a priority queue, with the specified + priority (See :class:`Priority` for details of priorities). + For hooks with the same priority, they will be triggered in the same + order as they are registered. + + Args: + hook (:obj:`Hook`): The hook to be registered. + priority (int or str or :obj:`Priority`): Hook priority. + Lower value means higher priority. + """ + assert isinstance(hook, Hook) + if hasattr(hook, 'priority'): + raise ValueError('"priority" is a reserved attribute for hooks') + priority = get_priority(priority) + hook.priority = priority + # insert the hook to a sorted list + inserted = False + for i in range(len(self._hooks) - 1, -1, -1): + if priority >= self._hooks[i].priority: + self._hooks.insert(i + 1, hook) + inserted = True + break + if not inserted: + self._hooks.insert(0, hook) + + def register_hook_from_cfg(self, hook_cfg): + """Register a hook from its cfg. + + Args: + hook_cfg (dict): Hook config. It should have at least keys 'type' + and 'priority' indicating its type and priority. + + Note: + The specific hook class to register should not use 'type' and + 'priority' arguments during initialization. + """ + hook_cfg = hook_cfg.copy() + priority = hook_cfg.pop('priority', 'NORMAL') + hook = mmcv.build_from_cfg(hook_cfg, HOOKS) + self.register_hook(hook, priority=priority) + + def call_hook(self, fn_name): + """Call all hooks. + + Args: + fn_name (str): The function name in each hook to be called, such as + "before_train_epoch". + """ + for hook in self._hooks: + getattr(hook, fn_name)(self) + + def get_hook_info(self): + # Get hooks info in each stage + stage_hook_map = {stage: [] for stage in Hook.stages} + for hook in self.hooks: + try: + priority = Priority(hook.priority).name + except ValueError: + priority = hook.priority + classname = hook.__class__.__name__ + hook_info = f'({priority:<12}) {classname:<35}' + for trigger_stage in hook.get_triggered_stages(): + stage_hook_map[trigger_stage].append(hook_info) + + stage_hook_infos = [] + for stage in Hook.stages: + hook_infos = stage_hook_map[stage] + if len(hook_infos) > 0: + info = f'{stage}:\n' + info += '\n'.join(hook_infos) + info += '\n -------------------- ' + stage_hook_infos.append(info) + return '\n'.join(stage_hook_infos) + + def load_checkpoint(self, + filename, + map_location='cpu', + strict=False, + revise_keys=[(r'^module.', '')]): + return load_checkpoint( + self.model, + filename, + map_location, + strict, + self.logger, + revise_keys=revise_keys) + + def resume(self, + checkpoint, + resume_optimizer=True, + map_location='default'): + if map_location == 'default': + if torch.cuda.is_available(): + device_id = torch.cuda.current_device() + checkpoint = self.load_checkpoint( + checkpoint, + map_location=lambda storage, loc: storage.cuda(device_id)) + else: + checkpoint = self.load_checkpoint(checkpoint) + else: + checkpoint = self.load_checkpoint( + checkpoint, map_location=map_location) + + self._epoch = checkpoint['meta']['epoch'] + self._iter = checkpoint['meta']['iter'] + if self.meta is None: + self.meta = {} + self.meta.setdefault('hook_msgs', {}) + # load `last_ckpt`, `best_score`, `best_ckpt`, etc. for hook messages + self.meta['hook_msgs'].update(checkpoint['meta'].get('hook_msgs', {})) + + # Re-calculate the number of iterations when resuming + # models with different number of GPUs + if 'config' in checkpoint['meta']: + config = mmcv.Config.fromstring( + checkpoint['meta']['config'], file_format='.py') + previous_gpu_ids = config.get('gpu_ids', None) + if previous_gpu_ids and len(previous_gpu_ids) > 0 and len( + previous_gpu_ids) != self.world_size: + self._iter = int(self._iter * len(previous_gpu_ids) / + self.world_size) + self.logger.info('the iteration number is changed due to ' + 'change of GPU number') + + # resume meta information meta + self.meta = checkpoint['meta'] + + if 'optimizer' in checkpoint and resume_optimizer: + if isinstance(self.optimizer, Optimizer): + self.optimizer.load_state_dict(checkpoint['optimizer']) + elif isinstance(self.optimizer, dict): + for k in self.optimizer.keys(): + self.optimizer[k].load_state_dict( + checkpoint['optimizer'][k]) + else: + raise TypeError( + 'Optimizer should be dict or torch.optim.Optimizer ' + f'but got {type(self.optimizer)}') + + self.logger.info('resumed epoch %d, iter %d', self.epoch, self.iter) + + def register_lr_hook(self, lr_config): + if lr_config is None: + return + elif isinstance(lr_config, dict): + assert 'policy' in lr_config + policy_type = lr_config.pop('policy') + # If the type of policy is all in lower case, e.g., 'cyclic', + # then its first letter will be capitalized, e.g., to be 'Cyclic'. + # This is for the convenient usage of Lr updater. + # Since this is not applicable for ` + # CosineAnnealingLrUpdater`, + # the string will not be changed if it contains capital letters. + if policy_type == policy_type.lower(): + policy_type = policy_type.title() + hook_type = policy_type + 'LrUpdaterHook' + lr_config['type'] = hook_type + hook = mmcv.build_from_cfg(lr_config, HOOKS) + else: + hook = lr_config + self.register_hook(hook, priority='VERY_HIGH') + + def register_momentum_hook(self, momentum_config): + if momentum_config is None: + return + if isinstance(momentum_config, dict): + assert 'policy' in momentum_config + policy_type = momentum_config.pop('policy') + # If the type of policy is all in lower case, e.g., 'cyclic', + # then its first letter will be capitalized, e.g., to be 'Cyclic'. + # This is for the convenient usage of momentum updater. + # Since this is not applicable for + # `CosineAnnealingMomentumUpdater`, + # the string will not be changed if it contains capital letters. + if policy_type == policy_type.lower(): + policy_type = policy_type.title() + hook_type = policy_type + 'MomentumUpdaterHook' + momentum_config['type'] = hook_type + hook = mmcv.build_from_cfg(momentum_config, HOOKS) + else: + hook = momentum_config + self.register_hook(hook, priority='HIGH') + + def register_optimizer_hook(self, optimizer_config): + if optimizer_config is None: + return + if isinstance(optimizer_config, dict): + optimizer_config.setdefault('type', 'OptimizerHook') + hook = mmcv.build_from_cfg(optimizer_config, HOOKS) + else: + hook = optimizer_config + self.register_hook(hook, priority='ABOVE_NORMAL') + + def register_checkpoint_hook(self, checkpoint_config): + if checkpoint_config is None: + return + if isinstance(checkpoint_config, dict): + checkpoint_config.setdefault('type', 'CheckpointHook') + hook = mmcv.build_from_cfg(checkpoint_config, HOOKS) + else: + hook = checkpoint_config + self.register_hook(hook, priority='NORMAL') + + def register_logger_hooks(self, log_config): + if log_config is None: + return + log_interval = log_config['interval'] + for info in log_config['hooks']: + logger_hook = mmcv.build_from_cfg( + info, HOOKS, default_args=dict(interval=log_interval)) + self.register_hook(logger_hook, priority='VERY_LOW') + + def register_timer_hook(self, timer_config): + if timer_config is None: + return + if isinstance(timer_config, dict): + timer_config_ = copy.deepcopy(timer_config) + hook = mmcv.build_from_cfg(timer_config_, HOOKS) + else: + hook = timer_config + self.register_hook(hook, priority='LOW') + + def register_custom_hooks(self, custom_config): + if custom_config is None: + return + + if not isinstance(custom_config, list): + custom_config = [custom_config] + + for item in custom_config: + if isinstance(item, dict): + self.register_hook_from_cfg(item) + else: + self.register_hook(item, priority='NORMAL') + + def register_profiler_hook(self, profiler_config): + if profiler_config is None: + return + if isinstance(profiler_config, dict): + profiler_config.setdefault('type', 'ProfilerHook') + hook = mmcv.build_from_cfg(profiler_config, HOOKS) + else: + hook = profiler_config + self.register_hook(hook) + + def register_training_hooks(self, + lr_config, + optimizer_config=None, + checkpoint_config=None, + log_config=None, + momentum_config=None, + timer_config=dict(type='IterTimerHook'), + custom_hooks_config=None): + """Register default and custom hooks for training. + + Default and custom hooks include: + + +----------------------+-------------------------+ + | Hooks | Priority | + +======================+=========================+ + | LrUpdaterHook | VERY_HIGH (10) | + +----------------------+-------------------------+ + | MomentumUpdaterHook | HIGH (30) | + +----------------------+-------------------------+ + | OptimizerStepperHook | ABOVE_NORMAL (40) | + +----------------------+-------------------------+ + | CheckpointSaverHook | NORMAL (50) | + +----------------------+-------------------------+ + | IterTimerHook | LOW (70) | + +----------------------+-------------------------+ + | LoggerHook(s) | VERY_LOW (90) | + +----------------------+-------------------------+ + | CustomHook(s) | defaults to NORMAL (50) | + +----------------------+-------------------------+ + + If custom hooks have same priority with default hooks, custom hooks + will be triggered after default hooks. + """ + self.register_lr_hook(lr_config) + self.register_momentum_hook(momentum_config) + self.register_optimizer_hook(optimizer_config) + self.register_checkpoint_hook(checkpoint_config) + self.register_timer_hook(timer_config) + self.register_logger_hooks(log_config) + self.register_custom_hooks(custom_hooks_config) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/builder.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/builder.py new file mode 100644 index 0000000000000000000000000000000000000000..77c96ba0b2f30ead9da23f293c5dc84dd3e4a74f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/builder.py @@ -0,0 +1,24 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy + +from ..utils import Registry + +RUNNERS = Registry('runner') +RUNNER_BUILDERS = Registry('runner builder') + + +def build_runner_constructor(cfg): + return RUNNER_BUILDERS.build(cfg) + + +def build_runner(cfg, default_args=None): + runner_cfg = copy.deepcopy(cfg) + constructor_type = runner_cfg.pop('constructor', + 'DefaultRunnerConstructor') + runner_constructor = build_runner_constructor( + dict( + type=constructor_type, + runner_cfg=runner_cfg, + default_args=default_args)) + runner = runner_constructor() + return runner diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/checkpoint.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/checkpoint.py new file mode 100644 index 0000000000000000000000000000000000000000..835ee725a035d6bb530e544ed8d381043e2397b1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/checkpoint.py @@ -0,0 +1,759 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import io +import os +import os.path as osp +import pkgutil +import re +import time +import warnings +from collections import OrderedDict +from importlib import import_module +from tempfile import TemporaryDirectory + +import torch +import torchvision +from torch.optim import Optimizer + +import mmcv +from ..fileio import FileClient +from ..fileio import load as load_file +from ..parallel import is_module_wrapper +from ..utils import digit_version, load_url, mkdir_or_exist +from .dist_utils import get_dist_info + +ENV_MMCV_HOME = 'MMCV_HOME' +ENV_XDG_CACHE_HOME = 'XDG_CACHE_HOME' +DEFAULT_CACHE_DIR = '~/.cache' + + +def _get_mmcv_home(): + mmcv_home = os.path.expanduser( + os.getenv( + ENV_MMCV_HOME, + os.path.join( + os.getenv(ENV_XDG_CACHE_HOME, DEFAULT_CACHE_DIR), 'mmcv'))) + + mkdir_or_exist(mmcv_home) + return mmcv_home + + +def load_state_dict(module, state_dict, strict=False, logger=None): + """Load state_dict to a module. + + This method is modified from :meth:`torch.nn.Module.load_state_dict`. + Default value for ``strict`` is set to ``False`` and the message for + param mismatch will be shown even if strict is False. + + Args: + module (Module): Module that receives the state_dict. + state_dict (OrderedDict): Weights. + strict (bool): whether to strictly enforce that the keys + in :attr:`state_dict` match the keys returned by this module's + :meth:`~torch.nn.Module.state_dict` function. Default: ``False``. + logger (:obj:`logging.Logger`, optional): Logger to log the error + message. If not specified, print function will be used. + """ + unexpected_keys = [] + all_missing_keys = [] + err_msg = [] + + metadata = getattr(state_dict, '_metadata', None) + state_dict = state_dict.copy() + if metadata is not None: + state_dict._metadata = metadata + + # use _load_from_state_dict to enable checkpoint version control + def load(module, prefix=''): + # recursively check parallel module in case that the model has a + # complicated structure, e.g., nn.Module(nn.Module(DDP)) + if is_module_wrapper(module): + module = module.module + local_metadata = {} if metadata is None else metadata.get( + prefix[:-1], {}) + module._load_from_state_dict(state_dict, prefix, local_metadata, True, + all_missing_keys, unexpected_keys, + err_msg) + for name, child in module._modules.items(): + if child is not None: + load(child, prefix + name + '.') + + load(module) + load = None # break load->load reference cycle + + # ignore "num_batches_tracked" of BN layers + missing_keys = [ + key for key in all_missing_keys if 'num_batches_tracked' not in key + ] + + if unexpected_keys: + err_msg.append('unexpected key in source ' + f'state_dict: {", ".join(unexpected_keys)}\n') + if missing_keys: + err_msg.append( + f'missing keys in source state_dict: {", ".join(missing_keys)}\n') + + rank, _ = get_dist_info() + if len(err_msg) > 0 and rank == 0: + err_msg.insert( + 0, 'The model and loaded state dict do not match exactly\n') + err_msg = '\n'.join(err_msg) + if strict: + raise RuntimeError(err_msg) + elif logger is not None: + logger.warning(err_msg) + else: + print(err_msg) + + +def get_torchvision_models(): + if digit_version(torchvision.__version__) < digit_version('0.13.0a0'): + model_urls = dict() + # When the version of torchvision is lower than 0.13, the model url is + # not declared in `torchvision.model.__init__.py`, so we need to + # iterate through `torchvision.models.__path__` to get the url for each + # model. + for _, name, ispkg in pkgutil.walk_packages( + torchvision.models.__path__): + if ispkg: + continue + _zoo = import_module(f'torchvision.models.{name}') + if hasattr(_zoo, 'model_urls'): + _urls = getattr(_zoo, 'model_urls') + model_urls.update(_urls) + else: + # Since torchvision bumps to v0.13, the weight loading logic, + # model keys and model urls have been changed. Here the URLs of old + # version is loaded to avoid breaking back compatibility. If the + # torchvision version>=0.13.0, new URLs will be added. Users can get + # the resnet50 checkpoint by setting 'resnet50.imagent1k_v1', + # 'resnet50' or 'ResNet50_Weights.IMAGENET1K_V1' in the config. + json_path = osp.join(mmcv.__path__[0], + 'model_zoo/torchvision_0.12.json') + model_urls = mmcv.load(json_path) + for cls_name, cls in torchvision.models.__dict__.items(): + # The name of torchvision model weights classes ends with + # `_Weights` such as `ResNet18_Weights`. However, some model weight + # classes, such as `MNASNet0_75_Weights` does not have any urls in + # torchvision 0.13.0 and cannot be iterated. Here we simply check + # `DEFAULT` attribute to ensure the class is not empty. + if (not cls_name.endswith('_Weights') + or not hasattr(cls, 'DEFAULT')): + continue + # Since `cls.DEFAULT` can not be accessed by iterating cls, we set + # default urls explicitly. + cls_key = cls_name.replace('_Weights', '').lower() + model_urls[f'{cls_key}.default'] = cls.DEFAULT.url + for weight_enum in cls: + cls_key = cls_name.replace('_Weights', '').lower() + cls_key = f'{cls_key}.{weight_enum.name.lower()}' + model_urls[cls_key] = weight_enum.url + + return model_urls + + +def get_external_models(): + mmcv_home = _get_mmcv_home() + default_json_path = osp.join(mmcv.__path__[0], 'model_zoo/open_mmlab.json') + default_urls = load_file(default_json_path) + assert isinstance(default_urls, dict) + external_json_path = osp.join(mmcv_home, 'open_mmlab.json') + if osp.exists(external_json_path): + external_urls = load_file(external_json_path) + assert isinstance(external_urls, dict) + default_urls.update(external_urls) + + return default_urls + + +def get_mmcls_models(): + mmcls_json_path = osp.join(mmcv.__path__[0], 'model_zoo/mmcls.json') + mmcls_urls = load_file(mmcls_json_path) + + return mmcls_urls + + +def get_deprecated_model_names(): + deprecate_json_path = osp.join(mmcv.__path__[0], + 'model_zoo/deprecated.json') + deprecate_urls = load_file(deprecate_json_path) + assert isinstance(deprecate_urls, dict) + + return deprecate_urls + + +def _process_mmcls_checkpoint(checkpoint): + if 'state_dict' in checkpoint: + state_dict = checkpoint['state_dict'] + else: + # Some checkpoints converted from 3rd-party repo don't + # have the "state_dict" key. + state_dict = checkpoint + new_state_dict = OrderedDict() + for k, v in state_dict.items(): + if k.startswith('backbone.'): + new_state_dict[k[9:]] = v + new_checkpoint = dict(state_dict=new_state_dict) + + return new_checkpoint + + +class CheckpointLoader: + """A general checkpoint loader to manage all schemes.""" + + _schemes = {} + + @classmethod + def _register_scheme(cls, prefixes, loader, force=False): + if isinstance(prefixes, str): + prefixes = [prefixes] + else: + assert isinstance(prefixes, (list, tuple)) + for prefix in prefixes: + if (prefix not in cls._schemes) or force: + cls._schemes[prefix] = loader + else: + raise KeyError( + f'{prefix} is already registered as a loader backend, ' + 'add "force=True" if you want to override it') + # sort, longer prefixes take priority + cls._schemes = OrderedDict( + sorted(cls._schemes.items(), key=lambda t: t[0], reverse=True)) + + @classmethod + def register_scheme(cls, prefixes, loader=None, force=False): + """Register a loader to CheckpointLoader. + + This method can be used as a normal class method or a decorator. + + Args: + prefixes (str or list[str] or tuple[str]): + The prefix of the registered loader. + loader (function, optional): The loader function to be registered. + When this method is used as a decorator, loader is None. + Defaults to None. + force (bool, optional): Whether to override the loader + if the prefix has already been registered. Defaults to False. + """ + + if loader is not None: + cls._register_scheme(prefixes, loader, force=force) + return + + def _register(loader_cls): + cls._register_scheme(prefixes, loader_cls, force=force) + return loader_cls + + return _register + + @classmethod + def _get_checkpoint_loader(cls, path): + """Finds a loader that supports the given path. Falls back to the local + loader if no other loader is found. + + Args: + path (str): checkpoint path + + Returns: + callable: checkpoint loader + """ + for p in cls._schemes: + # use regular match to handle some cases that where the prefix of + # loader has a prefix. For example, both 's3://path' and + # 'open-mmlab:s3://path' should return `load_from_ceph` + if re.match(p, path) is not None: + return cls._schemes[p] + + @classmethod + def load_checkpoint(cls, filename, map_location=None, logger=None): + """load checkpoint through URL scheme path. + + Args: + filename (str): checkpoint file name with given prefix + map_location (str, optional): Same as :func:`torch.load`. + Default: None + logger (:mod:`logging.Logger`, optional): The logger for message. + Default: None + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + + checkpoint_loader = cls._get_checkpoint_loader(filename) + class_name = checkpoint_loader.__name__ + mmcv.print_log( + f'load checkpoint from {class_name[10:]} path: {filename}', logger) + return checkpoint_loader(filename, map_location) + + +@CheckpointLoader.register_scheme(prefixes='') +def load_from_local(filename, map_location): + """load checkpoint by local file path. + + Args: + filename (str): local checkpoint file path + map_location (str, optional): Same as :func:`torch.load`. + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + filename = osp.expanduser(filename) + if not osp.isfile(filename): + raise FileNotFoundError(f'{filename} can not be found.') + checkpoint = torch.load(filename, map_location=map_location) + return checkpoint + + +@CheckpointLoader.register_scheme(prefixes=('http://', 'https://')) +def load_from_http(filename, map_location=None, model_dir=None): + """load checkpoint through HTTP or HTTPS scheme path. In distributed + setting, this function only download checkpoint at local rank 0. + + Args: + filename (str): checkpoint file path with modelzoo or + torchvision prefix + map_location (str, optional): Same as :func:`torch.load`. + model_dir (string, optional): directory in which to save the object, + Default: None + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + rank, world_size = get_dist_info() + if rank == 0: + checkpoint = load_url( + filename, model_dir=model_dir, map_location=map_location) + if world_size > 1: + torch.distributed.barrier() + if rank > 0: + checkpoint = load_url( + filename, model_dir=model_dir, map_location=map_location) + return checkpoint + + +@CheckpointLoader.register_scheme(prefixes='pavi://') +def load_from_pavi(filename, map_location=None): + """load checkpoint through the file path prefixed with pavi. In distributed + setting, this function download ckpt at all ranks to different temporary + directories. + + Args: + filename (str): checkpoint file path with pavi prefix + map_location (str, optional): Same as :func:`torch.load`. + Default: None + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + assert filename.startswith('pavi://'), \ + f'Expected filename startswith `pavi://`, but get {filename}' + model_path = filename[7:] + + try: + from pavi import modelcloud + except ImportError: + raise ImportError( + 'Please install pavi to load checkpoint from modelcloud.') + + model = modelcloud.get(model_path) + with TemporaryDirectory() as tmp_dir: + downloaded_file = osp.join(tmp_dir, model.name) + model.download(downloaded_file) + checkpoint = torch.load(downloaded_file, map_location=map_location) + return checkpoint + + +@CheckpointLoader.register_scheme(prefixes=r'(\S+\:)?s3://') +def load_from_ceph(filename, map_location=None, backend='petrel'): + """load checkpoint through the file path prefixed with s3. In distributed + setting, this function download ckpt at all ranks to different temporary + directories. + + Note: + Since v1.4.1, the registered scheme prefixes have been enhanced to + support bucket names in the path prefix, e.g. 's3://xx.xx/xx.path', + 'bucket1:s3://xx.xx/xx.path'. + + Args: + filename (str): checkpoint file path with s3 prefix + map_location (str, optional): Same as :func:`torch.load`. + backend (str, optional): The storage backend type. Options are 'ceph', + 'petrel'. Default: 'petrel'. + + .. warning:: + :class:`mmcv.fileio.file_client.CephBackend` will be deprecated, + please use :class:`mmcv.fileio.file_client.PetrelBackend` instead. + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + allowed_backends = ['ceph', 'petrel'] + if backend not in allowed_backends: + raise ValueError(f'Load from Backend {backend} is not supported.') + + if backend == 'ceph': + warnings.warn( + 'CephBackend will be deprecated, please use PetrelBackend instead', + DeprecationWarning) + + # CephClient and PetrelBackend have the same prefix 's3://' and the latter + # will be chosen as default. If PetrelBackend can not be instantiated + # successfully, the CephClient will be chosen. + try: + file_client = FileClient(backend=backend) + except ImportError: + allowed_backends.remove(backend) + file_client = FileClient(backend=allowed_backends[0]) + + with io.BytesIO(file_client.get(filename)) as buffer: + checkpoint = torch.load(buffer, map_location=map_location) + return checkpoint + + +@CheckpointLoader.register_scheme(prefixes=('modelzoo://', 'torchvision://')) +def load_from_torchvision(filename, map_location=None): + """load checkpoint through the file path prefixed with modelzoo or + torchvision. + + Args: + filename (str): checkpoint file path with modelzoo or + torchvision prefix + map_location (str, optional): Same as :func:`torch.load`. + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + model_urls = get_torchvision_models() + if filename.startswith('modelzoo://'): + warnings.warn( + 'The URL scheme of "modelzoo://" is deprecated, please ' + 'use "torchvision://" instead', DeprecationWarning) + model_name = filename[11:] + else: + model_name = filename[14:] + + # Support getting model urls in the same way as torchvision + # `ResNet50_Weights.IMAGENET1K_V1` will be mapped to + # resnet50.imagenet1k_v1. + model_name = model_name.lower().replace('_weights', '') + return load_from_http(model_urls[model_name], map_location=map_location) + + +@CheckpointLoader.register_scheme(prefixes=('open-mmlab://', 'openmmlab://')) +def load_from_openmmlab(filename, map_location=None): + """load checkpoint through the file path prefixed with open-mmlab or + openmmlab. + + Args: + filename (str): checkpoint file path with open-mmlab or + openmmlab prefix + map_location (str, optional): Same as :func:`torch.load`. + Default: None + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + + model_urls = get_external_models() + prefix_str = 'open-mmlab://' + if filename.startswith(prefix_str): + model_name = filename[13:] + else: + model_name = filename[12:] + prefix_str = 'openmmlab://' + + deprecated_urls = get_deprecated_model_names() + if model_name in deprecated_urls: + warnings.warn( + f'{prefix_str}{model_name} is deprecated in favor ' + f'of {prefix_str}{deprecated_urls[model_name]}', + DeprecationWarning) + model_name = deprecated_urls[model_name] + model_url = model_urls[model_name] + # check if is url + if model_url.startswith(('http://', 'https://')): + checkpoint = load_from_http(model_url, map_location=map_location) + else: + filename = osp.join(_get_mmcv_home(), model_url) + if not osp.isfile(filename): + raise FileNotFoundError(f'{filename} can not be found.') + checkpoint = torch.load(filename, map_location=map_location) + return checkpoint + + +@CheckpointLoader.register_scheme(prefixes='mmcls://') +def load_from_mmcls(filename, map_location=None): + """load checkpoint through the file path prefixed with mmcls. + + Args: + filename (str): checkpoint file path with mmcls prefix + map_location (str, optional): Same as :func:`torch.load`. + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + + model_urls = get_mmcls_models() + model_name = filename[8:] + checkpoint = load_from_http( + model_urls[model_name], map_location=map_location) + checkpoint = _process_mmcls_checkpoint(checkpoint) + return checkpoint + + +def _load_checkpoint(filename, map_location=None, logger=None): + """Load checkpoint from somewhere (modelzoo, file, url). + + Args: + filename (str): Accept local filepath, URL, ``torchvision://xxx``, + ``open-mmlab://xxx``. Please refer to ``docs/model_zoo.md`` for + details. + map_location (str, optional): Same as :func:`torch.load`. + Default: None. + logger (:mod:`logging.Logger`, optional): The logger for error message. + Default: None + + Returns: + dict or OrderedDict: The loaded checkpoint. It can be either an + OrderedDict storing model weights or a dict containing other + information, which depends on the checkpoint. + """ + return CheckpointLoader.load_checkpoint(filename, map_location, logger) + + +def _load_checkpoint_with_prefix(prefix, filename, map_location=None): + """Load partial pretrained model with specific prefix. + + Args: + prefix (str): The prefix of sub-module. + filename (str): Accept local filepath, URL, ``torchvision://xxx``, + ``open-mmlab://xxx``. Please refer to ``docs/model_zoo.md`` for + details. + map_location (str | None): Same as :func:`torch.load`. Default: None. + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + + checkpoint = _load_checkpoint(filename, map_location=map_location) + + if 'state_dict' in checkpoint: + state_dict = checkpoint['state_dict'] + else: + state_dict = checkpoint + if not prefix.endswith('.'): + prefix += '.' + prefix_len = len(prefix) + + state_dict = { + k[prefix_len:]: v + for k, v in state_dict.items() if k.startswith(prefix) + } + + assert state_dict, f'{prefix} is not in the pretrained model' + return state_dict + + +def load_checkpoint(model, + filename, + map_location=None, + strict=False, + logger=None, + revise_keys=[(r'^module\.', '')]): + """Load checkpoint from a file or URI. + + Args: + model (Module): Module to load checkpoint. + filename (str): Accept local filepath, URL, ``torchvision://xxx``, + ``open-mmlab://xxx``. Please refer to ``docs/model_zoo.md`` for + details. + map_location (str): Same as :func:`torch.load`. + strict (bool): Whether to allow different params for the model and + checkpoint. + logger (:mod:`logging.Logger` or None): The logger for error message. + revise_keys (list): A list of customized keywords to modify the + state_dict in checkpoint. Each item is a (pattern, replacement) + pair of the regular expression operations. Default: strip + the prefix 'module.' by [(r'^module\\.', '')]. + + Returns: + dict or OrderedDict: The loaded checkpoint. + """ + checkpoint = _load_checkpoint(filename, map_location, logger) + # OrderedDict is a subclass of dict + if not isinstance(checkpoint, dict): + raise RuntimeError( + f'No state_dict found in checkpoint file {filename}') + # get state_dict from checkpoint + if 'state_dict' in checkpoint: + state_dict = checkpoint['state_dict'] + else: + state_dict = checkpoint + + # strip prefix of state_dict + metadata = getattr(state_dict, '_metadata', OrderedDict()) + for p, r in revise_keys: + state_dict = OrderedDict( + {re.sub(p, r, k): v + for k, v in state_dict.items()}) + # Keep metadata in state_dict + state_dict._metadata = metadata + + # load state_dict + load_state_dict(model, state_dict, strict, logger) + return checkpoint + + +def weights_to_cpu(state_dict): + """Copy a model state_dict to cpu. + + Args: + state_dict (OrderedDict): Model weights on GPU. + + Returns: + OrderedDict: Model weights on GPU. + """ + state_dict_cpu = OrderedDict() + for key, val in state_dict.items(): + state_dict_cpu[key] = val.cpu() + # Keep metadata in state_dict + state_dict_cpu._metadata = getattr(state_dict, '_metadata', OrderedDict()) + return state_dict_cpu + + +def _save_to_state_dict(module, destination, prefix, keep_vars): + """Saves module state to `destination` dictionary. + + This method is modified from :meth:`torch.nn.Module._save_to_state_dict`. + + Args: + module (nn.Module): The module to generate state_dict. + destination (dict): A dict where state will be stored. + prefix (str): The prefix for parameters and buffers used in this + module. + """ + for name, param in module._parameters.items(): + if param is not None: + destination[prefix + name] = param if keep_vars else param.detach() + for name, buf in module._buffers.items(): + # remove check of _non_persistent_buffers_set to allow nn.BatchNorm2d + if buf is not None: + destination[prefix + name] = buf if keep_vars else buf.detach() + + +def get_state_dict(module, destination=None, prefix='', keep_vars=False): + """Returns a dictionary containing a whole state of the module. + + Both parameters and persistent buffers (e.g. running averages) are + included. Keys are corresponding parameter and buffer names. + + This method is modified from :meth:`torch.nn.Module.state_dict` to + recursively check parallel module in case that the model has a complicated + structure, e.g., nn.Module(nn.Module(DDP)). + + Args: + module (nn.Module): The module to generate state_dict. + destination (OrderedDict): Returned dict for the state of the + module. + prefix (str): Prefix of the key. + keep_vars (bool): Whether to keep the variable property of the + parameters. Default: False. + + Returns: + dict: A dictionary containing a whole state of the module. + """ + # recursively check parallel module in case that the model has a + # complicated structure, e.g., nn.Module(nn.Module(DDP)) + if is_module_wrapper(module): + module = module.module + + # below is the same as torch.nn.Module.state_dict() + if destination is None: + destination = OrderedDict() + destination._metadata = OrderedDict() + destination._metadata[prefix[:-1]] = local_metadata = dict( + version=module._version) + _save_to_state_dict(module, destination, prefix, keep_vars) + for name, child in module._modules.items(): + if child is not None: + get_state_dict( + child, destination, prefix + name + '.', keep_vars=keep_vars) + for hook in module._state_dict_hooks.values(): + hook_result = hook(module, destination, prefix, local_metadata) + if hook_result is not None: + destination = hook_result + return destination + + +def save_checkpoint(model, + filename, + optimizer=None, + meta=None, + file_client_args=None): + """Save checkpoint to file. + + The checkpoint will have 3 fields: ``meta``, ``state_dict`` and + ``optimizer``. By default ``meta`` will contain version and time info. + + Args: + model (Module): Module whose params are to be saved. + filename (str): Checkpoint filename. + optimizer (:obj:`Optimizer`, optional): Optimizer to be saved. + meta (dict, optional): Metadata to be saved in checkpoint. + file_client_args (dict, optional): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + `New in version 1.3.16.` + """ + if meta is None: + meta = {} + elif not isinstance(meta, dict): + raise TypeError(f'meta must be a dict or None, but got {type(meta)}') + meta.update(mmcv_version=mmcv.__version__, time=time.asctime()) + + if is_module_wrapper(model): + model = model.module + + if hasattr(model, 'CLASSES') and model.CLASSES is not None: + # save class name to the meta + meta.update(CLASSES=model.CLASSES) + + checkpoint = { + 'meta': meta, + 'state_dict': weights_to_cpu(get_state_dict(model)) + } + # save optimizer state dict in the checkpoint + if isinstance(optimizer, Optimizer): + checkpoint['optimizer'] = optimizer.state_dict() + elif isinstance(optimizer, dict): + checkpoint['optimizer'] = {} + for name, optim in optimizer.items(): + checkpoint['optimizer'][name] = optim.state_dict() + + if filename.startswith('pavi://'): + if file_client_args is not None: + raise ValueError( + 'file_client_args should be "None" if filename starts with' + f'"pavi://", but got {file_client_args}') + try: + from pavi import exception, modelcloud + except ImportError: + raise ImportError( + 'Please install pavi to load checkpoint from modelcloud.') + model_path = filename[7:] + root = modelcloud.Folder() + model_dir, model_name = osp.split(model_path) + try: + model = modelcloud.get(model_dir) + except exception.NodeNotFoundError: + model = root.create_training_model(model_dir) + with TemporaryDirectory() as tmp_dir: + checkpoint_file = osp.join(tmp_dir, model_name) + with open(checkpoint_file, 'wb') as f: + torch.save(checkpoint, f) + f.flush() + model.create_file(checkpoint_file, name=model_name) + else: + file_client = FileClient.infer_client(file_client_args, filename) + with io.BytesIO() as f: + torch.save(checkpoint, f) + file_client.put(f.getvalue(), filename) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/default_constructor.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/default_constructor.py new file mode 100644 index 0000000000000000000000000000000000000000..4a4f2cc646146ac7153e0fca22f623a8e5300f9a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/default_constructor.py @@ -0,0 +1,45 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .builder import RUNNER_BUILDERS, RUNNERS + + +@RUNNER_BUILDERS.register_module() +class DefaultRunnerConstructor: + """Default constructor for runners. + + Custom existing `Runner` like `EpocBasedRunner` though `RunnerConstructor`. + For example, We can inject some new properties and functions for `Runner`. + + Example: + >>> from mmcv.runner import RUNNER_BUILDERS, build_runner + >>> # Define a new RunnerReconstructor + >>> @RUNNER_BUILDERS.register_module() + >>> class MyRunnerConstructor: + ... def __init__(self, runner_cfg, default_args=None): + ... if not isinstance(runner_cfg, dict): + ... raise TypeError('runner_cfg should be a dict', + ... f'but got {type(runner_cfg)}') + ... self.runner_cfg = runner_cfg + ... self.default_args = default_args + ... + ... def __call__(self): + ... runner = RUNNERS.build(self.runner_cfg, + ... default_args=self.default_args) + ... # Add new properties for existing runner + ... runner.my_name = 'my_runner' + ... runner.my_function = lambda self: print(self.my_name) + ... ... + >>> # build your runner + >>> runner_cfg = dict(type='EpochBasedRunner', max_epochs=40, + ... constructor='MyRunnerConstructor') + >>> runner = build_runner(runner_cfg) + """ + + def __init__(self, runner_cfg, default_args=None): + if not isinstance(runner_cfg, dict): + raise TypeError('runner_cfg should be a dict', + f'but got {type(runner_cfg)}') + self.runner_cfg = runner_cfg + self.default_args = default_args + + def __call__(self): + return RUNNERS.build(self.runner_cfg, default_args=self.default_args) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/dist_utils.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/dist_utils.py new file mode 100644 index 0000000000000000000000000000000000000000..26d77a8f9572548d9a5f0984cfd150e3f0ed6932 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/dist_utils.py @@ -0,0 +1,204 @@ +# Copyright (c) OpenMMLab. All rights reserved. +# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved +import functools +import os +import socket +import subprocess +from collections import OrderedDict + +import torch +import torch.multiprocessing as mp +from torch import distributed as dist +from torch._utils import (_flatten_dense_tensors, _take_tensors, + _unflatten_dense_tensors) + +from mmcv.utils import IS_MLU_AVAILABLE + + +def _find_free_port(): + # Copied from https://github.com/facebookresearch/detectron2/blob/main/detectron2/engine/launch.py # noqa: E501 + sock = socket.socket(socket.AF_INET, socket.SOCK_STREAM) + # Binding to port 0 will cause the OS to find an available port for us + sock.bind(('', 0)) + port = sock.getsockname()[1] + sock.close() + # NOTE: there is still a chance the port could be taken by other processes. + return port + + +def _is_free_port(port): + ips = socket.gethostbyname_ex(socket.gethostname())[-1] + ips.append('localhost') + with socket.socket(socket.AF_INET, socket.SOCK_STREAM) as s: + return all(s.connect_ex((ip, port)) != 0 for ip in ips) + + +def init_dist(launcher, backend='nccl', **kwargs): + if mp.get_start_method(allow_none=True) is None: + mp.set_start_method('spawn') + if launcher == 'pytorch': + _init_dist_pytorch(backend, **kwargs) + elif launcher == 'mpi': + _init_dist_mpi(backend, **kwargs) + elif launcher == 'slurm': + _init_dist_slurm(backend, **kwargs) + else: + raise ValueError(f'Invalid launcher type: {launcher}') + + +def _init_dist_pytorch(backend, **kwargs): + # TODO: use local_rank instead of rank % num_gpus + rank = int(os.environ['RANK']) + if IS_MLU_AVAILABLE: + import torch_mlu # noqa: F401 + torch.mlu.set_device(rank) + dist.init_process_group( + backend='cncl', + rank=rank, + world_size=int(os.environ['WORLD_SIZE']), + **kwargs) + else: + num_gpus = torch.cuda.device_count() + torch.cuda.set_device(rank % num_gpus) + dist.init_process_group(backend=backend, **kwargs) + + +def _init_dist_mpi(backend, **kwargs): + local_rank = int(os.environ['OMPI_COMM_WORLD_LOCAL_RANK']) + torch.cuda.set_device(local_rank) + if 'MASTER_PORT' not in os.environ: + # 29500 is torch.distributed default port + os.environ['MASTER_PORT'] = '29500' + if 'MASTER_ADDR' not in os.environ: + raise KeyError('The environment variable MASTER_ADDR is not set') + os.environ['WORLD_SIZE'] = os.environ['OMPI_COMM_WORLD_SIZE'] + os.environ['RANK'] = os.environ['OMPI_COMM_WORLD_RANK'] + dist.init_process_group(backend=backend, **kwargs) + + +def _init_dist_slurm(backend, port=None): + """Initialize slurm distributed training environment. + + If argument ``port`` is not specified, then the master port will be system + environment variable ``MASTER_PORT``. If ``MASTER_PORT`` is not in system + environment variable, then a default port ``29500`` will be used. + + Args: + backend (str): Backend of torch.distributed. + port (int, optional): Master port. Defaults to None. + """ + proc_id = int(os.environ['SLURM_PROCID']) + ntasks = int(os.environ['SLURM_NTASKS']) + node_list = os.environ['SLURM_NODELIST'] + num_gpus = torch.cuda.device_count() + torch.cuda.set_device(proc_id % num_gpus) + addr = subprocess.getoutput( + f'scontrol show hostname {node_list} | head -n1') + # specify master port + if port is not None: + os.environ['MASTER_PORT'] = str(port) + elif 'MASTER_PORT' in os.environ: + pass # use MASTER_PORT in the environment variable + else: + # if torch.distributed default port(29500) is available + # then use it, else find a free port + if _is_free_port(29500): + os.environ['MASTER_PORT'] = '29500' + else: + os.environ['MASTER_PORT'] = str(_find_free_port()) + # use MASTER_ADDR in the environment variable if it already exists + if 'MASTER_ADDR' not in os.environ: + os.environ['MASTER_ADDR'] = addr + os.environ['WORLD_SIZE'] = str(ntasks) + os.environ['LOCAL_RANK'] = str(proc_id % num_gpus) + os.environ['RANK'] = str(proc_id) + dist.init_process_group(backend=backend) + + +def get_dist_info(): + if dist.is_available() and dist.is_initialized(): + rank = dist.get_rank() + world_size = dist.get_world_size() + else: + rank = 0 + world_size = 1 + return rank, world_size + + +def master_only(func): + + @functools.wraps(func) + def wrapper(*args, **kwargs): + rank, _ = get_dist_info() + if rank == 0: + return func(*args, **kwargs) + + return wrapper + + +def allreduce_params(params, coalesce=True, bucket_size_mb=-1): + """Allreduce parameters. + + Args: + params (list[torch.Parameters]): List of parameters or buffers of a + model. + coalesce (bool, optional): Whether allreduce parameters as a whole. + Defaults to True. + bucket_size_mb (int, optional): Size of bucket, the unit is MB. + Defaults to -1. + """ + _, world_size = get_dist_info() + if world_size == 1: + return + params = [param.data for param in params] + if coalesce: + _allreduce_coalesced(params, world_size, bucket_size_mb) + else: + for tensor in params: + dist.all_reduce(tensor.div_(world_size)) + + +def allreduce_grads(params, coalesce=True, bucket_size_mb=-1): + """Allreduce gradients. + + Args: + params (list[torch.Parameters]): List of parameters of a model + coalesce (bool, optional): Whether allreduce parameters as a whole. + Defaults to True. + bucket_size_mb (int, optional): Size of bucket, the unit is MB. + Defaults to -1. + """ + grads = [ + param.grad.data for param in params + if param.requires_grad and param.grad is not None + ] + _, world_size = get_dist_info() + if world_size == 1: + return + if coalesce: + _allreduce_coalesced(grads, world_size, bucket_size_mb) + else: + for tensor in grads: + dist.all_reduce(tensor.div_(world_size)) + + +def _allreduce_coalesced(tensors, world_size, bucket_size_mb=-1): + if bucket_size_mb > 0: + bucket_size_bytes = bucket_size_mb * 1024 * 1024 + buckets = _take_tensors(tensors, bucket_size_bytes) + else: + buckets = OrderedDict() + for tensor in tensors: + tp = tensor.type() + if tp not in buckets: + buckets[tp] = [] + buckets[tp].append(tensor) + buckets = buckets.values() + + for bucket in buckets: + flat_tensors = _flatten_dense_tensors(bucket) + dist.all_reduce(flat_tensors) + flat_tensors.div_(world_size) + for tensor, synced in zip( + bucket, _unflatten_dense_tensors(flat_tensors, bucket)): + tensor.copy_(synced) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/epoch_based_runner.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/epoch_based_runner.py new file mode 100644 index 0000000000000000000000000000000000000000..078e91df3f38b24cbb325deae0e13316b0655209 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/epoch_based_runner.py @@ -0,0 +1,188 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import platform +import shutil +import time +import warnings + +import torch + +import mmcv +from .base_runner import BaseRunner +from .builder import RUNNERS +from .checkpoint import save_checkpoint +from .utils import get_host_info + + +@RUNNERS.register_module() +class EpochBasedRunner(BaseRunner): + """Epoch-based Runner. + + This runner train models epoch by epoch. + """ + + def run_iter(self, data_batch, train_mode, **kwargs): + if self.batch_processor is not None: + outputs = self.batch_processor( + self.model, data_batch, train_mode=train_mode, **kwargs) + elif train_mode: + outputs = self.model.train_step(data_batch, self.optimizer, + **kwargs) + else: + outputs = self.model.val_step(data_batch, self.optimizer, **kwargs) + if not isinstance(outputs, dict): + raise TypeError('"batch_processor()" or "model.train_step()"' + 'and "model.val_step()" must return a dict') + if 'log_vars' in outputs: + self.log_buffer.update(outputs['log_vars'], outputs['num_samples']) + self.outputs = outputs + + def train(self, data_loader, **kwargs): + self.model.train() + self.mode = 'train' + self.data_loader = data_loader + self._max_iters = self._max_epochs * len(self.data_loader) + self.call_hook('before_train_epoch') + time.sleep(2) # Prevent possible deadlock during epoch transition + for i, data_batch in enumerate(self.data_loader): + self._inner_iter = i + self.call_hook('before_train_iter') + self.run_iter(data_batch, train_mode=True, **kwargs) + self.call_hook('after_train_iter') + self._iter += 1 + + self.call_hook('after_train_epoch') + self._epoch += 1 + + @torch.no_grad() + def val(self, data_loader, **kwargs): + self.model.eval() + self.mode = 'val' + self.data_loader = data_loader + self.call_hook('before_val_epoch') + time.sleep(2) # Prevent possible deadlock during epoch transition + for i, data_batch in enumerate(self.data_loader): + self._inner_iter = i + self.call_hook('before_val_iter') + self.run_iter(data_batch, train_mode=False) + self.call_hook('after_val_iter') + + self.call_hook('after_val_epoch') + + def run(self, data_loaders, workflow, max_epochs=None, **kwargs): + """Start running. + + Args: + data_loaders (list[:obj:`DataLoader`]): Dataloaders for training + and validation. + workflow (list[tuple]): A list of (phase, epochs) to specify the + running order and epochs. E.g, [('train', 2), ('val', 1)] means + running 2 epochs for training and 1 epoch for validation, + iteratively. + """ + assert isinstance(data_loaders, list) + assert mmcv.is_list_of(workflow, tuple) + assert len(data_loaders) == len(workflow) + if max_epochs is not None: + warnings.warn( + 'setting max_epochs in run is deprecated, ' + 'please set max_epochs in runner_config', DeprecationWarning) + self._max_epochs = max_epochs + + assert self._max_epochs is not None, ( + 'max_epochs must be specified during instantiation') + + for i, flow in enumerate(workflow): + mode, epochs = flow + if mode == 'train': + self._max_iters = self._max_epochs * len(data_loaders[i]) + break + + work_dir = self.work_dir if self.work_dir is not None else 'NONE' + self.logger.info('Start running, host: %s, work_dir: %s', + get_host_info(), work_dir) + self.logger.info('Hooks will be executed in the following order:\n%s', + self.get_hook_info()) + self.logger.info('workflow: %s, max: %d epochs', workflow, + self._max_epochs) + self.call_hook('before_run') + + while self.epoch < self._max_epochs: + for i, flow in enumerate(workflow): + mode, epochs = flow + if isinstance(mode, str): # self.train() + if not hasattr(self, mode): + raise ValueError( + f'runner has no method named "{mode}" to run an ' + 'epoch') + epoch_runner = getattr(self, mode) + else: + raise TypeError( + 'mode in workflow must be a str, but got {}'.format( + type(mode))) + + for _ in range(epochs): + if mode == 'train' and self.epoch >= self._max_epochs: + break + epoch_runner(data_loaders[i], **kwargs) + + time.sleep(1) # wait for some hooks like loggers to finish + self.call_hook('after_run') + + def save_checkpoint(self, + out_dir, + filename_tmpl='epoch_{}.pth', + save_optimizer=True, + meta=None, + create_symlink=True): + """Save the checkpoint. + + Args: + out_dir (str): The directory that checkpoints are saved. + filename_tmpl (str, optional): The checkpoint filename template, + which contains a placeholder for the epoch number. + Defaults to 'epoch_{}.pth'. + save_optimizer (bool, optional): Whether to save the optimizer to + the checkpoint. Defaults to True. + meta (dict, optional): The meta information to be saved in the + checkpoint. Defaults to None. + create_symlink (bool, optional): Whether to create a symlink + "latest.pth" to point to the latest checkpoint. + Defaults to True. + """ + if meta is None: + meta = {} + elif not isinstance(meta, dict): + raise TypeError( + f'meta should be a dict or None, but got {type(meta)}') + if self.meta is not None: + meta.update(self.meta) + # Note: meta.update(self.meta) should be done before + # meta.update(epoch=self.epoch + 1, iter=self.iter) otherwise + # there will be problems with resumed checkpoints. + # More details in https://github.com/open-mmlab/mmcv/pull/1108 + meta.update(epoch=self.epoch + 1, iter=self.iter) + + filename = filename_tmpl.format(self.epoch + 1) + filepath = osp.join(out_dir, filename) + optimizer = self.optimizer if save_optimizer else None + save_checkpoint(self.model, filepath, optimizer=optimizer, meta=meta) + # in some environments, `os.symlink` is not supported, you may need to + # set `create_symlink` to False + if create_symlink: + dst_file = osp.join(out_dir, 'latest.pth') + if platform.system() != 'Windows': + mmcv.symlink(filename, dst_file) + else: + shutil.copy(filepath, dst_file) + + +@RUNNERS.register_module() +class Runner(EpochBasedRunner): + """Deprecated name of EpochBasedRunner.""" + + def __init__(self, *args, **kwargs): + warnings.warn( + 'Runner was deprecated, please use EpochBasedRunner instead', + DeprecationWarning) + super().__init__(*args, **kwargs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/fp16_utils.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/fp16_utils.py new file mode 100644 index 0000000000000000000000000000000000000000..be3ac3a3149d215d22b8d3168cf7ac0048cc4af0 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/fp16_utils.py @@ -0,0 +1,423 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import functools +import warnings +from collections import abc +from inspect import getfullargspec + +import numpy as np +import torch +import torch.nn as nn + +from mmcv.utils import TORCH_VERSION, digit_version +from .dist_utils import allreduce_grads as _allreduce_grads + +try: + # If PyTorch version >= 1.6.0, torch.cuda.amp.autocast would be imported + # and used; otherwise, auto fp16 will adopt mmcv's implementation. + # Note that when PyTorch >= 1.6.0, we still cast tensor types to fp16 + # manually, so the behavior may not be consistent with real amp. + from torch.cuda.amp import autocast +except ImportError: + pass + + +def cast_tensor_type(inputs, src_type, dst_type): + """Recursively convert Tensor in inputs from src_type to dst_type. + + Note: + In v1.4.4 and later, ``cast_tersor_type`` will only convert the + torch.Tensor which is consistent with ``src_type`` to the ``dst_type``. + Before v1.4.4, it ignores the ``src_type`` argument, leading to some + potential problems. For example, + ``cast_tensor_type(inputs, torch.float, torch.half)`` will convert all + tensors in inputs to ``torch.half`` including those originally in + ``torch.Int`` or other types, which is not expected. + + Args: + inputs: Inputs that to be casted. + src_type (torch.dtype): Source type.. + dst_type (torch.dtype): Destination type. + + Returns: + The same type with inputs, but all contained Tensors have been cast. + """ + if isinstance(inputs, nn.Module): + return inputs + elif isinstance(inputs, torch.Tensor): + # we need to ensure that the type of inputs to be casted are the same + # as the argument `src_type`. + return inputs.to(dst_type) if inputs.dtype == src_type else inputs + elif isinstance(inputs, str): + return inputs + elif isinstance(inputs, np.ndarray): + return inputs + elif isinstance(inputs, abc.Mapping): + return type(inputs)({ + k: cast_tensor_type(v, src_type, dst_type) + for k, v in inputs.items() + }) + elif isinstance(inputs, abc.Iterable): + return type(inputs)( + cast_tensor_type(item, src_type, dst_type) for item in inputs) + else: + return inputs + + +def auto_fp16(apply_to=None, out_fp32=False, supported_types=(nn.Module, )): + """Decorator to enable fp16 training automatically. + + This decorator is useful when you write custom modules and want to support + mixed precision training. If inputs arguments are fp32 tensors, they will + be converted to fp16 automatically. Arguments other than fp32 tensors are + ignored. If you are using PyTorch >= 1.6, torch.cuda.amp is used as the + backend, otherwise, original mmcv implementation will be adopted. + + Args: + apply_to (Iterable, optional): The argument names to be converted. + `None` indicates all arguments. + out_fp32 (bool): Whether to convert the output back to fp32. + supported_types (tuple): Classes can be decorated by ``auto_fp16``. + `New in version 1.5.0.` + Example: + + >>> import torch.nn as nn + >>> class MyModule1(nn.Module): + >>> + >>> # Convert x and y to fp16 + >>> @auto_fp16() + >>> def forward(self, x, y): + >>> pass + + >>> import torch.nn as nn + >>> class MyModule2(nn.Module): + >>> + >>> # convert pred to fp16 + >>> @auto_fp16(apply_to=('pred', )) + >>> def do_something(self, pred, others): + >>> pass + """ + + def auto_fp16_wrapper(old_func): + + @functools.wraps(old_func) + def new_func(*args, **kwargs): + # check if the module has set the attribute `fp16_enabled`, if not, + # just fallback to the original method. + if not isinstance(args[0], supported_types): + raise TypeError('@auto_fp16 can only be used to decorate the ' + f'method of those classes {supported_types}') + if not (hasattr(args[0], 'fp16_enabled') and args[0].fp16_enabled): + return old_func(*args, **kwargs) + + # get the arg spec of the decorated method + args_info = getfullargspec(old_func) + # get the argument names to be casted + args_to_cast = args_info.args if apply_to is None else apply_to + # convert the args that need to be processed + new_args = [] + # NOTE: default args are not taken into consideration + if args: + arg_names = args_info.args[:len(args)] + for i, arg_name in enumerate(arg_names): + if arg_name in args_to_cast: + new_args.append( + cast_tensor_type(args[i], torch.float, torch.half)) + else: + new_args.append(args[i]) + # convert the kwargs that need to be processed + new_kwargs = {} + if kwargs: + for arg_name, arg_value in kwargs.items(): + if arg_name in args_to_cast: + new_kwargs[arg_name] = cast_tensor_type( + arg_value, torch.float, torch.half) + else: + new_kwargs[arg_name] = arg_value + # apply converted arguments to the decorated method + if (TORCH_VERSION != 'parrots' and + digit_version(TORCH_VERSION) >= digit_version('1.6.0')): + with autocast(enabled=True): + output = old_func(*new_args, **new_kwargs) + else: + output = old_func(*new_args, **new_kwargs) + # cast the results back to fp32 if necessary + if out_fp32: + output = cast_tensor_type(output, torch.half, torch.float) + return output + + return new_func + + return auto_fp16_wrapper + + +def force_fp32(apply_to=None, out_fp16=False): + """Decorator to convert input arguments to fp32 in force. + + This decorator is useful when you write custom modules and want to support + mixed precision training. If there are some inputs that must be processed + in fp32 mode, then this decorator can handle it. If inputs arguments are + fp16 tensors, they will be converted to fp32 automatically. Arguments other + than fp16 tensors are ignored. If you are using PyTorch >= 1.6, + torch.cuda.amp is used as the backend, otherwise, original mmcv + implementation will be adopted. + + Args: + apply_to (Iterable, optional): The argument names to be converted. + `None` indicates all arguments. + out_fp16 (bool): Whether to convert the output back to fp16. + + Example: + + >>> import torch.nn as nn + >>> class MyModule1(nn.Module): + >>> + >>> # Convert x and y to fp32 + >>> @force_fp32() + >>> def loss(self, x, y): + >>> pass + + >>> import torch.nn as nn + >>> class MyModule2(nn.Module): + >>> + >>> # convert pred to fp32 + >>> @force_fp32(apply_to=('pred', )) + >>> def post_process(self, pred, others): + >>> pass + """ + + def force_fp32_wrapper(old_func): + + @functools.wraps(old_func) + def new_func(*args, **kwargs): + # check if the module has set the attribute `fp16_enabled`, if not, + # just fallback to the original method. + if not isinstance(args[0], torch.nn.Module): + raise TypeError('@force_fp32 can only be used to decorate the ' + 'method of nn.Module') + if not (hasattr(args[0], 'fp16_enabled') and args[0].fp16_enabled): + return old_func(*args, **kwargs) + # get the arg spec of the decorated method + args_info = getfullargspec(old_func) + # get the argument names to be casted + args_to_cast = args_info.args if apply_to is None else apply_to + # convert the args that need to be processed + new_args = [] + if args: + arg_names = args_info.args[:len(args)] + for i, arg_name in enumerate(arg_names): + if arg_name in args_to_cast: + new_args.append( + cast_tensor_type(args[i], torch.half, torch.float)) + else: + new_args.append(args[i]) + # convert the kwargs that need to be processed + new_kwargs = dict() + if kwargs: + for arg_name, arg_value in kwargs.items(): + if arg_name in args_to_cast: + new_kwargs[arg_name] = cast_tensor_type( + arg_value, torch.half, torch.float) + else: + new_kwargs[arg_name] = arg_value + # apply converted arguments to the decorated method + if (TORCH_VERSION != 'parrots' and + digit_version(TORCH_VERSION) >= digit_version('1.6.0')): + with autocast(enabled=False): + output = old_func(*new_args, **new_kwargs) + else: + output = old_func(*new_args, **new_kwargs) + # cast the results back to fp32 if necessary + if out_fp16: + output = cast_tensor_type(output, torch.float, torch.half) + return output + + return new_func + + return force_fp32_wrapper + + +def allreduce_grads(params, coalesce=True, bucket_size_mb=-1): + warnings.warning( + '"mmcv.runner.fp16_utils.allreduce_grads" is deprecated, and will be ' + 'removed in v2.8. Please switch to "mmcv.runner.allreduce_grads', + DeprecationWarning) + _allreduce_grads(params, coalesce=coalesce, bucket_size_mb=bucket_size_mb) + + +def wrap_fp16_model(model): + """Wrap the FP32 model to FP16. + + If you are using PyTorch >= 1.6, torch.cuda.amp is used as the + backend, otherwise, original mmcv implementation will be adopted. + + For PyTorch >= 1.6, this function will + 1. Set fp16 flag inside the model to True. + + Otherwise: + 1. Convert FP32 model to FP16. + 2. Remain some necessary layers to be FP32, e.g., normalization layers. + 3. Set `fp16_enabled` flag inside the model to True. + + Args: + model (nn.Module): Model in FP32. + """ + if (TORCH_VERSION == 'parrots' + or digit_version(TORCH_VERSION) < digit_version('1.6.0')): + # convert model to fp16 + model.half() + # patch the normalization layers to make it work in fp32 mode + patch_norm_fp32(model) + # set `fp16_enabled` flag + for m in model.modules(): + if hasattr(m, 'fp16_enabled'): + m.fp16_enabled = True + + +def patch_norm_fp32(module): + """Recursively convert normalization layers from FP16 to FP32. + + Args: + module (nn.Module): The modules to be converted in FP16. + + Returns: + nn.Module: The converted module, the normalization layers have been + converted to FP32. + """ + if isinstance(module, (nn.modules.batchnorm._BatchNorm, nn.GroupNorm)): + module.float() + if isinstance(module, nn.GroupNorm) or torch.__version__ < '1.3': + module.forward = patch_forward_method(module.forward, torch.half, + torch.float) + for child in module.children(): + patch_norm_fp32(child) + return module + + +def patch_forward_method(func, src_type, dst_type, convert_output=True): + """Patch the forward method of a module. + + Args: + func (callable): The original forward method. + src_type (torch.dtype): Type of input arguments to be converted from. + dst_type (torch.dtype): Type of input arguments to be converted to. + convert_output (bool): Whether to convert the output back to src_type. + + Returns: + callable: The patched forward method. + """ + + def new_forward(*args, **kwargs): + output = func(*cast_tensor_type(args, src_type, dst_type), + **cast_tensor_type(kwargs, src_type, dst_type)) + if convert_output: + output = cast_tensor_type(output, dst_type, src_type) + return output + + return new_forward + + +class LossScaler: + """Class that manages loss scaling in mixed precision training which + supports both dynamic or static mode. + + The implementation refers to + https://github.com/NVIDIA/apex/blob/master/apex/fp16_utils/loss_scaler.py. + Indirectly, by supplying ``mode='dynamic'`` for dynamic loss scaling. + It's important to understand how :class:`LossScaler` operates. + Loss scaling is designed to combat the problem of underflowing + gradients encountered at long times when training fp16 networks. + Dynamic loss scaling begins by attempting a very high loss + scale. Ironically, this may result in OVERflowing gradients. + If overflowing gradients are encountered, :class:`FP16_Optimizer` then + skips the update step for this particular iteration/minibatch, + and :class:`LossScaler` adjusts the loss scale to a lower value. + If a certain number of iterations occur without overflowing gradients + detected,:class:`LossScaler` increases the loss scale once more. + In this way :class:`LossScaler` attempts to "ride the edge" of always + using the highest loss scale possible without incurring overflow. + + Args: + init_scale (float): Initial loss scale value, default: 2**32. + scale_factor (float): Factor used when adjusting the loss scale. + Default: 2. + mode (str): Loss scaling mode. 'dynamic' or 'static' + scale_window (int): Number of consecutive iterations without an + overflow to wait before increasing the loss scale. Default: 1000. + """ + + def __init__(self, + init_scale=2**32, + mode='dynamic', + scale_factor=2., + scale_window=1000): + self.cur_scale = init_scale + self.cur_iter = 0 + assert mode in ('dynamic', + 'static'), 'mode can only be dynamic or static' + self.mode = mode + self.last_overflow_iter = -1 + self.scale_factor = scale_factor + self.scale_window = scale_window + + def has_overflow(self, params): + """Check if params contain overflow.""" + if self.mode != 'dynamic': + return False + for p in params: + if p.grad is not None and LossScaler._has_inf_or_nan(p.grad.data): + return True + return False + + def _has_inf_or_nan(x): + """Check if params contain NaN.""" + try: + cpu_sum = float(x.float().sum()) + except RuntimeError as instance: + if 'value cannot be converted' not in instance.args[0]: + raise + return True + else: + if cpu_sum == float('inf') or cpu_sum == -float('inf') \ + or cpu_sum != cpu_sum: + return True + return False + + def update_scale(self, overflow): + """update the current loss scale value when overflow happens.""" + if self.mode != 'dynamic': + return + if overflow: + self.cur_scale = max(self.cur_scale / self.scale_factor, 1) + self.last_overflow_iter = self.cur_iter + else: + if (self.cur_iter - self.last_overflow_iter) % \ + self.scale_window == 0: + self.cur_scale *= self.scale_factor + self.cur_iter += 1 + + def state_dict(self): + """Returns the state of the scaler as a :class:`dict`.""" + return dict( + cur_scale=self.cur_scale, + cur_iter=self.cur_iter, + mode=self.mode, + last_overflow_iter=self.last_overflow_iter, + scale_factor=self.scale_factor, + scale_window=self.scale_window) + + def load_state_dict(self, state_dict): + """Loads the loss_scaler state dict. + + Args: + state_dict (dict): scaler state. + """ + self.cur_scale = state_dict['cur_scale'] + self.cur_iter = state_dict['cur_iter'] + self.mode = state_dict['mode'] + self.last_overflow_iter = state_dict['last_overflow_iter'] + self.scale_factor = state_dict['scale_factor'] + self.scale_window = state_dict['scale_window'] + + @property + def loss_scale(self): + return self.cur_scale diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..03e2a619e8dd6c516add4a3b23c3c790430255ba --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/__init__.py @@ -0,0 +1,48 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .checkpoint import CheckpointHook +from .closure import ClosureHook +from .ema import EMAHook +from .evaluation import DistEvalHook, EvalHook +from .hook import HOOKS, Hook +from .iter_timer import IterTimerHook +from .logger import (ClearMLLoggerHook, DvcliveLoggerHook, LoggerHook, + MlflowLoggerHook, NeptuneLoggerHook, PaviLoggerHook, + SegmindLoggerHook, TensorboardLoggerHook, TextLoggerHook, + WandbLoggerHook) +from .lr_updater import (CosineAnnealingLrUpdaterHook, + CosineRestartLrUpdaterHook, CyclicLrUpdaterHook, + ExpLrUpdaterHook, FixedLrUpdaterHook, + FlatCosineAnnealingLrUpdaterHook, InvLrUpdaterHook, + LinearAnnealingLrUpdaterHook, LrUpdaterHook, + OneCycleLrUpdaterHook, PolyLrUpdaterHook, + StepLrUpdaterHook) +from .memory import EmptyCacheHook +from .momentum_updater import (CosineAnnealingMomentumUpdaterHook, + CyclicMomentumUpdaterHook, + LinearAnnealingMomentumUpdaterHook, + MomentumUpdaterHook, + OneCycleMomentumUpdaterHook, + StepMomentumUpdaterHook) +from .optimizer import (Fp16OptimizerHook, GradientCumulativeFp16OptimizerHook, + GradientCumulativeOptimizerHook, OptimizerHook) +from .profiler import ProfilerHook +from .sampler_seed import DistSamplerSeedHook +from .sync_buffer import SyncBuffersHook + +__all__ = [ + 'HOOKS', 'Hook', 'CheckpointHook', 'ClosureHook', 'LrUpdaterHook', + 'FixedLrUpdaterHook', 'StepLrUpdaterHook', 'ExpLrUpdaterHook', + 'PolyLrUpdaterHook', 'InvLrUpdaterHook', 'CosineAnnealingLrUpdaterHook', + 'FlatCosineAnnealingLrUpdaterHook', 'CosineRestartLrUpdaterHook', + 'CyclicLrUpdaterHook', 'OneCycleLrUpdaterHook', 'OptimizerHook', + 'Fp16OptimizerHook', 'IterTimerHook', 'DistSamplerSeedHook', + 'EmptyCacheHook', 'LoggerHook', 'MlflowLoggerHook', 'PaviLoggerHook', + 'TextLoggerHook', 'TensorboardLoggerHook', 'NeptuneLoggerHook', + 'WandbLoggerHook', 'DvcliveLoggerHook', 'MomentumUpdaterHook', + 'StepMomentumUpdaterHook', 'CosineAnnealingMomentumUpdaterHook', + 'CyclicMomentumUpdaterHook', 'OneCycleMomentumUpdaterHook', + 'SyncBuffersHook', 'EMAHook', 'EvalHook', 'DistEvalHook', 'ProfilerHook', + 'GradientCumulativeOptimizerHook', 'GradientCumulativeFp16OptimizerHook', + 'SegmindLoggerHook', 'LinearAnnealingLrUpdaterHook', + 'LinearAnnealingMomentumUpdaterHook', 'ClearMLLoggerHook' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/checkpoint.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/checkpoint.py new file mode 100644 index 0000000000000000000000000000000000000000..7bb75f402a110daf18d6f1966ce5ee1483f16e45 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/checkpoint.py @@ -0,0 +1,167 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import warnings + +from mmcv.fileio import FileClient +from ..dist_utils import allreduce_params, master_only +from .hook import HOOKS, Hook + + +@HOOKS.register_module() +class CheckpointHook(Hook): + """Save checkpoints periodically. + + Args: + interval (int): The saving period. If ``by_epoch=True``, interval + indicates epochs, otherwise it indicates iterations. + Default: -1, which means "never". + by_epoch (bool): Saving checkpoints by epoch or by iteration. + Default: True. + save_optimizer (bool): Whether to save optimizer state_dict in the + checkpoint. It is usually used for resuming experiments. + Default: True. + out_dir (str, optional): The root directory to save checkpoints. If not + specified, ``runner.work_dir`` will be used by default. If + specified, the ``out_dir`` will be the concatenation of ``out_dir`` + and the last level directory of ``runner.work_dir``. + `Changed in version 1.3.16.` + max_keep_ckpts (int, optional): The maximum checkpoints to keep. + In some cases we want only the latest few checkpoints and would + like to delete old ones to save the disk space. + Default: -1, which means unlimited. + save_last (bool, optional): Whether to force the last checkpoint to be + saved regardless of interval. Default: True. + sync_buffer (bool, optional): Whether to synchronize buffers in + different gpus. Default: False. + file_client_args (dict, optional): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + `New in version 1.3.16.` + + .. warning:: + Before v1.3.16, the ``out_dir`` argument indicates the path where the + checkpoint is stored. However, since v1.3.16, ``out_dir`` indicates the + root directory and the final path to save checkpoint is the + concatenation of ``out_dir`` and the last level directory of + ``runner.work_dir``. Suppose the value of ``out_dir`` is "/path/of/A" + and the value of ``runner.work_dir`` is "/path/of/B", then the final + path will be "/path/of/A/B". + """ + + def __init__(self, + interval=-1, + by_epoch=True, + save_optimizer=True, + out_dir=None, + max_keep_ckpts=-1, + save_last=True, + sync_buffer=False, + file_client_args=None, + **kwargs): + self.interval = interval + self.by_epoch = by_epoch + self.save_optimizer = save_optimizer + self.out_dir = out_dir + self.max_keep_ckpts = max_keep_ckpts + self.save_last = save_last + self.args = kwargs + self.sync_buffer = sync_buffer + self.file_client_args = file_client_args + + def before_run(self, runner): + if not self.out_dir: + self.out_dir = runner.work_dir + + self.file_client = FileClient.infer_client(self.file_client_args, + self.out_dir) + + # if `self.out_dir` is not equal to `runner.work_dir`, it means that + # `self.out_dir` is set so the final `self.out_dir` is the + # concatenation of `self.out_dir` and the last level directory of + # `runner.work_dir` + if self.out_dir != runner.work_dir: + basename = osp.basename(runner.work_dir.rstrip(osp.sep)) + self.out_dir = self.file_client.join_path(self.out_dir, basename) + + runner.logger.info((f'Checkpoints will be saved to {self.out_dir} by ' + f'{self.file_client.name}.')) + + # disable the create_symlink option because some file backends do not + # allow to create a symlink + if 'create_symlink' in self.args: + if self.args[ + 'create_symlink'] and not self.file_client.allow_symlink: + self.args['create_symlink'] = False + warnings.warn( + ('create_symlink is set as True by the user but is changed' + 'to be False because creating symbolic link is not ' + f'allowed in {self.file_client.name}')) + else: + self.args['create_symlink'] = self.file_client.allow_symlink + + def after_train_epoch(self, runner): + if not self.by_epoch: + return + + # save checkpoint for following cases: + # 1. every ``self.interval`` epochs + # 2. reach the last epoch of training + if self.every_n_epochs( + runner, self.interval) or (self.save_last + and self.is_last_epoch(runner)): + runner.logger.info( + f'Saving checkpoint at {runner.epoch + 1} epochs') + if self.sync_buffer: + allreduce_params(runner.model.buffers()) + self._save_checkpoint(runner) + + @master_only + def _save_checkpoint(self, runner): + """Save the current checkpoint and delete unwanted checkpoint.""" + runner.save_checkpoint( + self.out_dir, save_optimizer=self.save_optimizer, **self.args) + if runner.meta is not None: + if self.by_epoch: + cur_ckpt_filename = self.args.get( + 'filename_tmpl', 'epoch_{}.pth').format(runner.epoch + 1) + else: + cur_ckpt_filename = self.args.get( + 'filename_tmpl', 'iter_{}.pth').format(runner.iter + 1) + runner.meta.setdefault('hook_msgs', dict()) + runner.meta['hook_msgs']['last_ckpt'] = self.file_client.join_path( + self.out_dir, cur_ckpt_filename) + # remove other checkpoints + if self.max_keep_ckpts > 0: + if self.by_epoch: + name = 'epoch_{}.pth' + current_ckpt = runner.epoch + 1 + else: + name = 'iter_{}.pth' + current_ckpt = runner.iter + 1 + redundant_ckpts = range( + current_ckpt - self.max_keep_ckpts * self.interval, 0, + -self.interval) + filename_tmpl = self.args.get('filename_tmpl', name) + for _step in redundant_ckpts: + ckpt_path = self.file_client.join_path( + self.out_dir, filename_tmpl.format(_step)) + if self.file_client.isfile(ckpt_path): + self.file_client.remove(ckpt_path) + else: + break + + def after_train_iter(self, runner): + if self.by_epoch: + return + + # save checkpoint for following cases: + # 1. every ``self.interval`` iterations + # 2. reach the last iteration of training + if self.every_n_iters( + runner, self.interval) or (self.save_last + and self.is_last_iter(runner)): + runner.logger.info( + f'Saving checkpoint at {runner.iter + 1} iterations') + if self.sync_buffer: + allreduce_params(runner.model.buffers()) + self._save_checkpoint(runner) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/closure.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/closure.py new file mode 100644 index 0000000000000000000000000000000000000000..b955f81f425be4ac3e6bb3f4aac653887989e872 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/closure.py @@ -0,0 +1,11 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .hook import HOOKS, Hook + + +@HOOKS.register_module() +class ClosureHook(Hook): + + def __init__(self, fn_name, fn): + assert hasattr(self, fn_name) + assert callable(fn) + setattr(self, fn_name, fn) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/ema.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/ema.py new file mode 100644 index 0000000000000000000000000000000000000000..6ed77b84e116cc8e548e18dfd44a3bbfacf8491d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/ema.py @@ -0,0 +1,89 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from ...parallel import is_module_wrapper +from ..hooks.hook import HOOKS, Hook + + +@HOOKS.register_module() +class EMAHook(Hook): + r"""Exponential Moving Average Hook. + + Use Exponential Moving Average on all parameters of model in training + process. All parameters have a ema backup, which update by the formula + as below. EMAHook takes priority over EvalHook and CheckpointSaverHook. + + .. math:: + + Xema\_{t+1} = (1 - \text{momentum}) \times + Xema\_{t} + \text{momentum} \times X_t + + Args: + momentum (float): The momentum used for updating ema parameter. + Defaults to 0.0002. + interval (int): Update ema parameter every interval iteration. + Defaults to 1. + warm_up (int): During first warm_up steps, we may use smaller momentum + to update ema parameters more slowly. Defaults to 100. + resume_from (str): The checkpoint path. Defaults to None. + """ + + def __init__(self, + momentum=0.0002, + interval=1, + warm_up=100, + resume_from=None): + assert isinstance(interval, int) and interval > 0 + self.warm_up = warm_up + self.interval = interval + assert momentum > 0 and momentum < 1 + self.momentum = momentum**interval + self.checkpoint = resume_from + + def before_run(self, runner): + """To resume model with it's ema parameters more friendly. + + Register ema parameter as ``named_buffer`` to model + """ + model = runner.model + if is_module_wrapper(model): + model = model.module + self.param_ema_buffer = {} + self.model_parameters = dict(model.named_parameters(recurse=True)) + for name, value in self.model_parameters.items(): + # "." is not allowed in module's buffer name + buffer_name = f"ema_{name.replace('.', '_')}" + self.param_ema_buffer[name] = buffer_name + model.register_buffer(buffer_name, value.data.clone()) + self.model_buffers = dict(model.named_buffers(recurse=True)) + if self.checkpoint is not None: + runner.resume(self.checkpoint) + + def after_train_iter(self, runner): + """Update ema parameter every self.interval iterations.""" + curr_step = runner.iter + # We warm up the momentum considering the instability at beginning + momentum = min(self.momentum, + (1 + curr_step) / (self.warm_up + curr_step)) + if curr_step % self.interval != 0: + return + for name, parameter in self.model_parameters.items(): + buffer_name = self.param_ema_buffer[name] + buffer_parameter = self.model_buffers[buffer_name] + buffer_parameter.mul_(1 - momentum).add_(momentum, parameter.data) + + def after_train_epoch(self, runner): + """We load parameter values from ema backup to model before the + EvalHook.""" + self._swap_ema_parameters() + + def before_train_epoch(self, runner): + """We recover model's parameter from ema backup after last epoch's + EvalHook.""" + self._swap_ema_parameters() + + def _swap_ema_parameters(self): + """Swap the parameter of model with parameter in ema_buffer.""" + for name, value in self.model_parameters.items(): + temp = value.data.clone() + ema_buffer = self.model_buffers[self.param_ema_buffer[name]] + value.data.copy_(ema_buffer.data) + ema_buffer.data.copy_(temp) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/evaluation.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/evaluation.py new file mode 100644 index 0000000000000000000000000000000000000000..2a57d535edf5e18796042a0c71346f8f6f5ead52 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/evaluation.py @@ -0,0 +1,511 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import warnings +from math import inf + +import torch.distributed as dist +from torch.nn.modules.batchnorm import _BatchNorm +from torch.utils.data import DataLoader + +from mmcv.fileio import FileClient +from mmcv.utils import is_seq_of +from .hook import Hook +from .logger import LoggerHook + + +class EvalHook(Hook): + """Non-Distributed evaluation hook. + + This hook will regularly perform evaluation in a given interval when + performing in non-distributed environment. + + Args: + dataloader (DataLoader): A PyTorch dataloader, whose dataset has + implemented ``evaluate`` function. + start (int | None, optional): Evaluation starting epoch. It enables + evaluation before the training starts if ``start`` <= the resuming + epoch. If None, whether to evaluate is merely decided by + ``interval``. Default: None. + interval (int): Evaluation interval. Default: 1. + by_epoch (bool): Determine perform evaluation by epoch or by iteration. + If set to True, it will perform by epoch. Otherwise, by iteration. + Default: True. + save_best (str, optional): If a metric is specified, it would measure + the best checkpoint during evaluation. The information about best + checkpoint would be saved in ``runner.meta['hook_msgs']`` to keep + best score value and best checkpoint path, which will be also + loaded when resume checkpoint. Options are the evaluation metrics + on the test dataset. e.g., ``bbox_mAP``, ``segm_mAP`` for bbox + detection and instance segmentation. ``AR@100`` for proposal + recall. If ``save_best`` is ``auto``, the first key of the returned + ``OrderedDict`` result will be used. Default: None. + rule (str | None, optional): Comparison rule for best score. If set to + None, it will infer a reasonable rule. Keys such as 'acc', 'top' + .etc will be inferred by 'greater' rule. Keys contain 'loss' will + be inferred by 'less' rule. Options are 'greater', 'less', None. + Default: None. + test_fn (callable, optional): test a model with samples from a + dataloader, and return the test results. If ``None``, the default + test function ``mmcv.engine.single_gpu_test`` will be used. + (default: ``None``) + greater_keys (List[str] | None, optional): Metric keys that will be + inferred by 'greater' comparison rule. If ``None``, + _default_greater_keys will be used. (default: ``None``) + less_keys (List[str] | None, optional): Metric keys that will be + inferred by 'less' comparison rule. If ``None``, _default_less_keys + will be used. (default: ``None``) + out_dir (str, optional): The root directory to save checkpoints. If not + specified, `runner.work_dir` will be used by default. If specified, + the `out_dir` will be the concatenation of `out_dir` and the last + level directory of `runner.work_dir`. + `New in version 1.3.16.` + file_client_args (dict): Arguments to instantiate a FileClient. + See :class:`mmcv.fileio.FileClient` for details. Default: None. + `New in version 1.3.16.` + **eval_kwargs: Evaluation arguments fed into the evaluate function of + the dataset. + + Note: + If new arguments are added for EvalHook, tools/test.py, + tools/eval_metric.py may be affected. + """ + + # Since the key for determine greater or less is related to the downstream + # tasks, downstream repos may need to overwrite the following inner + # variable accordingly. + + rule_map = {'greater': lambda x, y: x > y, 'less': lambda x, y: x < y} + init_value_map = {'greater': -inf, 'less': inf} + _default_greater_keys = [ + 'acc', 'top', 'AR@', 'auc', 'precision', 'mAP', 'mDice', 'mIoU', + 'mAcc', 'aAcc' + ] + _default_less_keys = ['loss'] + + def __init__(self, + dataloader, + start=None, + interval=1, + by_epoch=True, + save_best=None, + rule=None, + test_fn=None, + greater_keys=None, + less_keys=None, + out_dir=None, + file_client_args=None, + **eval_kwargs): + if not isinstance(dataloader, DataLoader): + raise TypeError(f'dataloader must be a pytorch DataLoader, ' + f'but got {type(dataloader)}') + + if interval <= 0: + raise ValueError(f'interval must be a positive number, ' + f'but got {interval}') + + assert isinstance(by_epoch, bool), '``by_epoch`` should be a boolean' + + if start is not None and start < 0: + raise ValueError(f'The evaluation start epoch {start} is smaller ' + f'than 0') + + self.dataloader = dataloader + self.interval = interval + self.start = start + self.by_epoch = by_epoch + + assert isinstance(save_best, str) or save_best is None, \ + '""save_best"" should be a str or None ' \ + f'rather than {type(save_best)}' + self.save_best = save_best + self.eval_kwargs = eval_kwargs + self.initial_flag = True + + if test_fn is None: + from mmcv.engine import single_gpu_test + self.test_fn = single_gpu_test + else: + self.test_fn = test_fn + + if greater_keys is None: + self.greater_keys = self._default_greater_keys + else: + if not isinstance(greater_keys, (list, tuple)): + greater_keys = (greater_keys, ) + assert is_seq_of(greater_keys, str) + self.greater_keys = greater_keys + + if less_keys is None: + self.less_keys = self._default_less_keys + else: + if not isinstance(less_keys, (list, tuple)): + less_keys = (less_keys, ) + assert is_seq_of(less_keys, str) + self.less_keys = less_keys + + if self.save_best is not None: + self.best_ckpt_path = None + self._init_rule(rule, self.save_best) + + self.out_dir = out_dir + self.file_client_args = file_client_args + + def _init_rule(self, rule, key_indicator): + """Initialize rule, key_indicator, comparison_func, and best score. + + Here is the rule to determine which rule is used for key indicator + when the rule is not specific (note that the key indicator matching + is case-insensitive): + 1. If the key indicator is in ``self.greater_keys``, the rule will be + specified as 'greater'. + 2. Or if the key indicator is in ``self.less_keys``, the rule will be + specified as 'less'. + 3. Or if the key indicator is equal to the substring in any one item + in ``self.greater_keys``, the rule will be specified as 'greater'. + 4. Or if the key indicator is equal to the substring in any one item + in ``self.less_keys``, the rule will be specified as 'less'. + + Args: + rule (str | None): Comparison rule for best score. + key_indicator (str | None): Key indicator to determine the + comparison rule. + """ + if rule not in self.rule_map and rule is not None: + raise KeyError(f'rule must be greater, less or None, ' + f'but got {rule}.') + + if rule is None: + if key_indicator != 'auto': + # `_lc` here means we use the lower case of keys for + # case-insensitive matching + key_indicator_lc = key_indicator.lower() + greater_keys = [key.lower() for key in self.greater_keys] + less_keys = [key.lower() for key in self.less_keys] + + if key_indicator_lc in greater_keys: + rule = 'greater' + elif key_indicator_lc in less_keys: + rule = 'less' + elif any(key in key_indicator_lc for key in greater_keys): + rule = 'greater' + elif any(key in key_indicator_lc for key in less_keys): + rule = 'less' + else: + raise ValueError(f'Cannot infer the rule for key ' + f'{key_indicator}, thus a specific rule ' + f'must be specified.') + self.rule = rule + self.key_indicator = key_indicator + if self.rule is not None: + self.compare_func = self.rule_map[self.rule] + + def before_run(self, runner): + if not self.out_dir: + self.out_dir = runner.work_dir + + self.file_client = FileClient.infer_client(self.file_client_args, + self.out_dir) + + # if `self.out_dir` is not equal to `runner.work_dir`, it means that + # `self.out_dir` is set so the final `self.out_dir` is the + # concatenation of `self.out_dir` and the last level directory of + # `runner.work_dir` + if self.out_dir != runner.work_dir: + basename = osp.basename(runner.work_dir.rstrip(osp.sep)) + self.out_dir = self.file_client.join_path(self.out_dir, basename) + runner.logger.info( + (f'The best checkpoint will be saved to {self.out_dir} by ' + f'{self.file_client.name}')) + + if self.save_best is not None: + if runner.meta is None: + warnings.warn('runner.meta is None. Creating an empty one.') + runner.meta = dict() + runner.meta.setdefault('hook_msgs', dict()) + self.best_ckpt_path = runner.meta['hook_msgs'].get( + 'best_ckpt', None) + + def before_train_iter(self, runner): + """Evaluate the model only at the start of training by iteration.""" + if self.by_epoch or not self.initial_flag: + return + if self.start is not None and runner.iter >= self.start: + self.after_train_iter(runner) + self.initial_flag = False + + def before_train_epoch(self, runner): + """Evaluate the model only at the start of training by epoch.""" + if not (self.by_epoch and self.initial_flag): + return + if self.start is not None and runner.epoch >= self.start: + self.after_train_epoch(runner) + self.initial_flag = False + + def after_train_iter(self, runner): + """Called after every training iter to evaluate the results.""" + if not self.by_epoch and self._should_evaluate(runner): + # Because the priority of EvalHook is higher than LoggerHook, the + # training log and the evaluating log are mixed. Therefore, + # we need to dump the training log and clear it before evaluating + # log is generated. In addition, this problem will only appear in + # `IterBasedRunner` whose `self.by_epoch` is False, because + # `EpochBasedRunner` whose `self.by_epoch` is True calls + # `_do_evaluate` in `after_train_epoch` stage, and at this stage + # the training log has been printed, so it will not cause any + # problem. more details at + # https://github.com/open-mmlab/mmsegmentation/issues/694 + for hook in runner._hooks: + if isinstance(hook, LoggerHook): + hook.after_train_iter(runner) + runner.log_buffer.clear() + + self._do_evaluate(runner) + + def after_train_epoch(self, runner): + """Called after every training epoch to evaluate the results.""" + if self.by_epoch and self._should_evaluate(runner): + self._do_evaluate(runner) + + def _do_evaluate(self, runner): + """perform evaluation and save ckpt.""" + results = self.test_fn(runner.model, self.dataloader) + runner.log_buffer.output['eval_iter_num'] = len(self.dataloader) + key_score = self.evaluate(runner, results) + # the key_score may be `None` so it needs to skip the action to save + # the best checkpoint + if self.save_best and key_score: + self._save_ckpt(runner, key_score) + + def _should_evaluate(self, runner): + """Judge whether to perform evaluation. + + Here is the rule to judge whether to perform evaluation: + 1. It will not perform evaluation during the epoch/iteration interval, + which is determined by ``self.interval``. + 2. It will not perform evaluation if the start time is larger than + current time. + 3. It will not perform evaluation when current time is larger than + the start time but during epoch/iteration interval. + + Returns: + bool: The flag indicating whether to perform evaluation. + """ + if self.by_epoch: + current = runner.epoch + check_time = self.every_n_epochs + else: + current = runner.iter + check_time = self.every_n_iters + + if self.start is None: + if not check_time(runner, self.interval): + # No evaluation during the interval. + return False + elif (current + 1) < self.start: + # No evaluation if start is larger than the current time. + return False + else: + # Evaluation only at epochs/iters 3, 5, 7... + # if start==3 and interval==2 + if (current + 1 - self.start) % self.interval: + return False + return True + + def _save_ckpt(self, runner, key_score): + """Save the best checkpoint. + + It will compare the score according to the compare function, write + related information (best score, best checkpoint path) and save the + best checkpoint into ``work_dir``. + """ + if self.by_epoch: + current = f'epoch_{runner.epoch + 1}' + cur_type, cur_time = 'epoch', runner.epoch + 1 + else: + current = f'iter_{runner.iter + 1}' + cur_type, cur_time = 'iter', runner.iter + 1 + + best_score = runner.meta['hook_msgs'].get( + 'best_score', self.init_value_map[self.rule]) + if self.compare_func(key_score, best_score): + best_score = key_score + runner.meta['hook_msgs']['best_score'] = best_score + + if self.best_ckpt_path and self.file_client.isfile( + self.best_ckpt_path): + self.file_client.remove(self.best_ckpt_path) + runner.logger.info( + (f'The previous best checkpoint {self.best_ckpt_path} was ' + 'removed')) + + best_ckpt_name = f'best_{self.key_indicator}_{current}.pth' + self.best_ckpt_path = self.file_client.join_path( + self.out_dir, best_ckpt_name) + runner.meta['hook_msgs']['best_ckpt'] = self.best_ckpt_path + + runner.save_checkpoint( + self.out_dir, + filename_tmpl=best_ckpt_name, + create_symlink=False) + runner.logger.info( + f'Now best checkpoint is saved as {best_ckpt_name}.') + runner.logger.info( + f'Best {self.key_indicator} is {best_score:0.4f} ' + f'at {cur_time} {cur_type}.') + + def evaluate(self, runner, results): + """Evaluate the results. + + Args: + runner (:obj:`mmcv.Runner`): The underlined training runner. + results (list): Output results. + """ + eval_res = self.dataloader.dataset.evaluate( + results, logger=runner.logger, **self.eval_kwargs) + + for name, val in eval_res.items(): + runner.log_buffer.output[name] = val + runner.log_buffer.ready = True + + if self.save_best is not None: + # If the performance of model is pool, the `eval_res` may be an + # empty dict and it will raise exception when `self.save_best` is + # not None. More details at + # https://github.com/open-mmlab/mmdetection/issues/6265. + if not eval_res: + warnings.warn( + 'Since `eval_res` is an empty dict, the behavior to save ' + 'the best checkpoint will be skipped in this evaluation.') + return None + + if self.key_indicator == 'auto': + # infer from eval_results + self._init_rule(self.rule, list(eval_res.keys())[0]) + return eval_res[self.key_indicator] + + return None + + +class DistEvalHook(EvalHook): + """Distributed evaluation hook. + + This hook will regularly perform evaluation in a given interval when + performing in distributed environment. + + Args: + dataloader (DataLoader): A PyTorch dataloader, whose dataset has + implemented ``evaluate`` function. + start (int | None, optional): Evaluation starting epoch. It enables + evaluation before the training starts if ``start`` <= the resuming + epoch. If None, whether to evaluate is merely decided by + ``interval``. Default: None. + interval (int): Evaluation interval. Default: 1. + by_epoch (bool): Determine perform evaluation by epoch or by iteration. + If set to True, it will perform by epoch. Otherwise, by iteration. + default: True. + save_best (str, optional): If a metric is specified, it would measure + the best checkpoint during evaluation. The information about best + checkpoint would be saved in ``runner.meta['hook_msgs']`` to keep + best score value and best checkpoint path, which will be also + loaded when resume checkpoint. Options are the evaluation metrics + on the test dataset. e.g., ``bbox_mAP``, ``segm_mAP`` for bbox + detection and instance segmentation. ``AR@100`` for proposal + recall. If ``save_best`` is ``auto``, the first key of the returned + ``OrderedDict`` result will be used. Default: None. + rule (str | None, optional): Comparison rule for best score. If set to + None, it will infer a reasonable rule. Keys such as 'acc', 'top' + .etc will be inferred by 'greater' rule. Keys contain 'loss' will + be inferred by 'less' rule. Options are 'greater', 'less', None. + Default: None. + test_fn (callable, optional): test a model with samples from a + dataloader in a multi-gpu manner, and return the test results. If + ``None``, the default test function ``mmcv.engine.multi_gpu_test`` + will be used. (default: ``None``) + tmpdir (str | None): Temporary directory to save the results of all + processes. Default: None. + gpu_collect (bool): Whether to use gpu or cpu to collect results. + Default: False. + broadcast_bn_buffer (bool): Whether to broadcast the + buffer(running_mean and running_var) of rank 0 to other rank + before evaluation. Default: True. + out_dir (str, optional): The root directory to save checkpoints. If not + specified, `runner.work_dir` will be used by default. If specified, + the `out_dir` will be the concatenation of `out_dir` and the last + level directory of `runner.work_dir`. + file_client_args (dict): Arguments to instantiate a FileClient. + See :class:`mmcv.fileio.FileClient` for details. Default: None. + **eval_kwargs: Evaluation arguments fed into the evaluate function of + the dataset. + """ + + def __init__(self, + dataloader, + start=None, + interval=1, + by_epoch=True, + save_best=None, + rule=None, + test_fn=None, + greater_keys=None, + less_keys=None, + broadcast_bn_buffer=True, + tmpdir=None, + gpu_collect=False, + out_dir=None, + file_client_args=None, + **eval_kwargs): + + if test_fn is None: + from mmcv.engine import multi_gpu_test + test_fn = multi_gpu_test + + super().__init__( + dataloader, + start=start, + interval=interval, + by_epoch=by_epoch, + save_best=save_best, + rule=rule, + test_fn=test_fn, + greater_keys=greater_keys, + less_keys=less_keys, + out_dir=out_dir, + file_client_args=file_client_args, + **eval_kwargs) + + self.broadcast_bn_buffer = broadcast_bn_buffer + self.tmpdir = tmpdir + self.gpu_collect = gpu_collect + + def _do_evaluate(self, runner): + """perform evaluation and save ckpt.""" + # Synchronization of BatchNorm's buffer (running_mean + # and running_var) is not supported in the DDP of pytorch, + # which may cause the inconsistent performance of models in + # different ranks, so we broadcast BatchNorm's buffers + # of rank 0 to other ranks to avoid this. + if self.broadcast_bn_buffer: + model = runner.model + for name, module in model.named_modules(): + if isinstance(module, + _BatchNorm) and module.track_running_stats: + dist.broadcast(module.running_var, 0) + dist.broadcast(module.running_mean, 0) + + tmpdir = self.tmpdir + if tmpdir is None: + tmpdir = osp.join(runner.work_dir, '.eval_hook') + + results = self.test_fn( + runner.model, + self.dataloader, + tmpdir=tmpdir, + gpu_collect=self.gpu_collect) + if runner.rank == 0: + print('\n') + runner.log_buffer.output['eval_iter_num'] = len(self.dataloader) + key_score = self.evaluate(runner, results) + # the key_score may be `None` so it needs to skip the action to + # save the best checkpoint + if self.save_best and key_score: + self._save_ckpt(runner, key_score) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/hook.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/hook.py new file mode 100644 index 0000000000000000000000000000000000000000..f2d1c9865b5a6b5927029774cce9731187c127ed --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/hook.py @@ -0,0 +1,92 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.utils import Registry, is_method_overridden + +HOOKS = Registry('hook') + + +class Hook: + stages = ('before_run', 'before_train_epoch', 'before_train_iter', + 'after_train_iter', 'after_train_epoch', 'before_val_epoch', + 'before_val_iter', 'after_val_iter', 'after_val_epoch', + 'after_run') + + def before_run(self, runner): + pass + + def after_run(self, runner): + pass + + def before_epoch(self, runner): + pass + + def after_epoch(self, runner): + pass + + def before_iter(self, runner): + pass + + def after_iter(self, runner): + pass + + def before_train_epoch(self, runner): + self.before_epoch(runner) + + def before_val_epoch(self, runner): + self.before_epoch(runner) + + def after_train_epoch(self, runner): + self.after_epoch(runner) + + def after_val_epoch(self, runner): + self.after_epoch(runner) + + def before_train_iter(self, runner): + self.before_iter(runner) + + def before_val_iter(self, runner): + self.before_iter(runner) + + def after_train_iter(self, runner): + self.after_iter(runner) + + def after_val_iter(self, runner): + self.after_iter(runner) + + def every_n_epochs(self, runner, n): + return (runner.epoch + 1) % n == 0 if n > 0 else False + + def every_n_inner_iters(self, runner, n): + return (runner.inner_iter + 1) % n == 0 if n > 0 else False + + def every_n_iters(self, runner, n): + return (runner.iter + 1) % n == 0 if n > 0 else False + + def end_of_epoch(self, runner): + return runner.inner_iter + 1 == len(runner.data_loader) + + def is_last_epoch(self, runner): + return runner.epoch + 1 == runner._max_epochs + + def is_last_iter(self, runner): + return runner.iter + 1 == runner._max_iters + + def get_triggered_stages(self): + trigger_stages = set() + for stage in Hook.stages: + if is_method_overridden(stage, Hook, self): + trigger_stages.add(stage) + + # some methods will be triggered in multi stages + # use this dict to map method to stages. + method_stages_map = { + 'before_epoch': ['before_train_epoch', 'before_val_epoch'], + 'after_epoch': ['after_train_epoch', 'after_val_epoch'], + 'before_iter': ['before_train_iter', 'before_val_iter'], + 'after_iter': ['after_train_iter', 'after_val_iter'], + } + + for method, map_stages in method_stages_map.items(): + if is_method_overridden(method, Hook, self): + trigger_stages.update(map_stages) + + return [stage for stage in Hook.stages if stage in trigger_stages] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/iter_timer.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/iter_timer.py new file mode 100644 index 0000000000000000000000000000000000000000..6bde4d52c35e0a13dd26a215853e2470072b09bc --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/iter_timer.py @@ -0,0 +1,30 @@ +# Copyright (c) 2022, Shanghai Iluvatar CoreX Semiconductor Co., Ltd. +# Copyright (c) OpenMMLab. All rights reserved. + +import time +import torch.distributed as dist + +from .hook import HOOKS, Hook + + +@HOOKS.register_module() +class IterTimerHook(Hook): + + def before_epoch(self, runner): + self.t = time.time() + + def before_iter(self, runner): + self.batch_size = runner.cfg.data.samples_per_gpu + runner.log_buffer.update({'data_time': time.time() - self.t}) + + def after_iter(self, runner): + iter_info = {'time': time.time() - self.t} + fps = self.batch_size / iter_info["time"] + + if dist.is_initialized(): + fps = fps * dist.get_world_size() + + iter_info["fps"] = fps + + runner.log_buffer.update(iter_info) + self.t = time.time() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..062709e704f08fc313a7f422cc7cd1e34bde5f68 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/__init__.py @@ -0,0 +1,18 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .base import LoggerHook +from .clearml import ClearMLLoggerHook +from .dvclive import DvcliveLoggerHook +from .mlflow import MlflowLoggerHook +from .neptune import NeptuneLoggerHook +from .pavi import PaviLoggerHook +from .segmind import SegmindLoggerHook +from .tensorboard import TensorboardLoggerHook +from .text import TextLoggerHook +from .wandb import WandbLoggerHook + +__all__ = [ + 'LoggerHook', 'MlflowLoggerHook', 'PaviLoggerHook', + 'TensorboardLoggerHook', 'TextLoggerHook', 'WandbLoggerHook', + 'NeptuneLoggerHook', 'DvcliveLoggerHook', 'SegmindLoggerHook', + 'ClearMLLoggerHook' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/base.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/base.py new file mode 100644 index 0000000000000000000000000000000000000000..9f1a51b3168f770448afd7a72f26d02ff34c9b07 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/base.py @@ -0,0 +1,167 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import numbers +from abc import ABCMeta, abstractmethod + +import numpy as np +import torch + +from ..hook import Hook + + +class LoggerHook(Hook): + """Base class for logger hooks. + + Args: + interval (int): Logging interval (every k iterations). Default 10. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. Default True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default False. + by_epoch (bool): Whether EpochBasedRunner is used. Default True. + """ + + __metaclass__ = ABCMeta + + def __init__(self, + interval=10, + ignore_last=True, + reset_flag=False, + by_epoch=True): + self.interval = interval + self.ignore_last = ignore_last + self.reset_flag = reset_flag + self.by_epoch = by_epoch + + @abstractmethod + def log(self, runner): + pass + + @staticmethod + def is_scalar(val, include_np=True, include_torch=True): + """Tell the input variable is a scalar or not. + + Args: + val: Input variable. + include_np (bool): Whether include 0-d np.ndarray as a scalar. + include_torch (bool): Whether include 0-d torch.Tensor as a scalar. + + Returns: + bool: True or False. + """ + if isinstance(val, numbers.Number): + return True + elif include_np and isinstance(val, np.ndarray) and val.ndim == 0: + return True + elif include_torch and isinstance(val, torch.Tensor) and len(val) == 1: + return True + else: + return False + + def get_mode(self, runner): + if runner.mode == 'train': + if 'time' in runner.log_buffer.output: + mode = 'train' + else: + mode = 'val' + elif runner.mode == 'val': + mode = 'val' + else: + raise ValueError(f"runner mode should be 'train' or 'val', " + f'but got {runner.mode}') + return mode + + def get_epoch(self, runner): + if runner.mode == 'train': + epoch = runner.epoch + 1 + elif runner.mode == 'val': + # normal val mode + # runner.epoch += 1 has been done before val workflow + epoch = runner.epoch + else: + raise ValueError(f"runner mode should be 'train' or 'val', " + f'but got {runner.mode}') + return epoch + + def get_iter(self, runner, inner_iter=False): + """Get the current training iteration step.""" + if self.by_epoch and inner_iter: + current_iter = runner.inner_iter + 1 + else: + current_iter = runner.iter + 1 + return current_iter + + def get_lr_tags(self, runner): + tags = {} + lrs = runner.current_lr() + if isinstance(lrs, dict): + for name, value in lrs.items(): + tags[f'learning_rate/{name}'] = value[0] + else: + tags['learning_rate'] = lrs[0] + return tags + + def get_momentum_tags(self, runner): + tags = {} + momentums = runner.current_momentum() + if isinstance(momentums, dict): + for name, value in momentums.items(): + tags[f'momentum/{name}'] = value[0] + else: + tags['momentum'] = momentums[0] + return tags + + def get_loggable_tags(self, + runner, + allow_scalar=True, + allow_text=False, + add_mode=True, + tags_to_skip=('time', 'data_time')): + tags = {} + for var, val in runner.log_buffer.output.items(): + if var in tags_to_skip: + continue + if self.is_scalar(val) and not allow_scalar: + continue + if isinstance(val, str) and not allow_text: + continue + if add_mode: + var = f'{self.get_mode(runner)}/{var}' + tags[var] = val + tags.update(self.get_lr_tags(runner)) + tags.update(self.get_momentum_tags(runner)) + return tags + + def before_run(self, runner): + for hook in runner.hooks[::-1]: + if isinstance(hook, LoggerHook): + hook.reset_flag = True + break + + def before_epoch(self, runner): + runner.log_buffer.clear() # clear logs of last epoch + + def after_train_iter(self, runner): + if self.by_epoch and self.every_n_inner_iters(runner, self.interval): + runner.log_buffer.average(self.interval) + elif not self.by_epoch and self.every_n_iters(runner, self.interval): + runner.log_buffer.average(self.interval) + elif self.end_of_epoch(runner) and not self.ignore_last: + # not precise but more stable + runner.log_buffer.average(self.interval) + + if runner.log_buffer.ready: + self.log(runner) + if self.reset_flag: + runner.log_buffer.clear_output() + + def after_train_epoch(self, runner): + if runner.log_buffer.ready: + self.log(runner) + if self.reset_flag: + runner.log_buffer.clear_output() + + def after_val_epoch(self, runner): + runner.log_buffer.average() + self.log(runner) + if self.reset_flag: + runner.log_buffer.clear_output() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/clearml.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/clearml.py new file mode 100644 index 0000000000000000000000000000000000000000..1de71e152b7bdf28285bd8925b134236a4d528e2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/clearml.py @@ -0,0 +1,62 @@ +# Copyright (c) OpenMMLab. All rights reserved. + +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class ClearMLLoggerHook(LoggerHook): + """Class to log metrics with clearml. + + It requires `clearml`_ to be installed. + + + Args: + init_kwargs (dict): A dict contains the `clearml.Task.init` + initialization keys. See `taskinit`_ for more details. + interval (int): Logging interval (every k iterations). Default 10. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. Default: True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default: False. + by_epoch (bool): Whether EpochBasedRunner is used. Default: True. + + .. _clearml: + https://clear.ml/docs/latest/docs/ + .. _taskinit: + https://clear.ml/docs/latest/docs/references/sdk/task/#taskinit + """ + + def __init__(self, + init_kwargs=None, + interval=10, + ignore_last=True, + reset_flag=False, + by_epoch=True): + super(ClearMLLoggerHook, self).__init__(interval, ignore_last, + reset_flag, by_epoch) + self.import_clearml() + self.init_kwargs = init_kwargs + + def import_clearml(self): + try: + import clearml + except ImportError: + raise ImportError( + 'Please run "pip install clearml" to install clearml') + self.clearml = clearml + + @master_only + def before_run(self, runner): + super(ClearMLLoggerHook, self).before_run(runner) + task_kwargs = self.init_kwargs if self.init_kwargs else {} + self.task = self.clearml.Task.init(**task_kwargs) + self.task_logger = self.task.get_logger() + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner) + for tag, val in tags.items(): + self.task_logger.report_scalar(tag, tag, val, + self.get_iter(runner)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/dvclive.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/dvclive.py new file mode 100644 index 0000000000000000000000000000000000000000..c79eefa75ed634d05b91b71eaf07049585e1be80 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/dvclive.py @@ -0,0 +1,68 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from pathlib import Path + +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class DvcliveLoggerHook(LoggerHook): + """Class to log metrics with dvclive. + + It requires `dvclive`_ to be installed. + + Args: + model_file (str): Default None. If not None, after each epoch the + model will be saved to {model_file}. + interval (int): Logging interval (every k iterations). Default 10. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. Default: True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default: False. + by_epoch (bool): Whether EpochBasedRunner is used. Default: True. + kwargs: Arguments for instantiating `Live`_. + + .. _dvclive: + https://dvc.org/doc/dvclive + + .. _Live: + https://dvc.org/doc/dvclive/api-reference/live#parameters + """ + + def __init__(self, + model_file=None, + interval=10, + ignore_last=True, + reset_flag=False, + by_epoch=True, + **kwargs): + super().__init__(interval, ignore_last, reset_flag, by_epoch) + self.model_file = model_file + self.import_dvclive(**kwargs) + + def import_dvclive(self, **kwargs): + try: + from dvclive import Live + except ImportError: + raise ImportError( + 'Please run "pip install dvclive" to install dvclive') + self.dvclive = Live(**kwargs) + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner) + if tags: + self.dvclive.set_step(self.get_iter(runner)) + for k, v in tags.items(): + self.dvclive.log(k, v) + + @master_only + def after_train_epoch(self, runner): + super().after_train_epoch(runner) + if self.model_file is not None: + runner.save_checkpoint( + Path(self.model_file).parent, + filename_tmpl=Path(self.model_file).name, + create_symlink=False, + ) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/mlflow.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/mlflow.py new file mode 100644 index 0000000000000000000000000000000000000000..dcd87bcb5e0826fbaa9fa10176f7be833c774eba --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/mlflow.py @@ -0,0 +1,80 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.utils import TORCH_VERSION +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class MlflowLoggerHook(LoggerHook): + """Class to log metrics and (optionally) a trained model to MLflow. + + It requires `MLflow`_ to be installed. + + Args: + exp_name (str, optional): Name of the experiment to be used. + Default None. If not None, set the active experiment. + If experiment does not exist, an experiment with provided name + will be created. + tags (Dict[str], optional): Tags for the current run. + Default None. If not None, set tags for the current run. + log_model (bool, optional): Whether to log an MLflow artifact. + Default True. If True, log runner.model as an MLflow artifact + for the current run. + interval (int): Logging interval (every k iterations). Default: 10. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. Default: True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default: False. + by_epoch (bool): Whether EpochBasedRunner is used. Default: True. + + .. _MLflow: + https://www.mlflow.org/docs/latest/index.html + """ + + def __init__(self, + exp_name=None, + tags=None, + log_model=True, + interval=10, + ignore_last=True, + reset_flag=False, + by_epoch=True): + super(MlflowLoggerHook, self).__init__(interval, ignore_last, + reset_flag, by_epoch) + self.import_mlflow() + self.exp_name = exp_name + self.tags = tags + self.log_model = log_model + + def import_mlflow(self): + try: + import mlflow + import mlflow.pytorch as mlflow_pytorch + except ImportError: + raise ImportError( + 'Please run "pip install mlflow" to install mlflow') + self.mlflow = mlflow + self.mlflow_pytorch = mlflow_pytorch + + @master_only + def before_run(self, runner): + super(MlflowLoggerHook, self).before_run(runner) + if self.exp_name is not None: + self.mlflow.set_experiment(self.exp_name) + if self.tags is not None: + self.mlflow.set_tags(self.tags) + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner) + if tags: + self.mlflow.log_metrics(tags, step=self.get_iter(runner)) + + @master_only + def after_run(self, runner): + if self.log_model: + self.mlflow_pytorch.log_model( + runner.model, + 'models', + pip_requirements=[f'torch=={TORCH_VERSION}']) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/neptune.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/neptune.py new file mode 100644 index 0000000000000000000000000000000000000000..e0aafe91d3dd87357ef69b0abb9fb57dc19b6194 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/neptune.py @@ -0,0 +1,88 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class NeptuneLoggerHook(LoggerHook): + """Class to log metrics to NeptuneAI. + + It requires `Neptune`_ to be installed. + + Args: + init_kwargs (dict): a dict contains the initialization keys as below: + + - project (str): Name of a project in a form of + namespace/project_name. If None, the value of NEPTUNE_PROJECT + environment variable will be taken. + - api_token (str): User’s API token. If None, the value of + NEPTUNE_API_TOKEN environment variable will be taken. Note: It is + strongly recommended to use NEPTUNE_API_TOKEN environment + variable rather than placing your API token in plain text in your + source code. + - name (str, optional, default is 'Untitled'): Editable name of the + run. Name is displayed in the run's Details and in Runs table as + a column. + + Check https://docs.neptune.ai/api-reference/neptune#init for more + init arguments. + interval (int): Logging interval (every k iterations). Default: 10. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than ``interval``. Default: True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default: True. + with_step (bool): If True, the step will be logged from + ``self.get_iters``. Otherwise, step will not be logged. + Default: True. + by_epoch (bool): Whether EpochBasedRunner is used. Default: True. + + .. _Neptune: + https://docs.neptune.ai + """ + + def __init__(self, + init_kwargs=None, + interval=10, + ignore_last=True, + reset_flag=True, + with_step=True, + by_epoch=True): + + super(NeptuneLoggerHook, self).__init__(interval, ignore_last, + reset_flag, by_epoch) + self.import_neptune() + self.init_kwargs = init_kwargs + self.with_step = with_step + + def import_neptune(self): + try: + import neptune.new as neptune + except ImportError: + raise ImportError( + 'Please run "pip install neptune-client" to install neptune') + self.neptune = neptune + self.run = None + + @master_only + def before_run(self, runner): + if self.init_kwargs: + self.run = self.neptune.init(**self.init_kwargs) + else: + self.run = self.neptune.init() + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner) + if tags: + for tag_name, tag_value in tags.items(): + if self.with_step: + self.run[tag_name].log( + tag_value, step=self.get_iter(runner)) + else: + tags['global_step'] = self.get_iter(runner) + self.run[tag_name].log(tags) + + @master_only + def after_run(self, runner): + self.run.stop() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/pavi.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/pavi.py new file mode 100644 index 0000000000000000000000000000000000000000..d5d61f9e505148e6271fccb935908c19a1c5822e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/pavi.py @@ -0,0 +1,132 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import json +import os +import os.path as osp + +import torch +import yaml + +import mmcv +from ....parallel.utils import is_module_wrapper +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class PaviLoggerHook(LoggerHook): + """Class to visual model, log metrics (for internal use). + + Args: + init_kwargs (dict): A dict contains the initialization keys. + add_graph (bool): Whether to visual model. Default: False. + add_last_ckpt (bool): Whether to save checkpoint after run. + Default: False. + interval (int): Logging interval (every k iterations). Default: True. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. Default: True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default: False. + by_epoch (bool): Whether EpochBasedRunner is used. Default: True. + img_key (string): Get image data from Dataset. Default: 'img_info'. + """ + + def __init__(self, + init_kwargs=None, + add_graph=False, + add_last_ckpt=False, + interval=10, + ignore_last=True, + reset_flag=False, + by_epoch=True, + img_key='img_info'): + super(PaviLoggerHook, self).__init__(interval, ignore_last, reset_flag, + by_epoch) + self.init_kwargs = init_kwargs + self.add_graph = add_graph + self.add_last_ckpt = add_last_ckpt + self.img_key = img_key + + @master_only + def before_run(self, runner): + super(PaviLoggerHook, self).before_run(runner) + try: + from pavi import SummaryWriter + except ImportError: + raise ImportError('Please run "pip install pavi" to install pavi.') + + self.run_name = runner.work_dir.split('/')[-1] + + if not self.init_kwargs: + self.init_kwargs = dict() + self.init_kwargs['name'] = self.run_name + self.init_kwargs['model'] = runner._model_name + if runner.meta is not None: + if 'config_dict' in runner.meta: + config_dict = runner.meta['config_dict'] + assert isinstance( + config_dict, + dict), ('meta["config_dict"] has to be of a dict, ' + f'but got {type(config_dict)}') + elif 'config_file' in runner.meta: + config_file = runner.meta['config_file'] + config_dict = dict(mmcv.Config.fromfile(config_file)) + else: + config_dict = None + if config_dict is not None: + # 'max_.*iter' is parsed in pavi sdk as the maximum iterations + # to properly set up the progress bar. + config_dict = config_dict.copy() + config_dict.setdefault('max_iter', runner.max_iters) + # non-serializable values are first converted in + # mmcv.dump to json + config_dict = json.loads( + mmcv.dump(config_dict, file_format='json')) + session_text = yaml.dump(config_dict) + self.init_kwargs['session_text'] = session_text + self.writer = SummaryWriter(**self.init_kwargs) + + def get_step(self, runner): + """Get the total training step/epoch.""" + if self.get_mode(runner) == 'val' and self.by_epoch: + return self.get_epoch(runner) + else: + return self.get_iter(runner) + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner, add_mode=False) + if tags: + self.writer.add_scalars( + self.get_mode(runner), tags, self.get_step(runner)) + + @master_only + def after_run(self, runner): + if self.add_last_ckpt: + ckpt_path = osp.join(runner.work_dir, 'latest.pth') + if osp.islink(ckpt_path): + ckpt_path = osp.join(runner.work_dir, os.readlink(ckpt_path)) + + if osp.isfile(ckpt_path): + # runner.epoch += 1 has been done before `after_run`. + iteration = runner.epoch if self.by_epoch else runner.iter + return self.writer.add_snapshot_file( + tag=self.run_name, + snapshot_file_path=ckpt_path, + iteration=iteration) + + # flush the buffer and send a task ending signal to Pavi + self.writer.close() + + @master_only + def before_epoch(self, runner): + if runner.epoch == 0 and self.add_graph: + if is_module_wrapper(runner.model): + _model = runner.model.module + else: + _model = runner.model + device = next(_model.parameters()).device + data = next(iter(runner.data_loader)) + image = data[self.img_key][0:1].to(device) + with torch.no_grad(): + self.writer.add_graph(_model, image) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/segmind.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/segmind.py new file mode 100644 index 0000000000000000000000000000000000000000..e262c7c1aa3905c966355200151746c21930cab5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/segmind.py @@ -0,0 +1,49 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class SegmindLoggerHook(LoggerHook): + """Class to log metrics to Segmind. + + It requires `Segmind`_ to be installed. + + Args: + interval (int): Logging interval (every k iterations). Default: 10. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. Default True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default False. + by_epoch (bool): Whether EpochBasedRunner is used. Default True. + + .. _Segmind: + https://docs.segmind.com/python-library + """ + + def __init__(self, + interval=10, + ignore_last=True, + reset_flag=False, + by_epoch=True): + super(SegmindLoggerHook, self).__init__(interval, ignore_last, + reset_flag, by_epoch) + self.import_segmind() + + def import_segmind(self): + try: + import segmind + except ImportError: + raise ImportError( + "Please run 'pip install segmind' to install segmind") + self.log_metrics = segmind.tracking.fluent.log_metrics + self.mlflow_log = segmind.utils.logging_utils.try_mlflow_log + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner) + if tags: + # logging metrics to segmind + self.mlflow_log( + self.log_metrics, tags, step=runner.epoch, epoch=runner.epoch) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/tensorboard.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/tensorboard.py new file mode 100644 index 0000000000000000000000000000000000000000..bf00d5742cb7a693f75205202c6178c6d96173ba --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/tensorboard.py @@ -0,0 +1,69 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp + +from mmcv.utils import TORCH_VERSION, digit_version +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class TensorboardLoggerHook(LoggerHook): + """Class to log metrics to Tensorboard. + + Args: + log_dir (string): Save directory location. Default: None. If default + values are used, directory location is ``runner.work_dir``/tf_logs. + interval (int): Logging interval (every k iterations). Default: True. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. Default: True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default: False. + by_epoch (bool): Whether EpochBasedRunner is used. Default: True. + """ + + def __init__(self, + log_dir=None, + interval=10, + ignore_last=True, + reset_flag=False, + by_epoch=True): + super(TensorboardLoggerHook, self).__init__(interval, ignore_last, + reset_flag, by_epoch) + self.log_dir = log_dir + + @master_only + def before_run(self, runner): + super(TensorboardLoggerHook, self).before_run(runner) + if (TORCH_VERSION == 'parrots' + or digit_version(TORCH_VERSION) < digit_version('1.1')): + try: + from tensorboardX import SummaryWriter + except ImportError: + raise ImportError('Please install tensorboardX to use ' + 'TensorboardLoggerHook.') + else: + try: + from torch.utils.tensorboard import SummaryWriter + except ImportError: + raise ImportError( + 'Please run "pip install future tensorboard" to install ' + 'the dependencies to use torch.utils.tensorboard ' + '(applicable to PyTorch 1.1 or higher)') + + if self.log_dir is None: + self.log_dir = osp.join(runner.work_dir, 'tf_logs') + self.writer = SummaryWriter(self.log_dir) + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner, allow_text=True) + for tag, val in tags.items(): + if isinstance(val, str): + self.writer.add_text(tag, val, self.get_iter(runner)) + else: + self.writer.add_scalar(tag, val, self.get_iter(runner)) + + @master_only + def after_run(self, runner): + self.writer.close() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/text.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/text.py new file mode 100644 index 0000000000000000000000000000000000000000..644ced2c996b22b163399b0cb3a102773abe8195 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/text.py @@ -0,0 +1,256 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import datetime +import os +import os.path as osp +from collections import OrderedDict + +import torch +import torch.distributed as dist + +import mmcv +from mmcv.fileio.file_client import FileClient +from mmcv.utils import is_tuple_of, scandir +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class TextLoggerHook(LoggerHook): + """Logger hook in text. + + In this logger hook, the information will be printed on terminal and + saved in json file. + + Args: + by_epoch (bool, optional): Whether EpochBasedRunner is used. + Default: True. + interval (int, optional): Logging interval (every k iterations). + Default: 10. + ignore_last (bool, optional): Ignore the log of last iterations in each + epoch if less than :attr:`interval`. Default: True. + reset_flag (bool, optional): Whether to clear the output buffer after + logging. Default: False. + interval_exp_name (int, optional): Logging interval for experiment + name. This feature is to help users conveniently get the experiment + information from screen or log file. Default: 1000. + out_dir (str, optional): Logs are saved in ``runner.work_dir`` default. + If ``out_dir`` is specified, logs will be copied to a new directory + which is the concatenation of ``out_dir`` and the last level + directory of ``runner.work_dir``. Default: None. + `New in version 1.3.16.` + out_suffix (str or tuple[str], optional): Those filenames ending with + ``out_suffix`` will be copied to ``out_dir``. + Default: ('.log.json', '.log', '.py'). + `New in version 1.3.16.` + keep_local (bool, optional): Whether to keep local log when + :attr:`out_dir` is specified. If False, the local log will be + removed. Default: True. + `New in version 1.3.16.` + file_client_args (dict, optional): Arguments to instantiate a + FileClient. See :class:`mmcv.fileio.FileClient` for details. + Default: None. + `New in version 1.3.16.` + """ + + def __init__(self, + by_epoch=True, + interval=10, + ignore_last=True, + reset_flag=False, + interval_exp_name=1000, + out_dir=None, + out_suffix=('.log.json', '.log', '.py'), + keep_local=True, + file_client_args=None): + super(TextLoggerHook, self).__init__(interval, ignore_last, reset_flag, + by_epoch) + self.by_epoch = by_epoch + self.time_sec_tot = 0 + self.interval_exp_name = interval_exp_name + + if out_dir is None and file_client_args is not None: + raise ValueError( + 'file_client_args should be "None" when `out_dir` is not' + 'specified.') + self.out_dir = out_dir + + if not (out_dir is None or isinstance(out_dir, str) + or is_tuple_of(out_dir, str)): + raise TypeError('out_dir should be "None" or string or tuple of ' + 'string, but got {out_dir}') + self.out_suffix = out_suffix + + self.keep_local = keep_local + self.file_client_args = file_client_args + if self.out_dir is not None: + self.file_client = FileClient.infer_client(file_client_args, + self.out_dir) + + def before_run(self, runner): + super(TextLoggerHook, self).before_run(runner) + + if self.out_dir is not None: + self.file_client = FileClient.infer_client(self.file_client_args, + self.out_dir) + # The final `self.out_dir` is the concatenation of `self.out_dir` + # and the last level directory of `runner.work_dir` + basename = osp.basename(runner.work_dir.rstrip(osp.sep)) + self.out_dir = self.file_client.join_path(self.out_dir, basename) + runner.logger.info( + (f'Text logs will be saved to {self.out_dir} by ' + f'{self.file_client.name} after the training process.')) + + self.start_iter = runner.iter + self.json_log_path = osp.join(runner.work_dir, + f'{runner.timestamp}.log.json') + if runner.meta is not None: + self._dump_log(runner.meta, runner) + + def _get_max_memory(self, runner): + device = getattr(runner.model, 'output_device', None) + mem = torch.cuda.max_memory_allocated(device=device) + mem_mb = torch.tensor([int(mem) // (1024 * 1024)], + dtype=torch.int, + device=device) + if runner.world_size > 1: + dist.reduce(mem_mb, 0, op=dist.ReduceOp.MAX) + return mem_mb.item() + + def _log_info(self, log_dict, runner): + # print exp name for users to distinguish experiments + # at every ``interval_exp_name`` iterations and the end of each epoch + if runner.meta is not None and 'exp_name' in runner.meta: + if (self.every_n_iters(runner, self.interval_exp_name)) or ( + self.by_epoch and self.end_of_epoch(runner)): + exp_info = f'Exp name: {runner.meta["exp_name"]}' + runner.logger.info(exp_info) + + if log_dict['mode'] == 'train': + if isinstance(log_dict['lr'], dict): + lr_str = [] + for k, val in log_dict['lr'].items(): + lr_str.append(f'lr_{k}: {val:.3e}') + lr_str = ' '.join(lr_str) + else: + lr_str = f'lr: {log_dict["lr"]:.3e}' + + # by epoch: Epoch [4][100/1000] + # by iter: Iter [100/100000] + if self.by_epoch: + log_str = f'Epoch [{log_dict["epoch"]}]' \ + f'[{log_dict["iter"]}/{len(runner.data_loader)}]\t' + else: + log_str = f'Iter [{log_dict["iter"]}/{runner.max_iters}]\t' + log_str += f'{lr_str}, ' + + if 'time' in log_dict.keys(): + self.time_sec_tot += (log_dict['time'] * self.interval) + time_sec_avg = self.time_sec_tot / ( + runner.iter - self.start_iter + 1) + eta_sec = time_sec_avg * (runner.max_iters - runner.iter - 1) + eta_str = str(datetime.timedelta(seconds=int(eta_sec))) + log_str += f'eta: {eta_str}, ' + log_str += f'time: {log_dict["time"]:.3f}, ' \ + f'data_time: {log_dict["data_time"]:.3f}, ' + # statistic memory + if torch.cuda.is_available(): + log_str += f'memory: {log_dict["memory"]}, ' + else: + # val/test time + # here 1000 is the length of the val dataloader + # by epoch: Epoch[val] [4][1000] + # by iter: Iter[val] [1000] + if self.by_epoch: + log_str = f'Epoch({log_dict["mode"]}) ' \ + f'[{log_dict["epoch"]}][{log_dict["iter"]}]\t' + else: + log_str = f'Iter({log_dict["mode"]}) [{log_dict["iter"]}]\t' + + log_items = [] + for name, val in log_dict.items(): + # TODO: resolve this hack + # these items have been in log_str + if name in [ + 'mode', 'Epoch', 'iter', 'lr', 'time', 'data_time', + 'memory', 'epoch' + ]: + continue + if isinstance(val, float): + val = f'{val:.4f}' + log_items.append(f'{name}: {val}') + log_str += ', '.join(log_items) + + runner.logger.info(log_str) + + def _dump_log(self, log_dict, runner): + # dump log in json format + json_log = OrderedDict() + for k, v in log_dict.items(): + json_log[k] = self._round_float(v) + # only append log at last line + if runner.rank == 0: + with open(self.json_log_path, 'a+') as f: + mmcv.dump(json_log, f, file_format='json') + f.write('\n') + + def _round_float(self, items): + if isinstance(items, list): + return [self._round_float(item) for item in items] + elif isinstance(items, float): + return round(items, 5) + else: + return items + + def log(self, runner): + if 'eval_iter_num' in runner.log_buffer.output: + # this doesn't modify runner.iter and is regardless of by_epoch + cur_iter = runner.log_buffer.output.pop('eval_iter_num') + else: + cur_iter = self.get_iter(runner, inner_iter=True) + + log_dict = OrderedDict( + mode=self.get_mode(runner), + epoch=self.get_epoch(runner), + iter=cur_iter) + + # only record lr of the first param group + cur_lr = runner.current_lr() + if isinstance(cur_lr, list): + log_dict['lr'] = cur_lr[0] + else: + assert isinstance(cur_lr, dict) + log_dict['lr'] = {} + for k, lr_ in cur_lr.items(): + assert isinstance(lr_, list) + log_dict['lr'].update({k: lr_[0]}) + + if 'time' in runner.log_buffer.output: + # statistic memory + if torch.cuda.is_available(): + log_dict['memory'] = self._get_max_memory(runner) + + log_dict = dict(log_dict, **runner.log_buffer.output) + + self._log_info(log_dict, runner) + self._dump_log(log_dict, runner) + return log_dict + + def after_run(self, runner): + # copy or upload logs to self.out_dir + if self.out_dir is not None: + for filename in scandir(runner.work_dir, self.out_suffix, True): + local_filepath = osp.join(runner.work_dir, filename) + out_filepath = self.file_client.join_path( + self.out_dir, filename) + with open(local_filepath, 'r') as f: + self.file_client.put_text(f.read(), out_filepath) + + runner.logger.info( + (f'The file {local_filepath} has been uploaded to ' + f'{out_filepath}.')) + + if not self.keep_local: + os.remove(local_filepath) + runner.logger.info( + (f'{local_filepath} was removed due to the ' + '`self.keep_local=False`')) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/wandb.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/wandb.py new file mode 100644 index 0000000000000000000000000000000000000000..78b890ee1bd855ae6b68674e9646b414cf2c4634 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/logger/wandb.py @@ -0,0 +1,107 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp + +from mmcv.utils import scandir +from ...dist_utils import master_only +from ..hook import HOOKS +from .base import LoggerHook + + +@HOOKS.register_module() +class WandbLoggerHook(LoggerHook): + """Class to log metrics with wandb. + + It requires `wandb`_ to be installed. + + + Args: + init_kwargs (dict): A dict contains the initialization keys. Check + https://docs.wandb.ai/ref/python/init for more init arguments. + interval (int): Logging interval (every k iterations). + Default 10. + ignore_last (bool): Ignore the log of last iterations in each epoch + if less than `interval`. + Default: True. + reset_flag (bool): Whether to clear the output buffer after logging. + Default: False. + commit (bool): Save the metrics dict to the wandb server and increment + the step. If false ``wandb.log`` just updates the current metrics + dict with the row argument and metrics won't be saved until + ``wandb.log`` is called with ``commit=True``. + Default: True. + by_epoch (bool): Whether EpochBasedRunner is used. + Default: True. + with_step (bool): If True, the step will be logged from + ``self.get_iters``. Otherwise, step will not be logged. + Default: True. + log_artifact (bool): If True, artifacts in {work_dir} will be uploaded + to wandb after training ends. + Default: True + `New in version 1.4.3.` + out_suffix (str or tuple[str], optional): Those filenames ending with + ``out_suffix`` will be uploaded to wandb. + Default: ('.log.json', '.log', '.py'). + `New in version 1.4.3.` + + .. _wandb: + https://docs.wandb.ai + """ + + def __init__(self, + init_kwargs=None, + interval=10, + ignore_last=True, + reset_flag=False, + commit=True, + by_epoch=True, + with_step=True, + log_artifact=True, + out_suffix=('.log.json', '.log', '.py')): + super(WandbLoggerHook, self).__init__(interval, ignore_last, + reset_flag, by_epoch) + self.import_wandb() + self.init_kwargs = init_kwargs + self.commit = commit + self.with_step = with_step + self.log_artifact = log_artifact + self.out_suffix = out_suffix + + def import_wandb(self): + try: + import wandb + except ImportError: + raise ImportError( + 'Please run "pip install wandb" to install wandb') + self.wandb = wandb + + @master_only + def before_run(self, runner): + super(WandbLoggerHook, self).before_run(runner) + if self.wandb is None: + self.import_wandb() + if self.init_kwargs: + self.wandb.init(**self.init_kwargs) + else: + self.wandb.init() + + @master_only + def log(self, runner): + tags = self.get_loggable_tags(runner) + if tags: + if self.with_step: + self.wandb.log( + tags, step=self.get_iter(runner), commit=self.commit) + else: + tags['global_step'] = self.get_iter(runner) + self.wandb.log(tags, commit=self.commit) + + @master_only + def after_run(self, runner): + if self.log_artifact: + wandb_artifact = self.wandb.Artifact( + name='artifacts', type='model') + for filename in scandir(runner.work_dir, self.out_suffix, True): + local_filepath = osp.join(runner.work_dir, filename) + wandb_artifact.add_file(local_filepath) + self.wandb.log_artifact(wandb_artifact) + self.wandb.join() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/lr_updater.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/lr_updater.py new file mode 100644 index 0000000000000000000000000000000000000000..ee2a53a65076ec2e8443978dcd16d19723e6a97e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/lr_updater.py @@ -0,0 +1,730 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import numbers +from math import cos, pi + +import mmcv +from .hook import HOOKS, Hook + + +class LrUpdaterHook(Hook): + """LR Scheduler in MMCV. + + Args: + by_epoch (bool): LR changes epoch by epoch + warmup (string): Type of warmup used. It can be None(use no warmup), + 'constant', 'linear' or 'exp' + warmup_iters (int): The number of iterations or epochs that warmup + lasts + warmup_ratio (float): LR used at the beginning of warmup equals to + warmup_ratio * initial_lr + warmup_by_epoch (bool): When warmup_by_epoch == True, warmup_iters + means the number of epochs that warmup lasts, otherwise means the + number of iteration that warmup lasts + """ + + def __init__(self, + by_epoch=True, + warmup=None, + warmup_iters=0, + warmup_ratio=0.1, + warmup_by_epoch=False): + # validate the "warmup" argument + if warmup is not None: + if warmup not in ['constant', 'linear', 'exp']: + raise ValueError( + f'"{warmup}" is not a supported type for warming up, valid' + ' types are "constant" and "linear"') + if warmup is not None: + assert warmup_iters > 0, \ + '"warmup_iters" must be a positive integer' + assert 0 < warmup_ratio <= 1.0, \ + '"warmup_ratio" must be in range (0,1]' + + self.by_epoch = by_epoch + self.warmup = warmup + self.warmup_iters = warmup_iters + self.warmup_ratio = warmup_ratio + self.warmup_by_epoch = warmup_by_epoch + + if self.warmup_by_epoch: + self.warmup_epochs = self.warmup_iters + self.warmup_iters = None + else: + self.warmup_epochs = None + + self.base_lr = [] # initial lr for all param groups + self.regular_lr = [] # expected lr if no warming up is performed + + def _set_lr(self, runner, lr_groups): + if isinstance(runner.optimizer, dict): + for k, optim in runner.optimizer.items(): + for param_group, lr in zip(optim.param_groups, lr_groups[k]): + param_group['lr'] = lr + else: + for param_group, lr in zip(runner.optimizer.param_groups, + lr_groups): + param_group['lr'] = lr + + def get_lr(self, runner, base_lr): + raise NotImplementedError + + def get_regular_lr(self, runner): + if isinstance(runner.optimizer, dict): + lr_groups = {} + for k in runner.optimizer.keys(): + _lr_group = [ + self.get_lr(runner, _base_lr) + for _base_lr in self.base_lr[k] + ] + lr_groups.update({k: _lr_group}) + + return lr_groups + else: + return [self.get_lr(runner, _base_lr) for _base_lr in self.base_lr] + + def get_warmup_lr(self, cur_iters): + + def _get_warmup_lr(cur_iters, regular_lr): + if self.warmup == 'constant': + warmup_lr = [_lr * self.warmup_ratio for _lr in regular_lr] + elif self.warmup == 'linear': + k = (1 - cur_iters / self.warmup_iters) * (1 - + self.warmup_ratio) + warmup_lr = [_lr * (1 - k) for _lr in regular_lr] + elif self.warmup == 'exp': + k = self.warmup_ratio**(1 - cur_iters / self.warmup_iters) + warmup_lr = [_lr * k for _lr in regular_lr] + return warmup_lr + + if isinstance(self.regular_lr, dict): + lr_groups = {} + for key, regular_lr in self.regular_lr.items(): + lr_groups[key] = _get_warmup_lr(cur_iters, regular_lr) + return lr_groups + else: + return _get_warmup_lr(cur_iters, self.regular_lr) + + def before_run(self, runner): + # NOTE: when resuming from a checkpoint, if 'initial_lr' is not saved, + # it will be set according to the optimizer params + if isinstance(runner.optimizer, dict): + self.base_lr = {} + for k, optim in runner.optimizer.items(): + for group in optim.param_groups: + group.setdefault('initial_lr', group['lr']) + _base_lr = [ + group['initial_lr'] for group in optim.param_groups + ] + self.base_lr.update({k: _base_lr}) + else: + for group in runner.optimizer.param_groups: + group.setdefault('initial_lr', group['lr']) + self.base_lr = [ + group['initial_lr'] for group in runner.optimizer.param_groups + ] + + def before_train_epoch(self, runner): + if self.warmup_iters is None: + epoch_len = len(runner.data_loader) + self.warmup_iters = self.warmup_epochs * epoch_len + + if not self.by_epoch: + return + + self.regular_lr = self.get_regular_lr(runner) + self._set_lr(runner, self.regular_lr) + + def before_train_iter(self, runner): + cur_iter = runner.iter + if not self.by_epoch: + self.regular_lr = self.get_regular_lr(runner) + if self.warmup is None or cur_iter >= self.warmup_iters: + self._set_lr(runner, self.regular_lr) + else: + warmup_lr = self.get_warmup_lr(cur_iter) + self._set_lr(runner, warmup_lr) + elif self.by_epoch: + if self.warmup is None or cur_iter > self.warmup_iters: + return + elif cur_iter == self.warmup_iters: + self._set_lr(runner, self.regular_lr) + else: + warmup_lr = self.get_warmup_lr(cur_iter) + self._set_lr(runner, warmup_lr) + + +@HOOKS.register_module() +class FixedLrUpdaterHook(LrUpdaterHook): + + def __init__(self, **kwargs): + super(FixedLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + return base_lr + + +@HOOKS.register_module() +class StepLrUpdaterHook(LrUpdaterHook): + """Step LR scheduler with min_lr clipping. + + Args: + step (int | list[int]): Step to decay the LR. If an int value is given, + regard it as the decay interval. If a list is given, decay LR at + these steps. + gamma (float, optional): Decay LR ratio. Default: 0.1. + min_lr (float, optional): Minimum LR value to keep. If LR after decay + is lower than `min_lr`, it will be clipped to this value. If None + is given, we don't perform lr clipping. Default: None. + """ + + def __init__(self, step, gamma=0.1, min_lr=None, **kwargs): + if isinstance(step, list): + assert mmcv.is_list_of(step, int) + assert all([s > 0 for s in step]) + elif isinstance(step, int): + assert step > 0 + else: + raise TypeError('"step" must be a list or integer') + self.step = step + self.gamma = gamma + self.min_lr = min_lr + super(StepLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + progress = runner.epoch if self.by_epoch else runner.iter + + # calculate exponential term + if isinstance(self.step, int): + exp = progress // self.step + else: + exp = len(self.step) + for i, s in enumerate(self.step): + if progress < s: + exp = i + break + + lr = base_lr * (self.gamma**exp) + if self.min_lr is not None: + # clip to a minimum value + lr = max(lr, self.min_lr) + return lr + + +@HOOKS.register_module() +class ExpLrUpdaterHook(LrUpdaterHook): + + def __init__(self, gamma, **kwargs): + self.gamma = gamma + super(ExpLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + progress = runner.epoch if self.by_epoch else runner.iter + return base_lr * self.gamma**progress + + +@HOOKS.register_module() +class PolyLrUpdaterHook(LrUpdaterHook): + + def __init__(self, power=1., min_lr=0., **kwargs): + self.power = power + self.min_lr = min_lr + super(PolyLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + if self.by_epoch: + progress = runner.epoch + max_progress = runner.max_epochs + else: + progress = runner.iter + max_progress = runner.max_iters + coeff = (1 - progress / max_progress)**self.power + return (base_lr - self.min_lr) * coeff + self.min_lr + + +@HOOKS.register_module() +class InvLrUpdaterHook(LrUpdaterHook): + + def __init__(self, gamma, power=1., **kwargs): + self.gamma = gamma + self.power = power + super(InvLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + progress = runner.epoch if self.by_epoch else runner.iter + return base_lr * (1 + self.gamma * progress)**(-self.power) + + +@HOOKS.register_module() +class CosineAnnealingLrUpdaterHook(LrUpdaterHook): + """CosineAnnealing LR scheduler. + + Args: + min_lr (float, optional): The minimum lr. Default: None. + min_lr_ratio (float, optional): The ratio of minimum lr to the base lr. + Either `min_lr` or `min_lr_ratio` should be specified. + Default: None. + """ + + def __init__(self, min_lr=None, min_lr_ratio=None, **kwargs): + assert (min_lr is None) ^ (min_lr_ratio is None) + self.min_lr = min_lr + self.min_lr_ratio = min_lr_ratio + super(CosineAnnealingLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + if self.by_epoch: + progress = runner.epoch + max_progress = runner.max_epochs + else: + progress = runner.iter + max_progress = runner.max_iters + + if self.min_lr_ratio is not None: + target_lr = base_lr * self.min_lr_ratio + else: + target_lr = self.min_lr + return annealing_cos(base_lr, target_lr, progress / max_progress) + + +@HOOKS.register_module() +class FlatCosineAnnealingLrUpdaterHook(LrUpdaterHook): + """Flat + Cosine lr schedule. + + Modified from https://github.com/fastai/fastai/blob/master/fastai/callback/schedule.py#L128 # noqa: E501 + + Args: + start_percent (float): When to start annealing the learning rate + after the percentage of the total training steps. + The value should be in range [0, 1). + Default: 0.75 + min_lr (float, optional): The minimum lr. Default: None. + min_lr_ratio (float, optional): The ratio of minimum lr to the base lr. + Either `min_lr` or `min_lr_ratio` should be specified. + Default: None. + """ + + def __init__(self, + start_percent=0.75, + min_lr=None, + min_lr_ratio=None, + **kwargs): + assert (min_lr is None) ^ (min_lr_ratio is None) + if start_percent < 0 or start_percent > 1 or not isinstance( + start_percent, float): + raise ValueError( + 'expected float between 0 and 1 start_percent, but ' + f'got {start_percent}') + self.start_percent = start_percent + self.min_lr = min_lr + self.min_lr_ratio = min_lr_ratio + super(FlatCosineAnnealingLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + if self.by_epoch: + start = round(runner.max_epochs * self.start_percent) + progress = runner.epoch - start + max_progress = runner.max_epochs - start + else: + start = round(runner.max_iters * self.start_percent) + progress = runner.iter - start + max_progress = runner.max_iters - start + + if self.min_lr_ratio is not None: + target_lr = base_lr * self.min_lr_ratio + else: + target_lr = self.min_lr + + if progress < 0: + return base_lr + else: + return annealing_cos(base_lr, target_lr, progress / max_progress) + + +@HOOKS.register_module() +class CosineRestartLrUpdaterHook(LrUpdaterHook): + """Cosine annealing with restarts learning rate scheme. + + Args: + periods (list[int]): Periods for each cosine anneling cycle. + restart_weights (list[float], optional): Restart weights at each + restart iteration. Default: [1]. + min_lr (float, optional): The minimum lr. Default: None. + min_lr_ratio (float, optional): The ratio of minimum lr to the base lr. + Either `min_lr` or `min_lr_ratio` should be specified. + Default: None. + """ + + def __init__(self, + periods, + restart_weights=[1], + min_lr=None, + min_lr_ratio=None, + **kwargs): + assert (min_lr is None) ^ (min_lr_ratio is None) + self.periods = periods + self.min_lr = min_lr + self.min_lr_ratio = min_lr_ratio + self.restart_weights = restart_weights + assert (len(self.periods) == len(self.restart_weights) + ), 'periods and restart_weights should have the same length.' + super(CosineRestartLrUpdaterHook, self).__init__(**kwargs) + + self.cumulative_periods = [ + sum(self.periods[0:i + 1]) for i in range(0, len(self.periods)) + ] + + def get_lr(self, runner, base_lr): + if self.by_epoch: + progress = runner.epoch + else: + progress = runner.iter + + if self.min_lr_ratio is not None: + target_lr = base_lr * self.min_lr_ratio + else: + target_lr = self.min_lr + + idx = get_position_from_periods(progress, self.cumulative_periods) + current_weight = self.restart_weights[idx] + nearest_restart = 0 if idx == 0 else self.cumulative_periods[idx - 1] + current_periods = self.periods[idx] + + alpha = min((progress - nearest_restart) / current_periods, 1) + return annealing_cos(base_lr, target_lr, alpha, current_weight) + + +def get_position_from_periods(iteration, cumulative_periods): + """Get the position from a period list. + + It will return the index of the right-closest number in the period list. + For example, the cumulative_periods = [100, 200, 300, 400], + if iteration == 50, return 0; + if iteration == 210, return 2; + if iteration == 300, return 3. + + Args: + iteration (int): Current iteration. + cumulative_periods (list[int]): Cumulative period list. + + Returns: + int: The position of the right-closest number in the period list. + """ + for i, period in enumerate(cumulative_periods): + if iteration < period: + return i + raise ValueError(f'Current iteration {iteration} exceeds ' + f'cumulative_periods {cumulative_periods}') + + +@HOOKS.register_module() +class CyclicLrUpdaterHook(LrUpdaterHook): + """Cyclic LR Scheduler. + + Implement the cyclical learning rate policy (CLR) described in + https://arxiv.org/pdf/1506.01186.pdf + + Different from the original paper, we use cosine annealing rather than + triangular policy inside a cycle. This improves the performance in the + 3D detection area. + + Args: + by_epoch (bool, optional): Whether to update LR by epoch. + target_ratio (tuple[float], optional): Relative ratio of the highest LR + and the lowest LR to the initial LR. + cyclic_times (int, optional): Number of cycles during training + step_ratio_up (float, optional): The ratio of the increasing process of + LR in the total cycle. + anneal_strategy (str, optional): {'cos', 'linear'} + Specifies the annealing strategy: 'cos' for cosine annealing, + 'linear' for linear annealing. Default: 'cos'. + gamma (float, optional): Cycle decay ratio. Default: 1. + It takes values in the range (0, 1]. The difference between the + maximum learning rate and the minimum learning rate decreases + periodically when it is less than 1. `New in version 1.4.4.` + """ + + def __init__(self, + by_epoch=False, + target_ratio=(10, 1e-4), + cyclic_times=1, + step_ratio_up=0.4, + anneal_strategy='cos', + gamma=1, + **kwargs): + if isinstance(target_ratio, float): + target_ratio = (target_ratio, target_ratio / 1e5) + elif isinstance(target_ratio, tuple): + target_ratio = (target_ratio[0], target_ratio[0] / 1e5) \ + if len(target_ratio) == 1 else target_ratio + else: + raise ValueError('target_ratio should be either float ' + f'or tuple, got {type(target_ratio)}') + + assert len(target_ratio) == 2, \ + '"target_ratio" must be list or tuple of two floats' + assert 0 <= step_ratio_up < 1.0, \ + '"step_ratio_up" must be in range [0,1)' + assert 0 < gamma <= 1, \ + '"gamma" must be in range (0, 1]' + + self.target_ratio = target_ratio + self.cyclic_times = cyclic_times + self.step_ratio_up = step_ratio_up + self.gamma = gamma + self.max_iter_per_phase = None + self.lr_phases = [] # init lr_phases + # validate anneal_strategy + if anneal_strategy not in ['cos', 'linear']: + raise ValueError('anneal_strategy must be one of "cos" or ' + f'"linear", instead got {anneal_strategy}') + elif anneal_strategy == 'cos': + self.anneal_func = annealing_cos + elif anneal_strategy == 'linear': + self.anneal_func = annealing_linear + + assert not by_epoch, \ + 'currently only support "by_epoch" = False' + super(CyclicLrUpdaterHook, self).__init__(by_epoch, **kwargs) + + def before_run(self, runner): + super(CyclicLrUpdaterHook, self).before_run(runner) + # initiate lr_phases + # total lr_phases are separated as up and down + self.max_iter_per_phase = runner.max_iters // self.cyclic_times + iter_up_phase = int(self.step_ratio_up * self.max_iter_per_phase) + self.lr_phases.append([0, iter_up_phase, 1, self.target_ratio[0]]) + self.lr_phases.append([ + iter_up_phase, self.max_iter_per_phase, self.target_ratio[0], + self.target_ratio[1] + ]) + + def get_lr(self, runner, base_lr): + curr_iter = runner.iter % self.max_iter_per_phase + curr_cycle = runner.iter // self.max_iter_per_phase + # Update weight decay + scale = self.gamma**curr_cycle + + for (start_iter, end_iter, start_ratio, end_ratio) in self.lr_phases: + if start_iter <= curr_iter < end_iter: + # Apply cycle scaling to gradually reduce the difference + # between max_lr and base lr. The target end_ratio can be + # expressed as: + # end_ratio = (base_lr + scale * (max_lr - base_lr)) / base_lr + # iteration: 0-iter_up_phase: + if start_iter == 0: + end_ratio = 1 - scale + end_ratio * scale + # iteration: iter_up_phase-self.max_iter_per_phase + else: + start_ratio = 1 - scale + start_ratio * scale + progress = curr_iter - start_iter + return self.anneal_func(base_lr * start_ratio, + base_lr * end_ratio, + progress / (end_iter - start_iter)) + + +@HOOKS.register_module() +class OneCycleLrUpdaterHook(LrUpdaterHook): + """One Cycle LR Scheduler. + + The 1cycle learning rate policy changes the learning rate after every + batch. The one cycle learning rate policy is described in + https://arxiv.org/pdf/1708.07120.pdf + + Args: + max_lr (float or list): Upper learning rate boundaries in the cycle + for each parameter group. + total_steps (int, optional): The total number of steps in the cycle. + Note that if a value is not provided here, it will be the max_iter + of runner. Default: None. + pct_start (float): The percentage of the cycle (in number of steps) + spent increasing the learning rate. + Default: 0.3 + anneal_strategy (str): {'cos', 'linear'} + Specifies the annealing strategy: 'cos' for cosine annealing, + 'linear' for linear annealing. + Default: 'cos' + div_factor (float): Determines the initial learning rate via + initial_lr = max_lr/div_factor + Default: 25 + final_div_factor (float): Determines the minimum learning rate via + min_lr = initial_lr/final_div_factor + Default: 1e4 + three_phase (bool): If three_phase is True, use a third phase of the + schedule to annihilate the learning rate according to + final_div_factor instead of modifying the second phase (the first + two phases will be symmetrical about the step indicated by + pct_start). + Default: False + """ + + def __init__(self, + max_lr, + total_steps=None, + pct_start=0.3, + anneal_strategy='cos', + div_factor=25, + final_div_factor=1e4, + three_phase=False, + **kwargs): + # validate by_epoch, currently only support by_epoch = False + if 'by_epoch' not in kwargs: + kwargs['by_epoch'] = False + else: + assert not kwargs['by_epoch'], \ + 'currently only support "by_epoch" = False' + if not isinstance(max_lr, (numbers.Number, list, dict)): + raise ValueError('the type of max_lr must be the one of list or ' + f'dict, but got {type(max_lr)}') + self._max_lr = max_lr + if total_steps is not None: + if not isinstance(total_steps, int): + raise ValueError('the type of total_steps must be int, but' + f'got {type(total_steps)}') + self.total_steps = total_steps + # validate pct_start + if pct_start < 0 or pct_start > 1 or not isinstance(pct_start, float): + raise ValueError('expected float between 0 and 1 pct_start, but ' + f'got {pct_start}') + self.pct_start = pct_start + # validate anneal_strategy + if anneal_strategy not in ['cos', 'linear']: + raise ValueError('anneal_strategy must be one of "cos" or ' + f'"linear", instead got {anneal_strategy}') + elif anneal_strategy == 'cos': + self.anneal_func = annealing_cos + elif anneal_strategy == 'linear': + self.anneal_func = annealing_linear + self.div_factor = div_factor + self.final_div_factor = final_div_factor + self.three_phase = three_phase + self.lr_phases = [] # init lr_phases + super(OneCycleLrUpdaterHook, self).__init__(**kwargs) + + def before_run(self, runner): + if hasattr(self, 'total_steps'): + total_steps = self.total_steps + else: + total_steps = runner.max_iters + if total_steps < runner.max_iters: + raise ValueError( + 'The total steps must be greater than or equal to max ' + f'iterations {runner.max_iters} of runner, but total steps ' + f'is {total_steps}.') + + if isinstance(runner.optimizer, dict): + self.base_lr = {} + for k, optim in runner.optimizer.items(): + _max_lr = format_param(k, optim, self._max_lr) + self.base_lr[k] = [lr / self.div_factor for lr in _max_lr] + for group, lr in zip(optim.param_groups, self.base_lr[k]): + group.setdefault('initial_lr', lr) + else: + k = type(runner.optimizer).__name__ + _max_lr = format_param(k, runner.optimizer, self._max_lr) + self.base_lr = [lr / self.div_factor for lr in _max_lr] + for group, lr in zip(runner.optimizer.param_groups, self.base_lr): + group.setdefault('initial_lr', lr) + + if self.three_phase: + self.lr_phases.append( + [float(self.pct_start * total_steps) - 1, 1, self.div_factor]) + self.lr_phases.append([ + float(2 * self.pct_start * total_steps) - 2, self.div_factor, 1 + ]) + self.lr_phases.append( + [total_steps - 1, 1, 1 / self.final_div_factor]) + else: + self.lr_phases.append( + [float(self.pct_start * total_steps) - 1, 1, self.div_factor]) + self.lr_phases.append( + [total_steps - 1, self.div_factor, 1 / self.final_div_factor]) + + def get_lr(self, runner, base_lr): + curr_iter = runner.iter + start_iter = 0 + for i, (end_iter, start_lr, end_lr) in enumerate(self.lr_phases): + if curr_iter <= end_iter: + pct = (curr_iter - start_iter) / (end_iter - start_iter) + lr = self.anneal_func(base_lr * start_lr, base_lr * end_lr, + pct) + break + start_iter = end_iter + return lr + + +@HOOKS.register_module() +class LinearAnnealingLrUpdaterHook(LrUpdaterHook): + """Linear annealing LR Scheduler decays the learning rate of each parameter + group linearly. + + Args: + min_lr (float, optional): The minimum lr. Default: None. + min_lr_ratio (float, optional): The ratio of minimum lr to the base lr. + Either `min_lr` or `min_lr_ratio` should be specified. + Default: None. + """ + + def __init__(self, min_lr=None, min_lr_ratio=None, **kwargs): + assert (min_lr is None) ^ (min_lr_ratio is None) + self.min_lr = min_lr + self.min_lr_ratio = min_lr_ratio + super(LinearAnnealingLrUpdaterHook, self).__init__(**kwargs) + + def get_lr(self, runner, base_lr): + if self.by_epoch: + progress = runner.epoch + max_progress = runner.max_epochs + else: + progress = runner.iter + max_progress = runner.max_iters + if self.min_lr_ratio is not None: + target_lr = base_lr * self.min_lr_ratio + else: + target_lr = self.min_lr + return annealing_linear(base_lr, target_lr, progress / max_progress) + + +def annealing_cos(start, end, factor, weight=1): + """Calculate annealing cos learning rate. + + Cosine anneal from `weight * start + (1 - weight) * end` to `end` as + percentage goes from 0.0 to 1.0. + + Args: + start (float): The starting learning rate of the cosine annealing. + end (float): The ending learing rate of the cosine annealing. + factor (float): The coefficient of `pi` when calculating the current + percentage. Range from 0.0 to 1.0. + weight (float, optional): The combination factor of `start` and `end` + when calculating the actual starting learning rate. Default to 1. + """ + cos_out = cos(pi * factor) + 1 + return end + 0.5 * weight * (start - end) * cos_out + + +def annealing_linear(start, end, factor): + """Calculate annealing linear learning rate. + + Linear anneal from `start` to `end` as percentage goes from 0.0 to 1.0. + + Args: + start (float): The starting learning rate of the linear annealing. + end (float): The ending learing rate of the linear annealing. + factor (float): The coefficient of `pi` when calculating the current + percentage. Range from 0.0 to 1.0. + """ + return start + (end - start) * factor + + +def format_param(name, optim, param): + if isinstance(param, numbers.Number): + return [param] * len(optim.param_groups) + elif isinstance(param, (list, tuple)): # multi param groups + if len(param) != len(optim.param_groups): + raise ValueError(f'expected {len(optim.param_groups)} ' + f'values for {name}, got {len(param)}') + return param + else: # multi optimizers + if name not in param: + raise KeyError(f'{name} is not found in {param.keys()}') + return param[name] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/memory.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/memory.py new file mode 100644 index 0000000000000000000000000000000000000000..70cf9a838fb314e3bd3c07aadbc00921a81e83ed --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/memory.py @@ -0,0 +1,25 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch + +from .hook import HOOKS, Hook + + +@HOOKS.register_module() +class EmptyCacheHook(Hook): + + def __init__(self, before_epoch=False, after_epoch=True, after_iter=False): + self._before_epoch = before_epoch + self._after_epoch = after_epoch + self._after_iter = after_iter + + def after_iter(self, runner): + if self._after_iter: + torch.cuda.empty_cache() + + def before_epoch(self, runner): + if self._before_epoch: + torch.cuda.empty_cache() + + def after_epoch(self, runner): + if self._after_epoch: + torch.cuda.empty_cache() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/momentum_updater.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/momentum_updater.py new file mode 100644 index 0000000000000000000000000000000000000000..aa15fe23c81e0a6d3d9b518c3b67f3778b167c6a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/momentum_updater.py @@ -0,0 +1,566 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import mmcv +from .hook import HOOKS, Hook +from .lr_updater import annealing_cos, annealing_linear, format_param + + +class MomentumUpdaterHook(Hook): + + def __init__(self, + by_epoch=True, + warmup=None, + warmup_iters=0, + warmup_ratio=0.9): + # validate the "warmup" argument + if warmup is not None: + if warmup not in ['constant', 'linear', 'exp']: + raise ValueError( + f'"{warmup}" is not a supported type for warming up, valid' + ' types are "constant" and "linear"') + if warmup is not None: + assert warmup_iters > 0, \ + '"warmup_iters" must be a positive integer' + assert 0 < warmup_ratio <= 1.0, \ + '"warmup_momentum" must be in range (0,1]' + + self.by_epoch = by_epoch + self.warmup = warmup + self.warmup_iters = warmup_iters + self.warmup_ratio = warmup_ratio + + self.base_momentum = [] # initial momentum for all param groups + self.regular_momentum = [ + ] # expected momentum if no warming up is performed + + def _set_momentum(self, runner, momentum_groups): + if isinstance(runner.optimizer, dict): + for k, optim in runner.optimizer.items(): + for param_group, mom in zip(optim.param_groups, + momentum_groups[k]): + if 'momentum' in param_group.keys(): + param_group['momentum'] = mom + elif 'betas' in param_group.keys(): + param_group['betas'] = (mom, param_group['betas'][1]) + else: + for param_group, mom in zip(runner.optimizer.param_groups, + momentum_groups): + if 'momentum' in param_group.keys(): + param_group['momentum'] = mom + elif 'betas' in param_group.keys(): + param_group['betas'] = (mom, param_group['betas'][1]) + + def get_momentum(self, runner, base_momentum): + raise NotImplementedError + + def get_regular_momentum(self, runner): + if isinstance(runner.optimizer, dict): + momentum_groups = {} + for k in runner.optimizer.keys(): + _momentum_group = [ + self.get_momentum(runner, _base_momentum) + for _base_momentum in self.base_momentum[k] + ] + momentum_groups.update({k: _momentum_group}) + return momentum_groups + else: + return [ + self.get_momentum(runner, _base_momentum) + for _base_momentum in self.base_momentum + ] + + def get_warmup_momentum(self, cur_iters): + + def _get_warmup_momentum(cur_iters, regular_momentum): + if self.warmup == 'constant': + warmup_momentum = [ + _momentum / self.warmup_ratio + for _momentum in regular_momentum + ] + elif self.warmup == 'linear': + k = (1 - cur_iters / self.warmup_iters) * (1 - + self.warmup_ratio) + warmup_momentum = [ + _momentum / (1 - k) for _momentum in regular_momentum + ] + elif self.warmup == 'exp': + k = self.warmup_ratio**(1 - cur_iters / self.warmup_iters) + warmup_momentum = [ + _momentum / k for _momentum in regular_momentum + ] + return warmup_momentum + + if isinstance(self.regular_momentum, dict): + momentum_groups = {} + for key, regular_momentum in self.regular_momentum.items(): + momentum_groups[key] = _get_warmup_momentum( + cur_iters, regular_momentum) + return momentum_groups + else: + return _get_warmup_momentum(cur_iters, self.regular_momentum) + + def before_run(self, runner): + # NOTE: when resuming from a checkpoint, + # if 'initial_momentum' is not saved, + # it will be set according to the optimizer params + if isinstance(runner.optimizer, dict): + self.base_momentum = {} + for k, optim in runner.optimizer.items(): + for group in optim.param_groups: + if 'momentum' in group.keys(): + group.setdefault('initial_momentum', group['momentum']) + else: + group.setdefault('initial_momentum', group['betas'][0]) + _base_momentum = [ + group['initial_momentum'] for group in optim.param_groups + ] + self.base_momentum.update({k: _base_momentum}) + else: + for group in runner.optimizer.param_groups: + if 'momentum' in group.keys(): + group.setdefault('initial_momentum', group['momentum']) + else: + group.setdefault('initial_momentum', group['betas'][0]) + self.base_momentum = [ + group['initial_momentum'] + for group in runner.optimizer.param_groups + ] + + def before_train_epoch(self, runner): + if not self.by_epoch: + return + self.regular_momentum = self.get_regular_momentum(runner) + self._set_momentum(runner, self.regular_momentum) + + def before_train_iter(self, runner): + cur_iter = runner.iter + if not self.by_epoch: + self.regular_momentum = self.get_regular_momentum(runner) + if self.warmup is None or cur_iter >= self.warmup_iters: + self._set_momentum(runner, self.regular_momentum) + else: + warmup_momentum = self.get_warmup_momentum(cur_iter) + self._set_momentum(runner, warmup_momentum) + elif self.by_epoch: + if self.warmup is None or cur_iter > self.warmup_iters: + return + elif cur_iter == self.warmup_iters: + self._set_momentum(runner, self.regular_momentum) + else: + warmup_momentum = self.get_warmup_momentum(cur_iter) + self._set_momentum(runner, warmup_momentum) + + +@HOOKS.register_module() +class StepMomentumUpdaterHook(MomentumUpdaterHook): + """Step momentum scheduler with min value clipping. + + Args: + step (int | list[int]): Step to decay the momentum. If an int value is + given, regard it as the decay interval. If a list is given, decay + momentum at these steps. + gamma (float, optional): Decay momentum ratio. Default: 0.5. + min_momentum (float, optional): Minimum momentum value to keep. If + momentum after decay is lower than this value, it will be clipped + accordingly. If None is given, we don't perform lr clipping. + Default: None. + """ + + def __init__(self, step, gamma=0.5, min_momentum=None, **kwargs): + if isinstance(step, list): + assert mmcv.is_list_of(step, int) + assert all([s > 0 for s in step]) + elif isinstance(step, int): + assert step > 0 + else: + raise TypeError('"step" must be a list or integer') + self.step = step + self.gamma = gamma + self.min_momentum = min_momentum + super(StepMomentumUpdaterHook, self).__init__(**kwargs) + + def get_momentum(self, runner, base_momentum): + progress = runner.epoch if self.by_epoch else runner.iter + + # calculate exponential term + if isinstance(self.step, int): + exp = progress // self.step + else: + exp = len(self.step) + for i, s in enumerate(self.step): + if progress < s: + exp = i + break + + momentum = base_momentum * (self.gamma**exp) + if self.min_momentum is not None: + # clip to a minimum value + momentum = max(momentum, self.min_momentum) + return momentum + + +@HOOKS.register_module() +class CosineAnnealingMomentumUpdaterHook(MomentumUpdaterHook): + """Cosine annealing LR Momentum decays the Momentum of each parameter group + linearly. + + Args: + min_momentum (float, optional): The minimum momentum. Default: None. + min_momentum_ratio (float, optional): The ratio of minimum momentum to + the base momentum. Either `min_momentum` or `min_momentum_ratio` + should be specified. Default: None. + """ + + def __init__(self, min_momentum=None, min_momentum_ratio=None, **kwargs): + assert (min_momentum is None) ^ (min_momentum_ratio is None) + self.min_momentum = min_momentum + self.min_momentum_ratio = min_momentum_ratio + super(CosineAnnealingMomentumUpdaterHook, self).__init__(**kwargs) + + def get_momentum(self, runner, base_momentum): + if self.by_epoch: + progress = runner.epoch + max_progress = runner.max_epochs + else: + progress = runner.iter + max_progress = runner.max_iters + if self.min_momentum_ratio is not None: + target_momentum = base_momentum * self.min_momentum_ratio + else: + target_momentum = self.min_momentum + return annealing_cos(base_momentum, target_momentum, + progress / max_progress) + + +@HOOKS.register_module() +class LinearAnnealingMomentumUpdaterHook(MomentumUpdaterHook): + """Linear annealing LR Momentum decays the Momentum of each parameter group + linearly. + + Args: + min_momentum (float, optional): The minimum momentum. Default: None. + min_momentum_ratio (float, optional): The ratio of minimum momentum to + the base momentum. Either `min_momentum` or `min_momentum_ratio` + should be specified. Default: None. + """ + + def __init__(self, min_momentum=None, min_momentum_ratio=None, **kwargs): + assert (min_momentum is None) ^ (min_momentum_ratio is None) + self.min_momentum = min_momentum + self.min_momentum_ratio = min_momentum_ratio + super(LinearAnnealingMomentumUpdaterHook, self).__init__(**kwargs) + + def get_momentum(self, runner, base_momentum): + if self.by_epoch: + progress = runner.epoch + max_progress = runner.max_epochs + else: + progress = runner.iter + max_progress = runner.max_iters + if self.min_momentum_ratio is not None: + target_momentum = base_momentum * self.min_momentum_ratio + else: + target_momentum = self.min_momentum + return annealing_linear(base_momentum, target_momentum, + progress / max_progress) + + +@HOOKS.register_module() +class CyclicMomentumUpdaterHook(MomentumUpdaterHook): + """Cyclic momentum Scheduler. + + Implement the cyclical momentum scheduler policy described in + https://arxiv.org/pdf/1708.07120.pdf + + This momentum scheduler usually used together with the CyclicLRUpdater + to improve the performance in the 3D detection area. + + Args: + target_ratio (tuple[float]): Relative ratio of the lowest momentum and + the highest momentum to the initial momentum. + cyclic_times (int): Number of cycles during training + step_ratio_up (float): The ratio of the increasing process of momentum + in the total cycle. + by_epoch (bool): Whether to update momentum by epoch. + anneal_strategy (str, optional): {'cos', 'linear'} + Specifies the annealing strategy: 'cos' for cosine annealing, + 'linear' for linear annealing. Default: 'cos'. + gamma (float, optional): Cycle decay ratio. Default: 1. + It takes values in the range (0, 1]. The difference between the + maximum learning rate and the minimum learning rate decreases + periodically when it is less than 1. `New in version 1.4.4.` + """ + + def __init__(self, + by_epoch=False, + target_ratio=(0.85 / 0.95, 1), + cyclic_times=1, + step_ratio_up=0.4, + anneal_strategy='cos', + gamma=1, + **kwargs): + if isinstance(target_ratio, float): + target_ratio = (target_ratio, target_ratio / 1e5) + elif isinstance(target_ratio, tuple): + target_ratio = (target_ratio[0], target_ratio[0] / 1e5) \ + if len(target_ratio) == 1 else target_ratio + else: + raise ValueError('target_ratio should be either float ' + f'or tuple, got {type(target_ratio)}') + + assert len(target_ratio) == 2, \ + '"target_ratio" must be list or tuple of two floats' + assert 0 <= step_ratio_up < 1.0, \ + '"step_ratio_up" must be in range [0,1)' + + self.target_ratio = target_ratio + self.cyclic_times = cyclic_times + self.step_ratio_up = step_ratio_up + self.gamma = gamma + self.momentum_phases = [] # init momentum_phases + + if anneal_strategy not in ['cos', 'linear']: + raise ValueError('anneal_strategy must be one of "cos" or ' + f'"linear", instead got {anneal_strategy}') + elif anneal_strategy == 'cos': + self.anneal_func = annealing_cos + elif anneal_strategy == 'linear': + self.anneal_func = annealing_linear + # currently only support by_epoch=False + assert not by_epoch, \ + 'currently only support "by_epoch" = False' + super(CyclicMomentumUpdaterHook, self).__init__(by_epoch, **kwargs) + + def before_run(self, runner): + super(CyclicMomentumUpdaterHook, self).before_run(runner) + # initiate momentum_phases + # total momentum_phases are separated as up and down + max_iter_per_phase = runner.max_iters // self.cyclic_times + iter_up_phase = int(self.step_ratio_up * max_iter_per_phase) + self.max_iter_per_phase = max_iter_per_phase + self.momentum_phases.append( + [0, iter_up_phase, 1, self.target_ratio[0]]) + self.momentum_phases.append([ + iter_up_phase, max_iter_per_phase, self.target_ratio[0], + self.target_ratio[1] + ]) + + def get_momentum(self, runner, base_momentum): + curr_iter = runner.iter % self.max_iter_per_phase + curr_cycle = runner.iter // self.max_iter_per_phase + scale = self.gamma**curr_cycle + for (start_iter, end_iter, start_ratio, end_ratio) \ + in self.momentum_phases: + if start_iter <= curr_iter < end_iter: + # Apply cycle scaling to gradually reduce the difference + # between max_momentum and base momentum. The target end_ratio + # can be expressed as: + # end_ratio = (base_momentum + scale * \ + # (max_momentum - base_momentum)) / base_momentum + # iteration: 0-iter_up_phase: + if start_iter == 0: + end_ratio = 1 - scale + end_ratio * scale + # iteration: iter_up_phase-self.max_iter_per_phase + else: + start_ratio = 1 - scale + start_ratio * scale + progress = curr_iter - start_iter + return self.anneal_func(base_momentum * start_ratio, + base_momentum * end_ratio, + progress / (end_iter - start_iter)) + + +@HOOKS.register_module() +class OneCycleMomentumUpdaterHook(MomentumUpdaterHook): + """OneCycle momentum Scheduler. + + This momentum scheduler usually used together with the OneCycleLrUpdater + to improve the performance. + + Args: + base_momentum (float or list): Lower momentum boundaries in the cycle + for each parameter group. Note that momentum is cycled inversely + to learning rate; at the peak of a cycle, momentum is + 'base_momentum' and learning rate is 'max_lr'. + Default: 0.85 + max_momentum (float or list): Upper momentum boundaries in the cycle + for each parameter group. Functionally, + it defines the cycle amplitude (max_momentum - base_momentum). + Note that momentum is cycled inversely + to learning rate; at the start of a cycle, momentum is + 'max_momentum' and learning rate is 'base_lr' + Default: 0.95 + pct_start (float): The percentage of the cycle (in number of steps) + spent increasing the learning rate. + Default: 0.3 + anneal_strategy (str): {'cos', 'linear'} + Specifies the annealing strategy: 'cos' for cosine annealing, + 'linear' for linear annealing. + Default: 'cos' + three_phase (bool): If three_phase is True, use a third phase of the + schedule to annihilate the learning rate according to + final_div_factor instead of modifying the second phase (the first + two phases will be symmetrical about the step indicated by + pct_start). + Default: False + """ + + def __init__(self, + base_momentum=0.85, + max_momentum=0.95, + pct_start=0.3, + anneal_strategy='cos', + three_phase=False, + **kwargs): + # validate by_epoch, currently only support by_epoch=False + if 'by_epoch' not in kwargs: + kwargs['by_epoch'] = False + else: + assert not kwargs['by_epoch'], \ + 'currently only support "by_epoch" = False' + if not isinstance(base_momentum, (float, list, dict)): + raise ValueError('base_momentum must be the type among of float,' + 'list or dict.') + self._base_momentum = base_momentum + if not isinstance(max_momentum, (float, list, dict)): + raise ValueError('max_momentum must be the type among of float,' + 'list or dict.') + self._max_momentum = max_momentum + # validate pct_start + if pct_start < 0 or pct_start > 1 or not isinstance(pct_start, float): + raise ValueError('Expected float between 0 and 1 pct_start, but ' + f'got {pct_start}') + self.pct_start = pct_start + # validate anneal_strategy + if anneal_strategy not in ['cos', 'linear']: + raise ValueError('anneal_strategy must by one of "cos" or ' + f'"linear", instead got {anneal_strategy}') + elif anneal_strategy == 'cos': + self.anneal_func = annealing_cos + elif anneal_strategy == 'linear': + self.anneal_func = annealing_linear + self.three_phase = three_phase + self.momentum_phases = [] # init momentum_phases + super(OneCycleMomentumUpdaterHook, self).__init__(**kwargs) + + def before_run(self, runner): + if isinstance(runner.optimizer, dict): + for k, optim in runner.optimizer.items(): + if ('momentum' not in optim.defaults + and 'betas' not in optim.defaults): + raise ValueError('optimizer must support momentum with' + 'option enabled') + self.use_beta1 = 'betas' in optim.defaults + _base_momentum = format_param(k, optim, self._base_momentum) + _max_momentum = format_param(k, optim, self._max_momentum) + for group, b_momentum, m_momentum in zip( + optim.param_groups, _base_momentum, _max_momentum): + if self.use_beta1: + _, beta2 = group['betas'] + group['betas'] = (m_momentum, beta2) + else: + group['momentum'] = m_momentum + group['base_momentum'] = b_momentum + group['max_momentum'] = m_momentum + else: + optim = runner.optimizer + if ('momentum' not in optim.defaults + and 'betas' not in optim.defaults): + raise ValueError('optimizer must support momentum with' + 'option enabled') + self.use_beta1 = 'betas' in optim.defaults + k = type(optim).__name__ + _base_momentum = format_param(k, optim, self._base_momentum) + _max_momentum = format_param(k, optim, self._max_momentum) + for group, b_momentum, m_momentum in zip(optim.param_groups, + _base_momentum, + _max_momentum): + if self.use_beta1: + _, beta2 = group['betas'] + group['betas'] = (m_momentum, beta2) + else: + group['momentum'] = m_momentum + group['base_momentum'] = b_momentum + group['max_momentum'] = m_momentum + + if self.three_phase: + self.momentum_phases.append({ + 'end_iter': + float(self.pct_start * runner.max_iters) - 1, + 'start_momentum': + 'max_momentum', + 'end_momentum': + 'base_momentum' + }) + self.momentum_phases.append({ + 'end_iter': + float(2 * self.pct_start * runner.max_iters) - 2, + 'start_momentum': + 'base_momentum', + 'end_momentum': + 'max_momentum' + }) + self.momentum_phases.append({ + 'end_iter': runner.max_iters - 1, + 'start_momentum': 'max_momentum', + 'end_momentum': 'max_momentum' + }) + else: + self.momentum_phases.append({ + 'end_iter': + float(self.pct_start * runner.max_iters) - 1, + 'start_momentum': + 'max_momentum', + 'end_momentum': + 'base_momentum' + }) + self.momentum_phases.append({ + 'end_iter': runner.max_iters - 1, + 'start_momentum': 'base_momentum', + 'end_momentum': 'max_momentum' + }) + + def _set_momentum(self, runner, momentum_groups): + if isinstance(runner.optimizer, dict): + for k, optim in runner.optimizer.items(): + for param_group, mom in zip(optim.param_groups, + momentum_groups[k]): + if 'momentum' in param_group.keys(): + param_group['momentum'] = mom + elif 'betas' in param_group.keys(): + param_group['betas'] = (mom, param_group['betas'][1]) + else: + for param_group, mom in zip(runner.optimizer.param_groups, + momentum_groups): + if 'momentum' in param_group.keys(): + param_group['momentum'] = mom + elif 'betas' in param_group.keys(): + param_group['betas'] = (mom, param_group['betas'][1]) + + def get_momentum(self, runner, param_group): + curr_iter = runner.iter + start_iter = 0 + for i, phase in enumerate(self.momentum_phases): + end_iter = phase['end_iter'] + if curr_iter <= end_iter or i == len(self.momentum_phases) - 1: + pct = (curr_iter - start_iter) / (end_iter - start_iter) + momentum = self.anneal_func( + param_group[phase['start_momentum']], + param_group[phase['end_momentum']], pct) + break + start_iter = end_iter + return momentum + + def get_regular_momentum(self, runner): + if isinstance(runner.optimizer, dict): + momentum_groups = {} + for k, optim in runner.optimizer.items(): + _momentum_group = [ + self.get_momentum(runner, param_group) + for param_group in optim.param_groups + ] + momentum_groups.update({k: _momentum_group}) + return momentum_groups + else: + momentum_groups = [] + for param_group in runner.optimizer.param_groups: + momentum_groups.append(self.get_momentum(runner, param_group)) + return momentum_groups diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/optimizer.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/optimizer.py new file mode 100644 index 0000000000000000000000000000000000000000..12c8851834911fc9802b314a4fb0ed0715eeee41 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/optimizer.py @@ -0,0 +1,556 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import logging +from collections import defaultdict +from itertools import chain + +from torch.nn.utils import clip_grad + +from mmcv.utils import TORCH_VERSION, _BatchNorm, digit_version +from ..dist_utils import allreduce_grads +from ..fp16_utils import LossScaler, wrap_fp16_model +from .hook import HOOKS, Hook + +try: + # If PyTorch version >= 1.6.0, torch.cuda.amp.GradScaler would be imported + # and used; otherwise, auto fp16 will adopt mmcv's implementation. + from torch.cuda.amp import GradScaler +except ImportError: + pass + + +@HOOKS.register_module() +class OptimizerHook(Hook): + """A hook contains custom operations for the optimizer. + + Args: + grad_clip (dict, optional): A config dict to control the clip_grad. + Default: None. + detect_anomalous_params (bool): This option is only used for + debugging which will slow down the training speed. + Detect anomalous parameters that are not included in + the computational graph with `loss` as the root. + There are two cases + + - Parameters were not used during + forward pass. + - Parameters were not used to produce + loss. + Default: False. + """ + + def __init__(self, grad_clip=None, detect_anomalous_params=False): + self.grad_clip = grad_clip + self.detect_anomalous_params = detect_anomalous_params + + def clip_grads(self, params): + params = list( + filter(lambda p: p.requires_grad and p.grad is not None, params)) + if len(params) > 0: + return clip_grad.clip_grad_norm_(params, **self.grad_clip) + + def after_train_iter(self, runner): + runner.optimizer.zero_grad() + if self.detect_anomalous_params: + self.detect_anomalous_parameters(runner.outputs['loss'], runner) + runner.outputs['loss'].backward() + + if self.grad_clip is not None: + grad_norm = self.clip_grads(runner.model.parameters()) + if grad_norm is not None: + # Add grad norm to the logger + runner.log_buffer.update({'grad_norm': float(grad_norm)}, + runner.outputs['num_samples']) + runner.optimizer.step() + + def detect_anomalous_parameters(self, loss, runner): + logger = runner.logger + parameters_in_graph = set() + visited = set() + + def traverse(grad_fn): + if grad_fn is None: + return + if grad_fn not in visited: + visited.add(grad_fn) + if hasattr(grad_fn, 'variable'): + parameters_in_graph.add(grad_fn.variable) + parents = grad_fn.next_functions + if parents is not None: + for parent in parents: + grad_fn = parent[0] + traverse(grad_fn) + + traverse(loss.grad_fn) + for n, p in runner.model.named_parameters(): + if p not in parameters_in_graph and p.requires_grad: + logger.log( + level=logging.ERROR, + msg=f'{n} with shape {p.size()} is not ' + f'in the computational graph \n') + + +@HOOKS.register_module() +class GradientCumulativeOptimizerHook(OptimizerHook): + """Optimizer Hook implements multi-iters gradient cumulating. + + Args: + cumulative_iters (int, optional): Num of gradient cumulative iters. + The optimizer will step every `cumulative_iters` iters. + Defaults to 1. + + Examples: + >>> # Use cumulative_iters to simulate a large batch size + >>> # It is helpful when the hardware cannot handle a large batch size. + >>> loader = DataLoader(data, batch_size=64) + >>> optim_hook = GradientCumulativeOptimizerHook(cumulative_iters=4) + >>> # almost equals to + >>> loader = DataLoader(data, batch_size=256) + >>> optim_hook = OptimizerHook() + """ + + def __init__(self, cumulative_iters=1, **kwargs): + super(GradientCumulativeOptimizerHook, self).__init__(**kwargs) + + assert isinstance(cumulative_iters, int) and cumulative_iters > 0, \ + f'cumulative_iters only accepts positive int, but got ' \ + f'{type(cumulative_iters)} instead.' + + self.cumulative_iters = cumulative_iters + self.divisible_iters = 0 + self.remainder_iters = 0 + self.initialized = False + + def has_batch_norm(self, module): + if isinstance(module, _BatchNorm): + return True + for m in module.children(): + if self.has_batch_norm(m): + return True + return False + + def _init(self, runner): + if runner.iter % self.cumulative_iters != 0: + runner.logger.warning( + 'Resume iter number is not divisible by cumulative_iters in ' + 'GradientCumulativeOptimizerHook, which means the gradient of ' + 'some iters is lost and the result may be influenced slightly.' + ) + + if self.has_batch_norm(runner.model) and self.cumulative_iters > 1: + runner.logger.warning( + 'GradientCumulativeOptimizerHook may slightly decrease ' + 'performance if the model has BatchNorm layers.') + + residual_iters = runner.max_iters - runner.iter + + self.divisible_iters = ( + residual_iters // self.cumulative_iters * self.cumulative_iters) + self.remainder_iters = residual_iters - self.divisible_iters + + self.initialized = True + + def after_train_iter(self, runner): + if not self.initialized: + self._init(runner) + + if runner.iter < self.divisible_iters: + loss_factor = self.cumulative_iters + else: + loss_factor = self.remainder_iters + loss = runner.outputs['loss'] + loss = loss / loss_factor + loss.backward() + + if (self.every_n_iters(runner, self.cumulative_iters) + or self.is_last_iter(runner)): + + if self.grad_clip is not None: + grad_norm = self.clip_grads(runner.model.parameters()) + if grad_norm is not None: + # Add grad norm to the logger + runner.log_buffer.update({'grad_norm': float(grad_norm)}, + runner.outputs['num_samples']) + runner.optimizer.step() + runner.optimizer.zero_grad() + + +if (TORCH_VERSION != 'parrots' + and digit_version(TORCH_VERSION) >= digit_version('1.6.0')): + + @HOOKS.register_module() + class Fp16OptimizerHook(OptimizerHook): + """FP16 optimizer hook (using PyTorch's implementation). + + If you are using PyTorch >= 1.6, torch.cuda.amp is used as the backend, + to take care of the optimization procedure. + + Args: + loss_scale (float | str | dict): Scale factor configuration. + If loss_scale is a float, static loss scaling will be used with + the specified scale. If loss_scale is a string, it must be + 'dynamic', then dynamic loss scaling will be used. + It can also be a dict containing arguments of GradScalar. + Defaults to 512. For Pytorch >= 1.6, mmcv uses official + implementation of GradScaler. If you use a dict version of + loss_scale to create GradScaler, please refer to: + https://pytorch.org/docs/stable/amp.html#torch.cuda.amp.GradScaler + for the parameters. + + Examples: + >>> loss_scale = dict( + ... init_scale=65536.0, + ... growth_factor=2.0, + ... backoff_factor=0.5, + ... growth_interval=2000 + ... ) + >>> optimizer_hook = Fp16OptimizerHook(loss_scale=loss_scale) + """ + + def __init__(self, + grad_clip=None, + coalesce=True, + bucket_size_mb=-1, + loss_scale=512., + distributed=True): + self.grad_clip = grad_clip + self.coalesce = coalesce + self.bucket_size_mb = bucket_size_mb + self.distributed = distributed + self._scale_update_param = None + if loss_scale == 'dynamic': + self.loss_scaler = GradScaler() + elif isinstance(loss_scale, float): + self._scale_update_param = loss_scale + self.loss_scaler = GradScaler(init_scale=loss_scale) + elif isinstance(loss_scale, dict): + self.loss_scaler = GradScaler(**loss_scale) + else: + raise ValueError('loss_scale must be of type float, dict, or ' + f'"dynamic", got {loss_scale}') + + def before_run(self, runner): + """Preparing steps before Mixed Precision Training.""" + # wrap model mode to fp16 + wrap_fp16_model(runner.model) + # resume from state dict + if 'fp16' in runner.meta and 'loss_scaler' in runner.meta['fp16']: + scaler_state_dict = runner.meta['fp16']['loss_scaler'] + self.loss_scaler.load_state_dict(scaler_state_dict) + + def copy_grads_to_fp32(self, fp16_net, fp32_weights): + """Copy gradients from fp16 model to fp32 weight copy.""" + for fp32_param, fp16_param in zip(fp32_weights, + fp16_net.parameters()): + if fp16_param.grad is not None: + if fp32_param.grad is None: + fp32_param.grad = fp32_param.data.new( + fp32_param.size()) + fp32_param.grad.copy_(fp16_param.grad) + + def copy_params_to_fp16(self, fp16_net, fp32_weights): + """Copy updated params from fp32 weight copy to fp16 model.""" + for fp16_param, fp32_param in zip(fp16_net.parameters(), + fp32_weights): + fp16_param.data.copy_(fp32_param.data) + + def after_train_iter(self, runner): + """Backward optimization steps for Mixed Precision Training. For + dynamic loss scaling, please refer to + https://pytorch.org/docs/stable/amp.html#torch.cuda.amp.GradScaler. + + 1. Scale the loss by a scale factor. + 2. Backward the loss to obtain the gradients. + 3. Unscale the optimizer’s gradient tensors. + 4. Call optimizer.step() and update scale factor. + 5. Save loss_scaler state_dict for resume purpose. + """ + # clear grads of last iteration + runner.model.zero_grad() + runner.optimizer.zero_grad() + + self.loss_scaler.scale(runner.outputs['loss']).backward() + self.loss_scaler.unscale_(runner.optimizer) + # grad clip + if self.grad_clip is not None: + grad_norm = self.clip_grads(runner.model.parameters()) + if grad_norm is not None: + # Add grad norm to the logger + runner.log_buffer.update({'grad_norm': float(grad_norm)}, + runner.outputs['num_samples']) + # backward and update scaler + self.loss_scaler.step(runner.optimizer) + self.loss_scaler.update(self._scale_update_param) + + # save state_dict of loss_scaler + runner.meta.setdefault( + 'fp16', {})['loss_scaler'] = self.loss_scaler.state_dict() + + @HOOKS.register_module() + class GradientCumulativeFp16OptimizerHook(GradientCumulativeOptimizerHook, + Fp16OptimizerHook): + """Fp16 optimizer Hook (using PyTorch's implementation) implements + multi-iters gradient cumulating. + + If you are using PyTorch >= 1.6, torch.cuda.amp is used as the backend, + to take care of the optimization procedure. + """ + + def __init__(self, *args, **kwargs): + super(GradientCumulativeFp16OptimizerHook, + self).__init__(*args, **kwargs) + + def after_train_iter(self, runner): + if not self.initialized: + self._init(runner) + + if runner.iter < self.divisible_iters: + loss_factor = self.cumulative_iters + else: + loss_factor = self.remainder_iters + loss = runner.outputs['loss'] + loss = loss / loss_factor + + self.loss_scaler.scale(loss).backward() + + if (self.every_n_iters(runner, self.cumulative_iters) + or self.is_last_iter(runner)): + + # copy fp16 grads in the model to fp32 params in the optimizer + self.loss_scaler.unscale_(runner.optimizer) + + if self.grad_clip is not None: + grad_norm = self.clip_grads(runner.model.parameters()) + if grad_norm is not None: + # Add grad norm to the logger + runner.log_buffer.update( + {'grad_norm': float(grad_norm)}, + runner.outputs['num_samples']) + + # backward and update scaler + self.loss_scaler.step(runner.optimizer) + self.loss_scaler.update(self._scale_update_param) + + # save state_dict of loss_scaler + runner.meta.setdefault( + 'fp16', {})['loss_scaler'] = self.loss_scaler.state_dict() + + # clear grads + runner.model.zero_grad() + runner.optimizer.zero_grad() + +else: + + @HOOKS.register_module() + class Fp16OptimizerHook(OptimizerHook): + """FP16 optimizer hook (mmcv's implementation). + + The steps of fp16 optimizer is as follows. + 1. Scale the loss value. + 2. BP in the fp16 model. + 2. Copy gradients from fp16 model to fp32 weights. + 3. Update fp32 weights. + 4. Copy updated parameters from fp32 weights to fp16 model. + + Refer to https://arxiv.org/abs/1710.03740 for more details. + + Args: + loss_scale (float | str | dict): Scale factor configuration. + If loss_scale is a float, static loss scaling will be used with + the specified scale. If loss_scale is a string, it must be + 'dynamic', then dynamic loss scaling will be used. + It can also be a dict containing arguments of LossScaler. + Defaults to 512. + """ + + def __init__(self, + grad_clip=None, + coalesce=True, + bucket_size_mb=-1, + loss_scale=512., + distributed=True): + self.grad_clip = grad_clip + self.coalesce = coalesce + self.bucket_size_mb = bucket_size_mb + self.distributed = distributed + if loss_scale == 'dynamic': + self.loss_scaler = LossScaler(mode='dynamic') + elif isinstance(loss_scale, float): + self.loss_scaler = LossScaler( + init_scale=loss_scale, mode='static') + elif isinstance(loss_scale, dict): + self.loss_scaler = LossScaler(**loss_scale) + else: + raise ValueError('loss_scale must be of type float, dict, or ' + f'"dynamic", got {loss_scale}') + + def before_run(self, runner): + """Preparing steps before Mixed Precision Training. + + 1. Make a master copy of fp32 weights for optimization. + 2. Convert the main model from fp32 to fp16. + """ + # keep a copy of fp32 weights + old_groups = runner.optimizer.param_groups + runner.optimizer.param_groups = copy.deepcopy( + runner.optimizer.param_groups) + state = defaultdict(dict) + p_map = { + old_p: p + for old_p, p in zip( + chain(*(g['params'] for g in old_groups)), + chain(*(g['params'] + for g in runner.optimizer.param_groups))) + } + for k, v in runner.optimizer.state.items(): + state[p_map[k]] = v + runner.optimizer.state = state + # convert model to fp16 + wrap_fp16_model(runner.model) + # resume from state dict + if 'fp16' in runner.meta and 'loss_scaler' in runner.meta['fp16']: + scaler_state_dict = runner.meta['fp16']['loss_scaler'] + self.loss_scaler.load_state_dict(scaler_state_dict) + + def copy_grads_to_fp32(self, fp16_net, fp32_weights): + """Copy gradients from fp16 model to fp32 weight copy.""" + for fp32_param, fp16_param in zip(fp32_weights, + fp16_net.parameters()): + if fp16_param.grad is not None: + if fp32_param.grad is None: + fp32_param.grad = fp32_param.data.new( + fp32_param.size()) + fp32_param.grad.copy_(fp16_param.grad) + + def copy_params_to_fp16(self, fp16_net, fp32_weights): + """Copy updated params from fp32 weight copy to fp16 model.""" + for fp16_param, fp32_param in zip(fp16_net.parameters(), + fp32_weights): + fp16_param.data.copy_(fp32_param.data) + + def after_train_iter(self, runner): + """Backward optimization steps for Mixed Precision Training. For + dynamic loss scaling, please refer `loss_scalar.py` + + 1. Scale the loss by a scale factor. + 2. Backward the loss to obtain the gradients (fp16). + 3. Copy gradients from the model to the fp32 weight copy. + 4. Scale the gradients back and update the fp32 weight copy. + 5. Copy back the params from fp32 weight copy to the fp16 model. + 6. Save loss_scaler state_dict for resume purpose. + """ + # clear grads of last iteration + runner.model.zero_grad() + runner.optimizer.zero_grad() + # scale the loss value + scaled_loss = runner.outputs['loss'] * self.loss_scaler.loss_scale + scaled_loss.backward() + # copy fp16 grads in the model to fp32 params in the optimizer + + fp32_weights = [] + for param_group in runner.optimizer.param_groups: + fp32_weights += param_group['params'] + self.copy_grads_to_fp32(runner.model, fp32_weights) + # allreduce grads + if self.distributed: + allreduce_grads(fp32_weights, self.coalesce, + self.bucket_size_mb) + + has_overflow = self.loss_scaler.has_overflow(fp32_weights) + # if has overflow, skip this iteration + if not has_overflow: + # scale the gradients back + for param in fp32_weights: + if param.grad is not None: + param.grad.div_(self.loss_scaler.loss_scale) + if self.grad_clip is not None: + grad_norm = self.clip_grads(fp32_weights) + if grad_norm is not None: + # Add grad norm to the logger + runner.log_buffer.update( + {'grad_norm': float(grad_norm)}, + runner.outputs['num_samples']) + # update fp32 params + runner.optimizer.step() + # copy fp32 params to the fp16 model + self.copy_params_to_fp16(runner.model, fp32_weights) + self.loss_scaler.update_scale(has_overflow) + if has_overflow: + runner.logger.warning('Check overflow, downscale loss scale ' + f'to {self.loss_scaler.cur_scale}') + + # save state_dict of loss_scaler + runner.meta.setdefault( + 'fp16', {})['loss_scaler'] = self.loss_scaler.state_dict() + + @HOOKS.register_module() + class GradientCumulativeFp16OptimizerHook(GradientCumulativeOptimizerHook, + Fp16OptimizerHook): + """Fp16 optimizer Hook (using mmcv implementation) implements multi- + iters gradient cumulating.""" + + def __init__(self, *args, **kwargs): + super(GradientCumulativeFp16OptimizerHook, + self).__init__(*args, **kwargs) + + def after_train_iter(self, runner): + if not self.initialized: + self._init(runner) + + if runner.iter < self.divisible_iters: + loss_factor = self.cumulative_iters + else: + loss_factor = self.remainder_iters + + loss = runner.outputs['loss'] + loss = loss / loss_factor + + # scale the loss value + scaled_loss = loss * self.loss_scaler.loss_scale + scaled_loss.backward() + + if (self.every_n_iters(runner, self.cumulative_iters) + or self.is_last_iter(runner)): + + # copy fp16 grads in the model to fp32 params in the optimizer + fp32_weights = [] + for param_group in runner.optimizer.param_groups: + fp32_weights += param_group['params'] + self.copy_grads_to_fp32(runner.model, fp32_weights) + # allreduce grads + if self.distributed: + allreduce_grads(fp32_weights, self.coalesce, + self.bucket_size_mb) + + has_overflow = self.loss_scaler.has_overflow(fp32_weights) + # if has overflow, skip this iteration + if not has_overflow: + # scale the gradients back + for param in fp32_weights: + if param.grad is not None: + param.grad.div_(self.loss_scaler.loss_scale) + if self.grad_clip is not None: + grad_norm = self.clip_grads(fp32_weights) + if grad_norm is not None: + # Add grad norm to the logger + runner.log_buffer.update( + {'grad_norm': float(grad_norm)}, + runner.outputs['num_samples']) + # update fp32 params + runner.optimizer.step() + # copy fp32 params to the fp16 model + self.copy_params_to_fp16(runner.model, fp32_weights) + else: + runner.logger.warning( + 'Check overflow, downscale loss scale ' + f'to {self.loss_scaler.cur_scale}') + + self.loss_scaler.update_scale(has_overflow) + + # save state_dict of loss_scaler + runner.meta.setdefault( + 'fp16', {})['loss_scaler'] = self.loss_scaler.state_dict() + + # clear grads + runner.model.zero_grad() + runner.optimizer.zero_grad() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/profiler.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/profiler.py new file mode 100644 index 0000000000000000000000000000000000000000..fef9adc13927df463c7ad37bc4c2f0b89c90a1f1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/profiler.py @@ -0,0 +1,180 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings +from typing import Callable, List, Optional, Union + +import torch + +from ..dist_utils import master_only +from .hook import HOOKS, Hook + + +@HOOKS.register_module() +class ProfilerHook(Hook): + """Profiler to analyze performance during training. + + PyTorch Profiler is a tool that allows the collection of the performance + metrics during the training. More details on Profiler can be found at + https://pytorch.org/docs/1.8.1/profiler.html#torch.profiler.profile + + Args: + by_epoch (bool): Profile performance by epoch or by iteration. + Default: True. + profile_iters (int): Number of iterations for profiling. + If ``by_epoch=True``, profile_iters indicates that they are the + first profile_iters epochs at the beginning of the + training, otherwise it indicates the first profile_iters + iterations. Default: 1. + activities (list[str]): List of activity groups (CPU, CUDA) to use in + profiling. Default: ['cpu', 'cuda']. + schedule (dict, optional): Config of generating the callable schedule. + if schedule is None, profiler will not add step markers into the + trace and table view. Default: None. + on_trace_ready (callable, dict): Either a handler or a dict of generate + handler. Default: None. + record_shapes (bool): Save information about operator's input shapes. + Default: False. + profile_memory (bool): Track tensor memory allocation/deallocation. + Default: False. + with_stack (bool): Record source information (file and line number) + for the ops. Default: False. + with_flops (bool): Use formula to estimate the FLOPS of specific + operators (matrix multiplication and 2D convolution). + Default: False. + json_trace_path (str, optional): Exports the collected trace in Chrome + JSON format. Default: None. + + Example: + >>> runner = ... # instantiate a Runner + >>> # tensorboard trace + >>> trace_config = dict(type='tb_trace', dir_name='work_dir') + >>> profiler_config = dict(on_trace_ready=trace_config) + >>> runner.register_profiler_hook(profiler_config) + >>> runner.run(data_loaders=[trainloader], workflow=[('train', 1)]) + """ + + def __init__(self, + by_epoch: bool = True, + profile_iters: int = 1, + activities: List[str] = ['cpu', 'cuda'], + schedule: Optional[dict] = None, + on_trace_ready: Optional[Union[Callable, dict]] = None, + record_shapes: bool = False, + profile_memory: bool = False, + with_stack: bool = False, + with_flops: bool = False, + json_trace_path: Optional[str] = None) -> None: + try: + from torch import profiler # torch version >= 1.8.1 + except ImportError: + raise ImportError('profiler is the new feature of torch1.8.1, ' + f'but your version is {torch.__version__}') + + assert isinstance(by_epoch, bool), '``by_epoch`` should be a boolean.' + self.by_epoch = by_epoch + + if profile_iters < 1: + raise ValueError('profile_iters should be greater than 0, but got ' + f'{profile_iters}') + self.profile_iters = profile_iters + + if not isinstance(activities, list): + raise ValueError( + f'activities should be list, but got {type(activities)}') + self.activities = [] + for activity in activities: + activity = activity.lower() + if activity == 'cpu': + self.activities.append(profiler.ProfilerActivity.CPU) + elif activity == 'cuda': + self.activities.append(profiler.ProfilerActivity.CUDA) + else: + raise ValueError( + f'activity should be "cpu" or "cuda", but got {activity}') + + if schedule is not None: + self.schedule = profiler.schedule(**schedule) + else: + self.schedule = None + + self.on_trace_ready = on_trace_ready + self.record_shapes = record_shapes + self.profile_memory = profile_memory + self.with_stack = with_stack + self.with_flops = with_flops + self.json_trace_path = json_trace_path + + @master_only + def before_run(self, runner): + if self.by_epoch and runner.max_epochs < self.profile_iters: + raise ValueError('self.profile_iters should not be greater than ' + f'{runner.max_epochs}') + + if not self.by_epoch and runner.max_iters < self.profile_iters: + raise ValueError('self.profile_iters should not be greater than ' + f'{runner.max_iters}') + + if callable(self.on_trace_ready): # handler + _on_trace_ready = self.on_trace_ready + elif isinstance(self.on_trace_ready, dict): # config of handler + trace_cfg = self.on_trace_ready.copy() + trace_type = trace_cfg.pop('type') # log_trace handler + if trace_type == 'log_trace': + + def _log_handler(prof): + print(prof.key_averages().table(**trace_cfg)) + + _on_trace_ready = _log_handler + elif trace_type == 'tb_trace': # tensorboard_trace handler + try: + import torch_tb_profiler # noqa: F401 + except ImportError: + raise ImportError('please run "pip install ' + 'torch-tb-profiler" to install ' + 'torch_tb_profiler') + _on_trace_ready = torch.profiler.tensorboard_trace_handler( + **trace_cfg) + else: + raise ValueError('trace_type should be "log_trace" or ' + f'"tb_trace", but got {trace_type}') + elif self.on_trace_ready is None: + _on_trace_ready = None # type: ignore + else: + raise ValueError('on_trace_ready should be handler, dict or None, ' + f'but got {type(self.on_trace_ready)}') + + if self.by_epoch and runner.max_epochs > 1: + warnings.warn(f'profiler will profile {runner.max_epochs} epochs ' + 'instead of 1 epoch. Since profiler will slow down ' + 'the training, it is recommended to train 1 epoch ' + 'with ProfilerHook and adjust your setting according' + ' to the profiler summary. During normal training ' + '(epoch > 1), you may disable the ProfilerHook.') + + self.profiler = torch.profiler.profile( + activities=self.activities, + schedule=self.schedule, + on_trace_ready=_on_trace_ready, + record_shapes=self.record_shapes, + profile_memory=self.profile_memory, + with_stack=self.with_stack, + with_flops=self.with_flops) + + self.profiler.__enter__() + runner.logger.info('profiler is profiling...') + + @master_only + def after_train_epoch(self, runner): + if self.by_epoch and runner.epoch == self.profile_iters - 1: + runner.logger.info('profiler may take a few minutes...') + self.profiler.__exit__(None, None, None) + if self.json_trace_path is not None: + self.profiler.export_chrome_trace(self.json_trace_path) + + @master_only + def after_train_iter(self, runner): + self.profiler.step() + if not self.by_epoch and runner.iter == self.profile_iters - 1: + runner.logger.info('profiler may take a few minutes...') + self.profiler.__exit__(None, None, None) + if self.json_trace_path is not None: + self.profiler.export_chrome_trace(self.json_trace_path) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/sampler_seed.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/sampler_seed.py new file mode 100644 index 0000000000000000000000000000000000000000..ee0dc6bdd8df5775857028aaed5444c0f59caf80 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/sampler_seed.py @@ -0,0 +1,20 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .hook import HOOKS, Hook + + +@HOOKS.register_module() +class DistSamplerSeedHook(Hook): + """Data-loading sampler for distributed training. + + When distributed training, it is only useful in conjunction with + :obj:`EpochBasedRunner`, while :obj:`IterBasedRunner` achieves the same + purpose with :obj:`IterLoader`. + """ + + def before_epoch(self, runner): + if hasattr(runner.data_loader.sampler, 'set_epoch'): + # in case the data loader uses `SequentialSampler` in Pytorch + runner.data_loader.sampler.set_epoch(runner.epoch) + elif hasattr(runner.data_loader.batch_sampler.sampler, 'set_epoch'): + # batch sampler in pytorch warps the sampler as its attributes. + runner.data_loader.batch_sampler.sampler.set_epoch(runner.epoch) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/sync_buffer.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/sync_buffer.py new file mode 100644 index 0000000000000000000000000000000000000000..6376b7ff894280cb2782243b25e8973650591577 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/hooks/sync_buffer.py @@ -0,0 +1,22 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from ..dist_utils import allreduce_params +from .hook import HOOKS, Hook + + +@HOOKS.register_module() +class SyncBuffersHook(Hook): + """Synchronize model buffers such as running_mean and running_var in BN at + the end of each epoch. + + Args: + distributed (bool): Whether distributed training is used. It is + effective only for distributed training. Defaults to True. + """ + + def __init__(self, distributed=True): + self.distributed = distributed + + def after_epoch(self, runner): + """All-reduce model buffers at the end of each epoch.""" + if self.distributed: + allreduce_params(runner.model.buffers()) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/iter_based_runner.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/iter_based_runner.py new file mode 100644 index 0000000000000000000000000000000000000000..9892b07a4a496f9f217d598ea0140c27ca187ffe --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/iter_based_runner.py @@ -0,0 +1,273 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import platform +import shutil +import time +import warnings + +import torch +from torch.optim import Optimizer + +import mmcv +from .base_runner import BaseRunner +from .builder import RUNNERS +from .checkpoint import save_checkpoint +from .hooks import IterTimerHook +from .utils import get_host_info + + +class IterLoader: + + def __init__(self, dataloader): + self._dataloader = dataloader + self.iter_loader = iter(self._dataloader) + self._epoch = 0 + + @property + def epoch(self): + return self._epoch + + def __next__(self): + try: + data = next(self.iter_loader) + except StopIteration: + self._epoch += 1 + if hasattr(self._dataloader.sampler, 'set_epoch'): + self._dataloader.sampler.set_epoch(self._epoch) + time.sleep(2) # Prevent possible deadlock during epoch transition + self.iter_loader = iter(self._dataloader) + data = next(self.iter_loader) + + return data + + def __len__(self): + return len(self._dataloader) + + +@RUNNERS.register_module() +class IterBasedRunner(BaseRunner): + """Iteration-based Runner. + + This runner train models iteration by iteration. + """ + + def train(self, data_loader, **kwargs): + self.model.train() + self.mode = 'train' + self.data_loader = data_loader + self._epoch = data_loader.epoch + data_batch = next(data_loader) + self.call_hook('before_train_iter') + outputs = self.model.train_step(data_batch, self.optimizer, **kwargs) + if not isinstance(outputs, dict): + raise TypeError('model.train_step() must return a dict') + if 'log_vars' in outputs: + self.log_buffer.update(outputs['log_vars'], outputs['num_samples']) + self.outputs = outputs + self.call_hook('after_train_iter') + self._inner_iter += 1 + self._iter += 1 + + @torch.no_grad() + def val(self, data_loader, **kwargs): + self.model.eval() + self.mode = 'val' + self.data_loader = data_loader + data_batch = next(data_loader) + self.call_hook('before_val_iter') + outputs = self.model.val_step(data_batch, **kwargs) + if not isinstance(outputs, dict): + raise TypeError('model.val_step() must return a dict') + if 'log_vars' in outputs: + self.log_buffer.update(outputs['log_vars'], outputs['num_samples']) + self.outputs = outputs + self.call_hook('after_val_iter') + self._inner_iter += 1 + + def run(self, data_loaders, workflow, max_iters=None, **kwargs): + """Start running. + + Args: + data_loaders (list[:obj:`DataLoader`]): Dataloaders for training + and validation. + workflow (list[tuple]): A list of (phase, iters) to specify the + running order and iterations. E.g, [('train', 10000), + ('val', 1000)] means running 10000 iterations for training and + 1000 iterations for validation, iteratively. + """ + assert isinstance(data_loaders, list) + assert mmcv.is_list_of(workflow, tuple) + assert len(data_loaders) == len(workflow) + if max_iters is not None: + warnings.warn( + 'setting max_iters in run is deprecated, ' + 'please set max_iters in runner_config', DeprecationWarning) + self._max_iters = max_iters + assert self._max_iters is not None, ( + 'max_iters must be specified during instantiation') + + work_dir = self.work_dir if self.work_dir is not None else 'NONE' + self.logger.info('Start running, host: %s, work_dir: %s', + get_host_info(), work_dir) + self.logger.info('Hooks will be executed in the following order:\n%s', + self.get_hook_info()) + self.logger.info('workflow: %s, max: %d iters', workflow, + self._max_iters) + self.call_hook('before_run') + + iter_loaders = [IterLoader(x) for x in data_loaders] + + self.call_hook('before_epoch') + + while self.iter < self._max_iters: + for i, flow in enumerate(workflow): + self._inner_iter = 0 + mode, iters = flow + if not isinstance(mode, str) or not hasattr(self, mode): + raise ValueError( + 'runner has no method named "{}" to run a workflow'. + format(mode)) + iter_runner = getattr(self, mode) + for _ in range(iters): + if mode == 'train' and self.iter >= self._max_iters: + break + iter_runner(iter_loaders[i], **kwargs) + + time.sleep(1) # wait for some hooks like loggers to finish + self.call_hook('after_epoch') + self.call_hook('after_run') + + def resume(self, + checkpoint, + resume_optimizer=True, + map_location='default'): + """Resume model from checkpoint. + + Args: + checkpoint (str): Checkpoint to resume from. + resume_optimizer (bool, optional): Whether resume the optimizer(s) + if the checkpoint file includes optimizer(s). Default to True. + map_location (str, optional): Same as :func:`torch.load`. + Default to 'default'. + """ + if map_location == 'default': + device_id = torch.cuda.current_device() + checkpoint = self.load_checkpoint( + checkpoint, + map_location=lambda storage, loc: storage.cuda(device_id)) + else: + checkpoint = self.load_checkpoint( + checkpoint, map_location=map_location) + + self._epoch = checkpoint['meta']['epoch'] + self._iter = checkpoint['meta']['iter'] + self._inner_iter = checkpoint['meta']['iter'] + if 'optimizer' in checkpoint and resume_optimizer: + if isinstance(self.optimizer, Optimizer): + self.optimizer.load_state_dict(checkpoint['optimizer']) + elif isinstance(self.optimizer, dict): + for k in self.optimizer.keys(): + self.optimizer[k].load_state_dict( + checkpoint['optimizer'][k]) + else: + raise TypeError( + 'Optimizer should be dict or torch.optim.Optimizer ' + f'but got {type(self.optimizer)}') + + self.logger.info(f'resumed from epoch: {self.epoch}, iter {self.iter}') + + def save_checkpoint(self, + out_dir, + filename_tmpl='iter_{}.pth', + meta=None, + save_optimizer=True, + create_symlink=True): + """Save checkpoint to file. + + Args: + out_dir (str): Directory to save checkpoint files. + filename_tmpl (str, optional): Checkpoint file template. + Defaults to 'iter_{}.pth'. + meta (dict, optional): Metadata to be saved in checkpoint. + Defaults to None. + save_optimizer (bool, optional): Whether save optimizer. + Defaults to True. + create_symlink (bool, optional): Whether create symlink to the + latest checkpoint file. Defaults to True. + """ + if meta is None: + meta = {} + elif not isinstance(meta, dict): + raise TypeError( + f'meta should be a dict or None, but got {type(meta)}') + if self.meta is not None: + meta.update(self.meta) + # Note: meta.update(self.meta) should be done before + # meta.update(epoch=self.epoch + 1, iter=self.iter) otherwise + # there will be problems with resumed checkpoints. + # More details in https://github.com/open-mmlab/mmcv/pull/1108 + meta.update(epoch=self.epoch + 1, iter=self.iter) + + filename = filename_tmpl.format(self.iter + 1) + filepath = osp.join(out_dir, filename) + optimizer = self.optimizer if save_optimizer else None + save_checkpoint(self.model, filepath, optimizer=optimizer, meta=meta) + # in some environments, `os.symlink` is not supported, you may need to + # set `create_symlink` to False + if create_symlink: + dst_file = osp.join(out_dir, 'latest.pth') + if platform.system() != 'Windows': + mmcv.symlink(filename, dst_file) + else: + shutil.copy(filepath, dst_file) + + def register_training_hooks(self, + lr_config, + optimizer_config=None, + checkpoint_config=None, + log_config=None, + momentum_config=None, + custom_hooks_config=None): + """Register default hooks for iter-based training. + + Checkpoint hook, optimizer stepper hook and logger hooks will be set to + `by_epoch=False` by default. + + Default hooks include: + + +----------------------+-------------------------+ + | Hooks | Priority | + +======================+=========================+ + | LrUpdaterHook | VERY_HIGH (10) | + +----------------------+-------------------------+ + | MomentumUpdaterHook | HIGH (30) | + +----------------------+-------------------------+ + | OptimizerStepperHook | ABOVE_NORMAL (40) | + +----------------------+-------------------------+ + | CheckpointSaverHook | NORMAL (50) | + +----------------------+-------------------------+ + | IterTimerHook | LOW (70) | + +----------------------+-------------------------+ + | LoggerHook(s) | VERY_LOW (90) | + +----------------------+-------------------------+ + | CustomHook(s) | defaults to NORMAL (50) | + +----------------------+-------------------------+ + + If custom hooks have same priority with default hooks, custom hooks + will be triggered after default hooks. + """ + if checkpoint_config is not None: + checkpoint_config.setdefault('by_epoch', False) + if lr_config is not None: + lr_config.setdefault('by_epoch', False) + if log_config is not None: + for info in log_config['hooks']: + info.setdefault('by_epoch', False) + super(IterBasedRunner, self).register_training_hooks( + lr_config=lr_config, + momentum_config=momentum_config, + optimizer_config=optimizer_config, + checkpoint_config=checkpoint_config, + log_config=log_config, + timer_config=IterTimerHook(), + custom_hooks_config=custom_hooks_config) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/log_buffer.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/log_buffer.py new file mode 100644 index 0000000000000000000000000000000000000000..d949e2941c5400088c7cd8a1dc893d8b233ae785 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/log_buffer.py @@ -0,0 +1,41 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from collections import OrderedDict + +import numpy as np + + +class LogBuffer: + + def __init__(self): + self.val_history = OrderedDict() + self.n_history = OrderedDict() + self.output = OrderedDict() + self.ready = False + + def clear(self): + self.val_history.clear() + self.n_history.clear() + self.clear_output() + + def clear_output(self): + self.output.clear() + self.ready = False + + def update(self, vars, count=1): + assert isinstance(vars, dict) + for key, var in vars.items(): + if key not in self.val_history: + self.val_history[key] = [] + self.n_history[key] = [] + self.val_history[key].append(var) + self.n_history[key].append(count) + + def average(self, n=0): + """Average latest n values or all values.""" + assert n >= 0 + for key in self.val_history: + values = np.array(self.val_history[key][-n:]) + nums = np.array(self.n_history[key][-n:]) + avg = np.sum(values * nums) / np.sum(nums) + self.output[key] = avg + self.ready = True diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..53c34d0470992cbc374f29681fdd00dc0e57968d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/__init__.py @@ -0,0 +1,9 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .builder import (OPTIMIZER_BUILDERS, OPTIMIZERS, build_optimizer, + build_optimizer_constructor) +from .default_constructor import DefaultOptimizerConstructor + +__all__ = [ + 'OPTIMIZER_BUILDERS', 'OPTIMIZERS', 'DefaultOptimizerConstructor', + 'build_optimizer', 'build_optimizer_constructor' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/builder.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/builder.py new file mode 100644 index 0000000000000000000000000000000000000000..f9234eed8f1f186d9d8dfda34562157ee39bdb3a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/builder.py @@ -0,0 +1,44 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import inspect + +import torch + +from ...utils import Registry, build_from_cfg + +OPTIMIZERS = Registry('optimizer') +OPTIMIZER_BUILDERS = Registry('optimizer builder') + + +def register_torch_optimizers(): + torch_optimizers = [] + for module_name in dir(torch.optim): + if module_name.startswith('__'): + continue + _optim = getattr(torch.optim, module_name) + if inspect.isclass(_optim) and issubclass(_optim, + torch.optim.Optimizer): + OPTIMIZERS.register_module()(_optim) + torch_optimizers.append(module_name) + return torch_optimizers + + +TORCH_OPTIMIZERS = register_torch_optimizers() + + +def build_optimizer_constructor(cfg): + return build_from_cfg(cfg, OPTIMIZER_BUILDERS) + + +def build_optimizer(model, cfg): + optimizer_cfg = copy.deepcopy(cfg) + constructor_type = optimizer_cfg.pop('constructor', + 'DefaultOptimizerConstructor') + paramwise_cfg = optimizer_cfg.pop('paramwise_cfg', None) + optim_constructor = build_optimizer_constructor( + dict( + type=constructor_type, + optimizer_cfg=optimizer_cfg, + paramwise_cfg=paramwise_cfg)) + optimizer = optim_constructor(model) + return optimizer diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/default_constructor.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/default_constructor.py new file mode 100644 index 0000000000000000000000000000000000000000..ae97db8801343192b0169f06ebd9ad88c54c01b4 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/optimizer/default_constructor.py @@ -0,0 +1,250 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch +from torch.nn import GroupNorm, LayerNorm + +from mmcv.utils import _BatchNorm, _InstanceNorm, build_from_cfg, is_list_of +from mmcv.utils.ext_loader import check_ops_exist +from .builder import OPTIMIZER_BUILDERS, OPTIMIZERS + + +@OPTIMIZER_BUILDERS.register_module() +class DefaultOptimizerConstructor: + """Default constructor for optimizers. + + By default each parameter share the same optimizer settings, and we + provide an argument ``paramwise_cfg`` to specify parameter-wise settings. + It is a dict and may contain the following fields: + + - ``custom_keys`` (dict): Specified parameters-wise settings by keys. If + one of the keys in ``custom_keys`` is a substring of the name of one + parameter, then the setting of the parameter will be specified by + ``custom_keys[key]`` and other setting like ``bias_lr_mult`` etc. will + be ignored. It should be noted that the aforementioned ``key`` is the + longest key that is a substring of the name of the parameter. If there + are multiple matched keys with the same length, then the key with lower + alphabet order will be chosen. + ``custom_keys[key]`` should be a dict and may contain fields ``lr_mult`` + and ``decay_mult``. See Example 2 below. + - ``bias_lr_mult`` (float): It will be multiplied to the learning + rate for all bias parameters (except for those in normalization + layers and offset layers of DCN). + - ``bias_decay_mult`` (float): It will be multiplied to the weight + decay for all bias parameters (except for those in + normalization layers, depthwise conv layers, offset layers of DCN). + - ``norm_decay_mult`` (float): It will be multiplied to the weight + decay for all weight and bias parameters of normalization + layers. + - ``dwconv_decay_mult`` (float): It will be multiplied to the weight + decay for all weight and bias parameters of depthwise conv + layers. + - ``dcn_offset_lr_mult`` (float): It will be multiplied to the learning + rate for parameters of offset layer in the deformable convs + of a model. + - ``bypass_duplicate`` (bool): If true, the duplicate parameters + would not be added into optimizer. Default: False. + + Note: + + 1. If the option ``dcn_offset_lr_mult`` is used, the constructor will + override the effect of ``bias_lr_mult`` in the bias of offset layer. + So be careful when using both ``bias_lr_mult`` and + ``dcn_offset_lr_mult``. If you wish to apply both of them to the offset + layer in deformable convs, set ``dcn_offset_lr_mult`` to the original + ``dcn_offset_lr_mult`` * ``bias_lr_mult``. + + 2. If the option ``dcn_offset_lr_mult`` is used, the constructor will + apply it to all the DCN layers in the model. So be careful when the + model contains multiple DCN layers in places other than backbone. + + Args: + model (:obj:`nn.Module`): The model with parameters to be optimized. + optimizer_cfg (dict): The config dict of the optimizer. + Positional fields are + + - `type`: class name of the optimizer. + + Optional fields are + + - any arguments of the corresponding optimizer type, e.g., + lr, weight_decay, momentum, etc. + paramwise_cfg (dict, optional): Parameter-wise options. + + Example 1: + >>> model = torch.nn.modules.Conv1d(1, 1, 1) + >>> optimizer_cfg = dict(type='SGD', lr=0.01, momentum=0.9, + >>> weight_decay=0.0001) + >>> paramwise_cfg = dict(norm_decay_mult=0.) + >>> optim_builder = DefaultOptimizerConstructor( + >>> optimizer_cfg, paramwise_cfg) + >>> optimizer = optim_builder(model) + + Example 2: + >>> # assume model have attribute model.backbone and model.cls_head + >>> optimizer_cfg = dict(type='SGD', lr=0.01, weight_decay=0.95) + >>> paramwise_cfg = dict(custom_keys={ + 'backbone': dict(lr_mult=0.1, decay_mult=0.9)}) + >>> optim_builder = DefaultOptimizerConstructor( + >>> optimizer_cfg, paramwise_cfg) + >>> optimizer = optim_builder(model) + >>> # Then the `lr` and `weight_decay` for model.backbone is + >>> # (0.01 * 0.1, 0.95 * 0.9). `lr` and `weight_decay` for + >>> # model.cls_head is (0.01, 0.95). + """ + + def __init__(self, optimizer_cfg, paramwise_cfg=None): + if not isinstance(optimizer_cfg, dict): + raise TypeError('optimizer_cfg should be a dict', + f'but got {type(optimizer_cfg)}') + self.optimizer_cfg = optimizer_cfg + self.paramwise_cfg = {} if paramwise_cfg is None else paramwise_cfg + self.base_lr = optimizer_cfg.get('lr', None) + self.base_wd = optimizer_cfg.get('weight_decay', None) + self._validate_cfg() + + def _validate_cfg(self): + if not isinstance(self.paramwise_cfg, dict): + raise TypeError('paramwise_cfg should be None or a dict, ' + f'but got {type(self.paramwise_cfg)}') + + if 'custom_keys' in self.paramwise_cfg: + if not isinstance(self.paramwise_cfg['custom_keys'], dict): + raise TypeError( + 'If specified, custom_keys must be a dict, ' + f'but got {type(self.paramwise_cfg["custom_keys"])}') + if self.base_wd is None: + for key in self.paramwise_cfg['custom_keys']: + if 'decay_mult' in self.paramwise_cfg['custom_keys'][key]: + raise ValueError('base_wd should not be None') + + # get base lr and weight decay + # weight_decay must be explicitly specified if mult is specified + if ('bias_decay_mult' in self.paramwise_cfg + or 'norm_decay_mult' in self.paramwise_cfg + or 'dwconv_decay_mult' in self.paramwise_cfg): + if self.base_wd is None: + raise ValueError('base_wd should not be None') + + def _is_in(self, param_group, param_group_list): + assert is_list_of(param_group_list, dict) + param = set(param_group['params']) + param_set = set() + for group in param_group_list: + param_set.update(set(group['params'])) + + return not param.isdisjoint(param_set) + + def add_params(self, params, module, prefix='', is_dcn_module=None): + """Add all parameters of module to the params list. + + The parameters of the given module will be added to the list of param + groups, with specific rules defined by paramwise_cfg. + + Args: + params (list[dict]): A list of param groups, it will be modified + in place. + module (nn.Module): The module to be added. + prefix (str): The prefix of the module + is_dcn_module (int|float|None): If the current module is a + submodule of DCN, `is_dcn_module` will be passed to + control conv_offset layer's learning rate. Defaults to None. + """ + # get param-wise options + custom_keys = self.paramwise_cfg.get('custom_keys', {}) + # first sort with alphabet order and then sort with reversed len of str + sorted_keys = sorted(sorted(custom_keys.keys()), key=len, reverse=True) + + bias_lr_mult = self.paramwise_cfg.get('bias_lr_mult', 1.) + bias_decay_mult = self.paramwise_cfg.get('bias_decay_mult', 1.) + norm_decay_mult = self.paramwise_cfg.get('norm_decay_mult', 1.) + dwconv_decay_mult = self.paramwise_cfg.get('dwconv_decay_mult', 1.) + bypass_duplicate = self.paramwise_cfg.get('bypass_duplicate', False) + dcn_offset_lr_mult = self.paramwise_cfg.get('dcn_offset_lr_mult', 1.) + + # special rules for norm layers and depth-wise conv layers + is_norm = isinstance(module, + (_BatchNorm, _InstanceNorm, GroupNorm, LayerNorm)) + is_dwconv = ( + isinstance(module, torch.nn.Conv2d) + and module.in_channels == module.groups) + + for name, param in module.named_parameters(recurse=False): + param_group = {'params': [param]} + if not param.requires_grad: + params.append(param_group) + continue + if bypass_duplicate and self._is_in(param_group, params): + warnings.warn(f'{prefix} is duplicate. It is skipped since ' + f'bypass_duplicate={bypass_duplicate}') + continue + # if the parameter match one of the custom keys, ignore other rules + is_custom = False + for key in sorted_keys: + if key in f'{prefix}.{name}': + is_custom = True + lr_mult = custom_keys[key].get('lr_mult', 1.) + param_group['lr'] = self.base_lr * lr_mult + if self.base_wd is not None: + decay_mult = custom_keys[key].get('decay_mult', 1.) + param_group['weight_decay'] = self.base_wd * decay_mult + break + + if not is_custom: + # bias_lr_mult affects all bias parameters + # except for norm.bias dcn.conv_offset.bias + if name == 'bias' and not (is_norm or is_dcn_module): + param_group['lr'] = self.base_lr * bias_lr_mult + + if (prefix.find('conv_offset') != -1 and is_dcn_module + and isinstance(module, torch.nn.Conv2d)): + # deal with both dcn_offset's bias & weight + param_group['lr'] = self.base_lr * dcn_offset_lr_mult + + # apply weight decay policies + if self.base_wd is not None: + # norm decay + if is_norm: + param_group[ + 'weight_decay'] = self.base_wd * norm_decay_mult + # depth-wise conv + elif is_dwconv: + param_group[ + 'weight_decay'] = self.base_wd * dwconv_decay_mult + # bias lr and decay + elif name == 'bias' and not is_dcn_module: + # TODO: current bias_decay_mult will have affect on DCN + param_group[ + 'weight_decay'] = self.base_wd * bias_decay_mult + params.append(param_group) + + if check_ops_exist(): + from mmcv.ops import DeformConv2d, ModulatedDeformConv2d + is_dcn_module = isinstance(module, + (DeformConv2d, ModulatedDeformConv2d)) + else: + is_dcn_module = False + for child_name, child_mod in module.named_children(): + child_prefix = f'{prefix}.{child_name}' if prefix else child_name + self.add_params( + params, + child_mod, + prefix=child_prefix, + is_dcn_module=is_dcn_module) + + def __call__(self, model): + if hasattr(model, 'module'): + model = model.module + + optimizer_cfg = self.optimizer_cfg.copy() + # if no paramwise option is specified, just use the global setting + if not self.paramwise_cfg: + optimizer_cfg['params'] = model.parameters() + return build_from_cfg(optimizer_cfg, OPTIMIZERS) + + # set param-wise lr and weight decay recursively + params = [] + self.add_params(params, model) + optimizer_cfg['params'] = params + + return build_from_cfg(optimizer_cfg, OPTIMIZERS) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/priority.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/priority.py new file mode 100644 index 0000000000000000000000000000000000000000..64cc4e3a05f8d5b89ab6eb32461e6e80f1d62e67 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/priority.py @@ -0,0 +1,60 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from enum import Enum + + +class Priority(Enum): + """Hook priority levels. + + +--------------+------------+ + | Level | Value | + +==============+============+ + | HIGHEST | 0 | + +--------------+------------+ + | VERY_HIGH | 10 | + +--------------+------------+ + | HIGH | 30 | + +--------------+------------+ + | ABOVE_NORMAL | 40 | + +--------------+------------+ + | NORMAL | 50 | + +--------------+------------+ + | BELOW_NORMAL | 60 | + +--------------+------------+ + | LOW | 70 | + +--------------+------------+ + | VERY_LOW | 90 | + +--------------+------------+ + | LOWEST | 100 | + +--------------+------------+ + """ + + HIGHEST = 0 + VERY_HIGH = 10 + HIGH = 30 + ABOVE_NORMAL = 40 + NORMAL = 50 + BELOW_NORMAL = 60 + LOW = 70 + VERY_LOW = 90 + LOWEST = 100 + + +def get_priority(priority): + """Get priority value. + + Args: + priority (int or str or :obj:`Priority`): Priority. + + Returns: + int: The priority value. + """ + if isinstance(priority, int): + if priority < 0 or priority > 100: + raise ValueError('priority must be between 0 and 100') + return priority + elif isinstance(priority, Priority): + return priority.value + elif isinstance(priority, str): + return Priority[priority.upper()].value + else: + raise TypeError('priority must be an integer or Priority enum value') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/utils.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/utils.py new file mode 100644 index 0000000000000000000000000000000000000000..144d11e1a857f740964ab20e13daf52be7c3d848 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/runner/utils.py @@ -0,0 +1,93 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os +import random +import sys +import time +import warnings +from getpass import getuser +from socket import gethostname + +import numpy as np +import torch + +import mmcv + + +def get_host_info(): + """Get hostname and username. + + Return empty string if exception raised, e.g. ``getpass.getuser()`` will + lead to error in docker container + """ + host = '' + try: + host = f'{getuser()}@{gethostname()}' + except Exception as e: + warnings.warn(f'Host or user not found: {str(e)}') + finally: + return host + + +def get_time_str(): + return time.strftime('%Y%m%d_%H%M%S', time.localtime()) + + +def obj_from_dict(info, parent=None, default_args=None): + """Initialize an object from dict. + + The dict must contain the key "type", which indicates the object type, it + can be either a string or type, such as "list" or ``list``. Remaining + fields are treated as the arguments for constructing the object. + + Args: + info (dict): Object types and arguments. + parent (:class:`module`): Module which may containing expected object + classes. + default_args (dict, optional): Default arguments for initializing the + object. + + Returns: + any type: Object built from the dict. + """ + assert isinstance(info, dict) and 'type' in info + assert isinstance(default_args, dict) or default_args is None + args = info.copy() + obj_type = args.pop('type') + if mmcv.is_str(obj_type): + if parent is not None: + obj_type = getattr(parent, obj_type) + else: + obj_type = sys.modules[obj_type] + elif not isinstance(obj_type, type): + raise TypeError('type must be a str or valid type, but ' + f'got {type(obj_type)}') + if default_args is not None: + for name, value in default_args.items(): + args.setdefault(name, value) + return obj_type(**args) + + +def set_random_seed(seed, deterministic=False, use_rank_shift=False): + """Set random seed. + + Args: + seed (int): Seed to be used. + deterministic (bool): Whether to set the deterministic option for + CUDNN backend, i.e., set `torch.backends.cudnn.deterministic` + to True and `torch.backends.cudnn.benchmark` to False. + Default: False. + rank_shift (bool): Whether to add rank number to the random seed to + have different random seed in different threads. Default: False. + """ + if use_rank_shift: + rank, _ = mmcv.runner.get_dist_info() + seed += rank + random.seed(seed) + np.random.seed(seed) + torch.manual_seed(seed) + torch.cuda.manual_seed(seed) + torch.cuda.manual_seed_all(seed) + os.environ['PYTHONHASHSEED'] = str(seed) + if deterministic: + torch.backends.cudnn.deterministic = True + torch.backends.cudnn.benchmark = False diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e0825ed6709692849d43f3fec88ad3165c6d69a6 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/__init__.py @@ -0,0 +1,78 @@ +# flake8: noqa +# Copyright (c) OpenMMLab. All rights reserved. +from .config import Config, ConfigDict, DictAction +from .misc import (check_prerequisites, concat_list, deprecated_api_warning, + has_method, import_modules_from_strings, is_list_of, + is_method_overridden, is_seq_of, is_str, is_tuple_of, + iter_cast, list_cast, requires_executable, requires_package, + slice_list, to_1tuple, to_2tuple, to_3tuple, to_4tuple, + to_ntuple, tuple_cast) +from .path import (check_file_exist, fopen, is_filepath, mkdir_or_exist, + scandir, symlink) +from .progressbar import (ProgressBar, track_iter_progress, + track_parallel_progress, track_progress) +from .testing import (assert_attrs_equal, assert_dict_contains_subset, + assert_dict_has_keys, assert_is_norm_layer, + assert_keys_equal, assert_params_all_zeros, + check_python_script) +from .timer import Timer, TimerError, check_time +from .version_utils import digit_version, get_git_hash + +try: + import torch +except ImportError: + __all__ = [ + 'Config', 'ConfigDict', 'DictAction', 'is_str', 'iter_cast', + 'list_cast', 'tuple_cast', 'is_seq_of', 'is_list_of', 'is_tuple_of', + 'slice_list', 'concat_list', 'check_prerequisites', 'requires_package', + 'requires_executable', 'is_filepath', 'fopen', 'check_file_exist', + 'mkdir_or_exist', 'symlink', 'scandir', 'ProgressBar', + 'track_progress', 'track_iter_progress', 'track_parallel_progress', + 'Timer', 'TimerError', 'check_time', 'deprecated_api_warning', + 'digit_version', 'get_git_hash', 'import_modules_from_strings', + 'assert_dict_contains_subset', 'assert_attrs_equal', + 'assert_dict_has_keys', 'assert_keys_equal', 'check_python_script', + 'to_1tuple', 'to_2tuple', 'to_3tuple', 'to_4tuple', 'to_ntuple', + 'is_method_overridden', 'has_method' + ] +else: + from .device_type import IS_IPU_AVAILABLE, IS_MLU_AVAILABLE + from .env import collect_env + from .hub import load_url + from .logging import get_logger, print_log + from .parrots_jit import jit, skip_no_elena + # yapf: disable + from .parrots_wrapper import (IS_CUDA_AVAILABLE, TORCH_VERSION, + BuildExtension, CppExtension, CUDAExtension, + DataLoader, PoolDataLoader, SyncBatchNorm, + _AdaptiveAvgPoolNd, _AdaptiveMaxPoolNd, + _AvgPoolNd, _BatchNorm, _ConvNd, + _ConvTransposeMixin, _get_cuda_home, + _InstanceNorm, _MaxPoolNd, get_build_config, + is_rocm_pytorch) + # yapf: enable + from .registry import Registry, build_from_cfg + from .seed import worker_init_fn + from .trace import is_jit_tracing + __all__ = [ + 'Config', 'ConfigDict', 'DictAction', 'collect_env', 'get_logger', + 'print_log', 'is_str', 'iter_cast', 'list_cast', 'tuple_cast', + 'is_seq_of', 'is_list_of', 'is_tuple_of', 'slice_list', 'concat_list', + 'check_prerequisites', 'requires_package', 'requires_executable', + 'is_filepath', 'fopen', 'check_file_exist', 'mkdir_or_exist', + 'symlink', 'scandir', 'ProgressBar', 'track_progress', + 'track_iter_progress', 'track_parallel_progress', 'Registry', + 'build_from_cfg', 'Timer', 'TimerError', 'check_time', 'SyncBatchNorm', + '_AdaptiveAvgPoolNd', '_AdaptiveMaxPoolNd', '_AvgPoolNd', '_BatchNorm', + '_ConvNd', '_ConvTransposeMixin', '_InstanceNorm', '_MaxPoolNd', + 'get_build_config', 'BuildExtension', 'CppExtension', 'CUDAExtension', + 'DataLoader', 'PoolDataLoader', 'TORCH_VERSION', + 'deprecated_api_warning', 'digit_version', 'get_git_hash', + 'import_modules_from_strings', 'jit', 'skip_no_elena', + 'assert_dict_contains_subset', 'assert_attrs_equal', + 'assert_dict_has_keys', 'assert_keys_equal', 'assert_is_norm_layer', + 'assert_params_all_zeros', 'check_python_script', + 'is_method_overridden', 'is_jit_tracing', 'is_rocm_pytorch', + '_get_cuda_home', 'load_url', 'has_method', 'IS_CUDA_AVAILABLE', + 'worker_init_fn', 'IS_MLU_AVAILABLE', 'IS_IPU_AVAILABLE' + ] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/config.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/config.py new file mode 100644 index 0000000000000000000000000000000000000000..8efbc24e275764453a36158878020ab2e8472b1f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/config.py @@ -0,0 +1,719 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import ast +import copy +import os +import os.path as osp +import platform +import shutil +import sys +import tempfile +import types +import uuid +import warnings +from argparse import Action, ArgumentParser +from collections import abc +from importlib import import_module +from pathlib import Path + +from addict import Dict +from yapf.yapflib.yapf_api import FormatCode + +from .misc import import_modules_from_strings +from .path import check_file_exist + +if platform.system() == 'Windows': + import regex as re +else: + import re + +BASE_KEY = '_base_' +DELETE_KEY = '_delete_' +DEPRECATION_KEY = '_deprecation_' +RESERVED_KEYS = ['filename', 'text', 'pretty_text'] + + +class ConfigDict(Dict): + + def __missing__(self, name): + raise KeyError(name) + + def __getattr__(self, name): + try: + value = super(ConfigDict, self).__getattr__(name) + except KeyError: + ex = AttributeError(f"'{self.__class__.__name__}' object has no " + f"attribute '{name}'") + except Exception as e: + ex = e + else: + return value + raise ex + + +def add_args(parser, cfg, prefix=''): + for k, v in cfg.items(): + if isinstance(v, str): + parser.add_argument('--' + prefix + k) + elif isinstance(v, int): + parser.add_argument('--' + prefix + k, type=int) + elif isinstance(v, float): + parser.add_argument('--' + prefix + k, type=float) + elif isinstance(v, bool): + parser.add_argument('--' + prefix + k, action='store_true') + elif isinstance(v, dict): + add_args(parser, v, prefix + k + '.') + elif isinstance(v, abc.Iterable): + parser.add_argument('--' + prefix + k, type=type(v[0]), nargs='+') + else: + print(f'cannot parse key {prefix + k} of type {type(v)}') + return parser + + +class Config: + """A facility for config and config files. + + It supports common file formats as configs: python/json/yaml. The interface + is the same as a dict object and also allows access config values as + attributes. + + Example: + >>> cfg = Config(dict(a=1, b=dict(b1=[0, 1]))) + >>> cfg.a + 1 + >>> cfg.b + {'b1': [0, 1]} + >>> cfg.b.b1 + [0, 1] + >>> cfg = Config.fromfile('tests/data/config/a.py') + >>> cfg.filename + "/home/kchen/projects/mmcv/tests/data/config/a.py" + >>> cfg.item4 + 'test' + >>> cfg + "Config [path: /home/kchen/projects/mmcv/tests/data/config/a.py]: " + "{'item1': [1, 2], 'item2': {'a': 0}, 'item3': True, 'item4': 'test'}" + """ + + @staticmethod + def _validate_py_syntax(filename): + with open(filename, 'r', encoding='utf-8') as f: + # Setting encoding explicitly to resolve coding issue on windows + content = f.read() + try: + ast.parse(content) + except SyntaxError as e: + raise SyntaxError('There are syntax errors in config ' + f'file {filename}: {e}') + + @staticmethod + def _substitute_predefined_vars(filename, temp_config_name): + file_dirname = osp.dirname(filename) + file_basename = osp.basename(filename) + file_basename_no_extension = osp.splitext(file_basename)[0] + file_extname = osp.splitext(filename)[1] + support_templates = dict( + fileDirname=file_dirname, + fileBasename=file_basename, + fileBasenameNoExtension=file_basename_no_extension, + fileExtname=file_extname) + with open(filename, 'r', encoding='utf-8') as f: + # Setting encoding explicitly to resolve coding issue on windows + config_file = f.read() + for key, value in support_templates.items(): + regexp = r'\{\{\s*' + str(key) + r'\s*\}\}' + value = value.replace('\\', '/') + config_file = re.sub(regexp, value, config_file) + with open(temp_config_name, 'w', encoding='utf-8') as tmp_config_file: + tmp_config_file.write(config_file) + + @staticmethod + def _pre_substitute_base_vars(filename, temp_config_name): + """Substitute base variable placehoders to string, so that parsing + would work.""" + with open(filename, 'r', encoding='utf-8') as f: + # Setting encoding explicitly to resolve coding issue on windows + config_file = f.read() + base_var_dict = {} + regexp = r'\{\{\s*' + BASE_KEY + r'\.([\w\.]+)\s*\}\}' + base_vars = set(re.findall(regexp, config_file)) + for base_var in base_vars: + randstr = f'_{base_var}_{uuid.uuid4().hex.lower()[:6]}' + base_var_dict[randstr] = base_var + regexp = r'\{\{\s*' + BASE_KEY + r'\.' + base_var + r'\s*\}\}' + config_file = re.sub(regexp, f'"{randstr}"', config_file) + with open(temp_config_name, 'w', encoding='utf-8') as tmp_config_file: + tmp_config_file.write(config_file) + return base_var_dict + + @staticmethod + def _substitute_base_vars(cfg, base_var_dict, base_cfg): + """Substitute variable strings to their actual values.""" + cfg = copy.deepcopy(cfg) + + if isinstance(cfg, dict): + for k, v in cfg.items(): + if isinstance(v, str) and v in base_var_dict: + new_v = base_cfg + for new_k in base_var_dict[v].split('.'): + new_v = new_v[new_k] + cfg[k] = new_v + elif isinstance(v, (list, tuple, dict)): + cfg[k] = Config._substitute_base_vars( + v, base_var_dict, base_cfg) + elif isinstance(cfg, tuple): + cfg = tuple( + Config._substitute_base_vars(c, base_var_dict, base_cfg) + for c in cfg) + elif isinstance(cfg, list): + cfg = [ + Config._substitute_base_vars(c, base_var_dict, base_cfg) + for c in cfg + ] + elif isinstance(cfg, str) and cfg in base_var_dict: + new_v = base_cfg + for new_k in base_var_dict[cfg].split('.'): + new_v = new_v[new_k] + cfg = new_v + + return cfg + + @staticmethod + def _file2dict(filename, use_predefined_variables=True): + filename = osp.abspath(osp.expanduser(filename)) + check_file_exist(filename) + fileExtname = osp.splitext(filename)[1] + if fileExtname not in ['.py', '.json', '.yaml', '.yml']: + raise IOError('Only py/yml/yaml/json type are supported now!') + + with tempfile.TemporaryDirectory() as temp_config_dir: + temp_config_file = tempfile.NamedTemporaryFile( + dir=temp_config_dir, suffix=fileExtname) + if platform.system() == 'Windows': + temp_config_file.close() + temp_config_name = osp.basename(temp_config_file.name) + # Substitute predefined variables + if use_predefined_variables: + Config._substitute_predefined_vars(filename, + temp_config_file.name) + else: + shutil.copyfile(filename, temp_config_file.name) + # Substitute base variables from placeholders to strings + base_var_dict = Config._pre_substitute_base_vars( + temp_config_file.name, temp_config_file.name) + + if filename.endswith('.py'): + temp_module_name = osp.splitext(temp_config_name)[0] + sys.path.insert(0, temp_config_dir) + Config._validate_py_syntax(filename) + mod = import_module(temp_module_name) + sys.path.pop(0) + cfg_dict = { + name: value + for name, value in mod.__dict__.items() + if not name.startswith('__') + and not isinstance(value, types.ModuleType) + and not isinstance(value, types.FunctionType) + } + # delete imported module + del sys.modules[temp_module_name] + elif filename.endswith(('.yml', '.yaml', '.json')): + import mmcv + cfg_dict = mmcv.load(temp_config_file.name) + # close temp file + temp_config_file.close() + + # check deprecation information + if DEPRECATION_KEY in cfg_dict: + deprecation_info = cfg_dict.pop(DEPRECATION_KEY) + warning_msg = f'The config file {filename} will be deprecated ' \ + 'in the future.' + if 'expected' in deprecation_info: + warning_msg += f' Please use {deprecation_info["expected"]} ' \ + 'instead.' + if 'reference' in deprecation_info: + warning_msg += ' More information can be found at ' \ + f'{deprecation_info["reference"]}' + warnings.warn(warning_msg, DeprecationWarning) + + cfg_text = filename + '\n' + with open(filename, 'r', encoding='utf-8') as f: + # Setting encoding explicitly to resolve coding issue on windows + cfg_text += f.read() + + if BASE_KEY in cfg_dict: + cfg_dir = osp.dirname(filename) + base_filename = cfg_dict.pop(BASE_KEY) + base_filename = base_filename if isinstance( + base_filename, list) else [base_filename] + + cfg_dict_list = list() + cfg_text_list = list() + for f in base_filename: + _cfg_dict, _cfg_text = Config._file2dict(osp.join(cfg_dir, f)) + cfg_dict_list.append(_cfg_dict) + cfg_text_list.append(_cfg_text) + + base_cfg_dict = dict() + for c in cfg_dict_list: + duplicate_keys = base_cfg_dict.keys() & c.keys() + if len(duplicate_keys) > 0: + raise KeyError('Duplicate key is not allowed among bases. ' + f'Duplicate keys: {duplicate_keys}') + base_cfg_dict.update(c) + + # Substitute base variables from strings to their actual values + cfg_dict = Config._substitute_base_vars(cfg_dict, base_var_dict, + base_cfg_dict) + + base_cfg_dict = Config._merge_a_into_b(cfg_dict, base_cfg_dict) + cfg_dict = base_cfg_dict + + # merge cfg_text + cfg_text_list.append(cfg_text) + cfg_text = '\n'.join(cfg_text_list) + + return cfg_dict, cfg_text + + @staticmethod + def _merge_a_into_b(a, b, allow_list_keys=False): + """merge dict ``a`` into dict ``b`` (non-inplace). + + Values in ``a`` will overwrite ``b``. ``b`` is copied first to avoid + in-place modifications. + + Args: + a (dict): The source dict to be merged into ``b``. + b (dict): The origin dict to be fetch keys from ``a``. + allow_list_keys (bool): If True, int string keys (e.g. '0', '1') + are allowed in source ``a`` and will replace the element of the + corresponding index in b if b is a list. Default: False. + + Returns: + dict: The modified dict of ``b`` using ``a``. + + Examples: + # Normally merge a into b. + >>> Config._merge_a_into_b( + ... dict(obj=dict(a=2)), dict(obj=dict(a=1))) + {'obj': {'a': 2}} + + # Delete b first and merge a into b. + >>> Config._merge_a_into_b( + ... dict(obj=dict(_delete_=True, a=2)), dict(obj=dict(a=1))) + {'obj': {'a': 2}} + + # b is a list + >>> Config._merge_a_into_b( + ... {'0': dict(a=2)}, [dict(a=1), dict(b=2)], True) + [{'a': 2}, {'b': 2}] + """ + b = b.copy() + for k, v in a.items(): + if allow_list_keys and k.isdigit() and isinstance(b, list): + k = int(k) + if len(b) <= k: + raise KeyError(f'Index {k} exceeds the length of list {b}') + b[k] = Config._merge_a_into_b(v, b[k], allow_list_keys) + elif isinstance(v, dict): + if k in b and not v.pop(DELETE_KEY, False): + allowed_types = (dict, list) if allow_list_keys else dict + if not isinstance(b[k], allowed_types): + raise TypeError( + f'{k}={v} in child config cannot inherit from ' + f'base because {k} is a dict in the child config ' + f'but is of type {type(b[k])} in base config. ' + f'You may set `{DELETE_KEY}=True` to ignore the ' + f'base config.') + b[k] = Config._merge_a_into_b(v, b[k], allow_list_keys) + else: + b[k] = ConfigDict(v) + else: + b[k] = v + return b + + @staticmethod + def fromfile(filename, + use_predefined_variables=True, + import_custom_modules=True): + if isinstance(filename, Path): + filename = str(filename) + cfg_dict, cfg_text = Config._file2dict(filename, + use_predefined_variables) + if import_custom_modules and cfg_dict.get('custom_imports', None): + import_modules_from_strings(**cfg_dict['custom_imports']) + return Config(cfg_dict, cfg_text=cfg_text, filename=filename) + + @staticmethod + def fromstring(cfg_str, file_format): + """Generate config from config str. + + Args: + cfg_str (str): Config str. + file_format (str): Config file format corresponding to the + config str. Only py/yml/yaml/json type are supported now! + + Returns: + :obj:`Config`: Config obj. + """ + if file_format not in ['.py', '.json', '.yaml', '.yml']: + raise IOError('Only py/yml/yaml/json type are supported now!') + if file_format != '.py' and 'dict(' in cfg_str: + # check if users specify a wrong suffix for python + warnings.warn( + 'Please check "file_format", the file format may be .py') + with tempfile.NamedTemporaryFile( + 'w', encoding='utf-8', suffix=file_format, + delete=False) as temp_file: + temp_file.write(cfg_str) + # on windows, previous implementation cause error + # see PR 1077 for details + cfg = Config.fromfile(temp_file.name) + os.remove(temp_file.name) + return cfg + + @staticmethod + def auto_argparser(description=None): + """Generate argparser from config file automatically (experimental)""" + partial_parser = ArgumentParser(description=description) + partial_parser.add_argument('config', help='config file path') + cfg_file = partial_parser.parse_known_args()[0].config + cfg = Config.fromfile(cfg_file) + parser = ArgumentParser(description=description) + parser.add_argument('config', help='config file path') + add_args(parser, cfg) + return parser, cfg + + def __init__(self, cfg_dict=None, cfg_text=None, filename=None): + if cfg_dict is None: + cfg_dict = dict() + elif not isinstance(cfg_dict, dict): + raise TypeError('cfg_dict must be a dict, but ' + f'got {type(cfg_dict)}') + for key in cfg_dict: + if key in RESERVED_KEYS: + raise KeyError(f'{key} is reserved for config file') + + if isinstance(filename, Path): + filename = str(filename) + + super(Config, self).__setattr__('_cfg_dict', ConfigDict(cfg_dict)) + super(Config, self).__setattr__('_filename', filename) + if cfg_text: + text = cfg_text + elif filename: + with open(filename, 'r') as f: + text = f.read() + else: + text = '' + super(Config, self).__setattr__('_text', text) + + @property + def filename(self): + return self._filename + + @property + def text(self): + return self._text + + @property + def pretty_text(self): + + indent = 4 + + def _indent(s_, num_spaces): + s = s_.split('\n') + if len(s) == 1: + return s_ + first = s.pop(0) + s = [(num_spaces * ' ') + line for line in s] + s = '\n'.join(s) + s = first + '\n' + s + return s + + def _format_basic_types(k, v, use_mapping=False): + if isinstance(v, str): + v_str = f"'{v}'" + else: + v_str = str(v) + + if use_mapping: + k_str = f"'{k}'" if isinstance(k, str) else str(k) + attr_str = f'{k_str}: {v_str}' + else: + attr_str = f'{str(k)}={v_str}' + attr_str = _indent(attr_str, indent) + + return attr_str + + def _format_list(k, v, use_mapping=False): + # check if all items in the list are dict + if all(isinstance(_, dict) for _ in v): + v_str = '[\n' + v_str += '\n'.join( + f'dict({_indent(_format_dict(v_), indent)}),' + for v_ in v).rstrip(',') + if use_mapping: + k_str = f"'{k}'" if isinstance(k, str) else str(k) + attr_str = f'{k_str}: {v_str}' + else: + attr_str = f'{str(k)}={v_str}' + attr_str = _indent(attr_str, indent) + ']' + else: + attr_str = _format_basic_types(k, v, use_mapping) + return attr_str + + def _contain_invalid_identifier(dict_str): + contain_invalid_identifier = False + for key_name in dict_str: + contain_invalid_identifier |= \ + (not str(key_name).isidentifier()) + return contain_invalid_identifier + + def _format_dict(input_dict, outest_level=False): + r = '' + s = [] + + use_mapping = _contain_invalid_identifier(input_dict) + if use_mapping: + r += '{' + for idx, (k, v) in enumerate(input_dict.items()): + is_last = idx >= len(input_dict) - 1 + end = '' if outest_level or is_last else ',' + if isinstance(v, dict): + v_str = '\n' + _format_dict(v) + if use_mapping: + k_str = f"'{k}'" if isinstance(k, str) else str(k) + attr_str = f'{k_str}: dict({v_str}' + else: + attr_str = f'{str(k)}=dict({v_str}' + attr_str = _indent(attr_str, indent) + ')' + end + elif isinstance(v, list): + attr_str = _format_list(k, v, use_mapping) + end + else: + attr_str = _format_basic_types(k, v, use_mapping) + end + + s.append(attr_str) + r += '\n'.join(s) + if use_mapping: + r += '}' + return r + + cfg_dict = self._cfg_dict.to_dict() + text = _format_dict(cfg_dict, outest_level=True) + # copied from setup.cfg + yapf_style = dict( + based_on_style='pep8', + blank_line_before_nested_class_or_def=True, + split_before_expression_after_opening_paren=True) + text, _ = FormatCode(text, style_config=yapf_style, verify=True) + + return text + + def __repr__(self): + return f'Config (path: {self.filename}): {self._cfg_dict.__repr__()}' + + def __len__(self): + return len(self._cfg_dict) + + def __getattr__(self, name): + return getattr(self._cfg_dict, name) + + def __getitem__(self, name): + return self._cfg_dict.__getitem__(name) + + def __setattr__(self, name, value): + if isinstance(value, dict): + value = ConfigDict(value) + self._cfg_dict.__setattr__(name, value) + + def __setitem__(self, name, value): + if isinstance(value, dict): + value = ConfigDict(value) + self._cfg_dict.__setitem__(name, value) + + def __iter__(self): + return iter(self._cfg_dict) + + def __getstate__(self): + return (self._cfg_dict, self._filename, self._text) + + def __copy__(self): + cls = self.__class__ + other = cls.__new__(cls) + other.__dict__.update(self.__dict__) + + return other + + def __deepcopy__(self, memo): + cls = self.__class__ + other = cls.__new__(cls) + memo[id(self)] = other + + for key, value in self.__dict__.items(): + super(Config, other).__setattr__(key, copy.deepcopy(value, memo)) + + return other + + def __setstate__(self, state): + _cfg_dict, _filename, _text = state + super(Config, self).__setattr__('_cfg_dict', _cfg_dict) + super(Config, self).__setattr__('_filename', _filename) + super(Config, self).__setattr__('_text', _text) + + def dump(self, file=None): + cfg_dict = super(Config, self).__getattribute__('_cfg_dict').to_dict() + if self.filename.endswith('.py'): + if file is None: + return self.pretty_text + else: + with open(file, 'w', encoding='utf-8') as f: + f.write(self.pretty_text) + else: + import mmcv + if file is None: + file_format = self.filename.split('.')[-1] + return mmcv.dump(cfg_dict, file_format=file_format) + else: + mmcv.dump(cfg_dict, file) + + def merge_from_dict(self, options, allow_list_keys=True): + """Merge list into cfg_dict. + + Merge the dict parsed by MultipleKVAction into this cfg. + + Examples: + >>> options = {'model.backbone.depth': 50, + ... 'model.backbone.with_cp':True} + >>> cfg = Config(dict(model=dict(backbone=dict(type='ResNet')))) + >>> cfg.merge_from_dict(options) + >>> cfg_dict = super(Config, self).__getattribute__('_cfg_dict') + >>> assert cfg_dict == dict( + ... model=dict(backbone=dict(depth=50, with_cp=True))) + + >>> # Merge list element + >>> cfg = Config(dict(pipeline=[ + ... dict(type='LoadImage'), dict(type='LoadAnnotations')])) + >>> options = dict(pipeline={'0': dict(type='SelfLoadImage')}) + >>> cfg.merge_from_dict(options, allow_list_keys=True) + >>> cfg_dict = super(Config, self).__getattribute__('_cfg_dict') + >>> assert cfg_dict == dict(pipeline=[ + ... dict(type='SelfLoadImage'), dict(type='LoadAnnotations')]) + + Args: + options (dict): dict of configs to merge from. + allow_list_keys (bool): If True, int string keys (e.g. '0', '1') + are allowed in ``options`` and will replace the element of the + corresponding index in the config if the config is a list. + Default: True. + """ + option_cfg_dict = {} + for full_key, v in options.items(): + d = option_cfg_dict + key_list = full_key.split('.') + for subkey in key_list[:-1]: + d.setdefault(subkey, ConfigDict()) + d = d[subkey] + subkey = key_list[-1] + d[subkey] = v + + cfg_dict = super(Config, self).__getattribute__('_cfg_dict') + super(Config, self).__setattr__( + '_cfg_dict', + Config._merge_a_into_b( + option_cfg_dict, cfg_dict, allow_list_keys=allow_list_keys)) + + +class DictAction(Action): + """ + argparse action to split an argument into KEY=VALUE form + on the first = and append to a dictionary. List options can + be passed as comma separated values, i.e 'KEY=V1,V2,V3', or with explicit + brackets, i.e. 'KEY=[V1,V2,V3]'. It also support nested brackets to build + list/tuple values. e.g. 'KEY=[(V1,V2),(V3,V4)]' + """ + + @staticmethod + def _parse_int_float_bool(val): + try: + return int(val) + except ValueError: + pass + try: + return float(val) + except ValueError: + pass + if val.lower() in ['true', 'false']: + return True if val.lower() == 'true' else False + if val == 'None': + return None + return val + + @staticmethod + def _parse_iterable(val): + """Parse iterable values in the string. + + All elements inside '()' or '[]' are treated as iterable values. + + Args: + val (str): Value string. + + Returns: + list | tuple: The expanded list or tuple from the string. + + Examples: + >>> DictAction._parse_iterable('1,2,3') + [1, 2, 3] + >>> DictAction._parse_iterable('[a, b, c]') + ['a', 'b', 'c'] + >>> DictAction._parse_iterable('[(1, 2, 3), [a, b], c]') + [(1, 2, 3), ['a', 'b'], 'c'] + """ + + def find_next_comma(string): + """Find the position of next comma in the string. + + If no ',' is found in the string, return the string length. All + chars inside '()' and '[]' are treated as one element and thus ',' + inside these brackets are ignored. + """ + assert (string.count('(') == string.count(')')) and ( + string.count('[') == string.count(']')), \ + f'Imbalanced brackets exist in {string}' + end = len(string) + for idx, char in enumerate(string): + pre = string[:idx] + # The string before this ',' is balanced + if ((char == ',') and (pre.count('(') == pre.count(')')) + and (pre.count('[') == pre.count(']'))): + end = idx + break + return end + + # Strip ' and " characters and replace whitespace. + val = val.strip('\'\"').replace(' ', '') + is_tuple = False + if val.startswith('(') and val.endswith(')'): + is_tuple = True + val = val[1:-1] + elif val.startswith('[') and val.endswith(']'): + val = val[1:-1] + elif ',' not in val: + # val is a single value + return DictAction._parse_int_float_bool(val) + + values = [] + while len(val) > 0: + comma_idx = find_next_comma(val) + element = DictAction._parse_iterable(val[:comma_idx]) + values.append(element) + val = val[comma_idx + 1:] + if is_tuple: + values = tuple(values) + return values + + def __call__(self, parser, namespace, values, option_string=None): + options = {} + for kv in values: + key, val = kv.split('=', maxsplit=1) + options[key] = self._parse_iterable(val) + setattr(namespace, self.dest, options) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/device_type.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/device_type.py new file mode 100644 index 0000000000000000000000000000000000000000..c29d944ab1246a8d9711a12b17a73b4372bd4eaa --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/device_type.py @@ -0,0 +1,24 @@ +# Copyright (c) OpenMMLab. All rights reserved. + + +def is_ipu_available(): + try: + import poptorch + return poptorch.ipuHardwareIsAvailable() + except ImportError: + return False + + +IS_IPU_AVAILABLE = is_ipu_available() + + +def is_mlu_available(): + try: + import torch + return (hasattr(torch, 'is_mlu_available') + and torch.is_mlu_available()) + except Exception: + return False + + +IS_MLU_AVAILABLE = is_mlu_available() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/env.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/env.py new file mode 100644 index 0000000000000000000000000000000000000000..511332506f88774efee9c01b0236e70462af41f7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/env.py @@ -0,0 +1,120 @@ +# Copyright (c) OpenMMLab. All rights reserved. +"""This file holding some environment constant for sharing by other files.""" + +import os.path as osp +import subprocess +import sys +from collections import defaultdict + +import cv2 +import torch + +import mmcv +from .parrots_wrapper import get_build_config + + +def collect_env(): + """Collect the information of the running environments. + + Returns: + dict: The environment information. The following fields are contained. + + - sys.platform: The variable of ``sys.platform``. + - Python: Python version. + - CUDA available: Bool, indicating if CUDA is available. + - GPU devices: Device type of each GPU. + - CUDA_HOME (optional): The env var ``CUDA_HOME``. + - NVCC (optional): NVCC version. + - GCC: GCC version, "n/a" if GCC is not installed. + - MSVC: Microsoft Virtual C++ Compiler version, Windows only. + - PyTorch: PyTorch version. + - PyTorch compiling details: The output of \ + ``torch.__config__.show()``. + - TorchVision (optional): TorchVision version. + - OpenCV: OpenCV version. + - MMCV: MMCV version. + - MMCV Compiler: The GCC version for compiling MMCV ops. + - MMCV CUDA Compiler: The CUDA version for compiling MMCV ops. + """ + env_info = {} + env_info['sys.platform'] = sys.platform + env_info['Python'] = sys.version.replace('\n', '') + + cuda_available = torch.cuda.is_available() + env_info['CUDA available'] = cuda_available + + if cuda_available: + devices = defaultdict(list) + for k in range(torch.cuda.device_count()): + devices[torch.cuda.get_device_name(k)].append(str(k)) + for name, device_ids in devices.items(): + env_info['GPU ' + ','.join(device_ids)] = name + + from mmcv.utils.parrots_wrapper import _get_cuda_home + CUDA_HOME = _get_cuda_home() + env_info['CUDA_HOME'] = CUDA_HOME + + if CUDA_HOME is not None and osp.isdir(CUDA_HOME): + try: + nvcc = osp.join(CUDA_HOME, 'bin/nvcc') + nvcc = subprocess.check_output(f'"{nvcc}" -V', shell=True) + nvcc = nvcc.decode('utf-8').strip() + release = nvcc.rfind('Cuda compilation tools') + build = nvcc.rfind('Build ') + nvcc = nvcc[release:build].strip() + except subprocess.SubprocessError: + nvcc = 'Not Available' + env_info['NVCC'] = nvcc + + try: + # Check C++ Compiler. + # For Unix-like, sysconfig has 'CC' variable like 'gcc -pthread ...', + # indicating the compiler used, we use this to get the compiler name + import sysconfig + cc = sysconfig.get_config_var('CC') + if cc: + cc = osp.basename(cc.split()[0]) + cc_info = subprocess.check_output(f'{cc} --version', shell=True) + env_info['GCC'] = cc_info.decode('utf-8').partition( + '\n')[0].strip() + else: + # on Windows, cl.exe is not in PATH. We need to find the path. + # distutils.ccompiler.new_compiler() returns a msvccompiler + # object and after initialization, path to cl.exe is found. + import locale + import os + from distutils.ccompiler import new_compiler + ccompiler = new_compiler() + ccompiler.initialize() + cc = subprocess.check_output( + f'{ccompiler.cc}', stderr=subprocess.STDOUT, shell=True) + encoding = os.device_encoding( + sys.stdout.fileno()) or locale.getpreferredencoding() + env_info['MSVC'] = cc.decode(encoding).partition('\n')[0].strip() + env_info['GCC'] = 'n/a' + except subprocess.CalledProcessError: + env_info['GCC'] = 'n/a' + + env_info['PyTorch'] = torch.__version__ + env_info['PyTorch compiling details'] = get_build_config() + + try: + import torchvision + env_info['TorchVision'] = torchvision.__version__ + except ModuleNotFoundError: + pass + + env_info['OpenCV'] = cv2.__version__ + + env_info['MMCV'] = mmcv.__version__ + + try: + from mmcv.ops import get_compiler_version, get_compiling_cuda_version + except ModuleNotFoundError: + env_info['MMCV Compiler'] = 'n/a' + env_info['MMCV CUDA Compiler'] = 'n/a' + else: + env_info['MMCV Compiler'] = get_compiler_version() + env_info['MMCV CUDA Compiler'] = get_compiling_cuda_version() + + return env_info diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/ext_loader.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/ext_loader.py new file mode 100644 index 0000000000000000000000000000000000000000..f82c6d5681d863c48a56a8976826f4fcb918e825 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/ext_loader.py @@ -0,0 +1,72 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import importlib +import os +import pkgutil +import warnings +from collections import namedtuple + +import torch + +if torch.__version__ != 'parrots': + + def load_ext(name, funcs): + ext = importlib.import_module('mmcv.' + name) + for fun in funcs: + assert hasattr(ext, fun), f'{fun} miss in module {name}' + return ext +else: + from parrots import extension + from parrots.base import ParrotsException + + has_return_value_ops = [ + 'nms', + 'softnms', + 'nms_match', + 'nms_rotated', + 'top_pool_forward', + 'top_pool_backward', + 'bottom_pool_forward', + 'bottom_pool_backward', + 'left_pool_forward', + 'left_pool_backward', + 'right_pool_forward', + 'right_pool_backward', + 'fused_bias_leakyrelu', + 'upfirdn2d', + 'ms_deform_attn_forward', + 'pixel_group', + 'contour_expand', + 'diff_iou_rotated_sort_vertices_forward', + ] + + def get_fake_func(name, e): + + def fake_func(*args, **kwargs): + warnings.warn(f'{name} is not supported in parrots now') + raise e + + return fake_func + + def load_ext(name, funcs): + ExtModule = namedtuple('ExtModule', funcs) + ext_list = [] + lib_root = os.path.dirname(os.path.dirname(os.path.realpath(__file__))) + for fun in funcs: + try: + ext_fun = extension.load(fun, name, lib_dir=lib_root) + except ParrotsException as e: + if 'No element registered' not in e.message: + warnings.warn(e.message) + ext_fun = get_fake_func(fun, e) + ext_list.append(ext_fun) + else: + if fun in has_return_value_ops: + ext_list.append(ext_fun.op) + else: + ext_list.append(ext_fun.op_) + return ExtModule(*ext_list) + + +def check_ops_exist(): + ext_loader = pkgutil.find_loader('mmcv._ext') + return ext_loader is not None diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/hub.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/hub.py new file mode 100644 index 0000000000000000000000000000000000000000..12fbff2ee48217c9adeb50c1cb640bcb38907a9a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/hub.py @@ -0,0 +1,131 @@ +# Copyright (c) OpenMMLab. All rights reserved. +# The 1.6 release of PyTorch switched torch.save to use a new zipfile-based +# file format. It will cause RuntimeError when a checkpoint was saved in +# torch >= 1.6.0 but loaded in torch < 1.7.0. +# More details at https://github.com/open-mmlab/mmpose/issues/904 +from .parrots_wrapper import TORCH_VERSION +from .path import mkdir_or_exist +from .version_utils import digit_version + +if TORCH_VERSION != 'parrots' and digit_version(TORCH_VERSION) < digit_version( + '1.7.0'): + # Modified from https://github.com/pytorch/pytorch/blob/master/torch/hub.py + import os + import sys + import warnings + import zipfile + from urllib.parse import urlparse + + import torch + from torch.hub import HASH_REGEX, _get_torch_home, download_url_to_file + + # Hub used to support automatically extracts from zipfile manually + # compressed by users. The legacy zip format expects only one file from + # torch.save() < 1.6 in the zip. We should remove this support since + # zipfile is now default zipfile format for torch.save(). + def _is_legacy_zip_format(filename): + if zipfile.is_zipfile(filename): + infolist = zipfile.ZipFile(filename).infolist() + return len(infolist) == 1 and not infolist[0].is_dir() + return False + + def _legacy_zip_load(filename, model_dir, map_location): + warnings.warn( + 'Falling back to the old format < 1.6. This support will' + ' be deprecated in favor of default zipfile format ' + 'introduced in 1.6. Please redo torch.save() to save it ' + 'in the new zipfile format.', DeprecationWarning) + # Note: extractall() defaults to overwrite file if exists. No need to + # clean up beforehand. We deliberately don't handle tarfile here + # since our legacy serialization format was in tar. + # E.g. resnet18-5c106cde.pth which is widely used. + with zipfile.ZipFile(filename) as f: + members = f.infolist() + if len(members) != 1: + raise RuntimeError( + 'Only one file(not dir) is allowed in the zipfile') + f.extractall(model_dir) + extraced_name = members[0].filename + extracted_file = os.path.join(model_dir, extraced_name) + return torch.load(extracted_file, map_location=map_location) + + def load_url(url, + model_dir=None, + map_location=None, + progress=True, + check_hash=False, + file_name=None): + r"""Loads the Torch serialized object at the given URL. + + If downloaded file is a zip file, it will be automatically decompressed + + If the object is already present in `model_dir`, it's deserialized and + returned. + The default value of ``model_dir`` is ``/checkpoints`` where + ``hub_dir`` is the directory returned by :func:`~torch.hub.get_dir`. + + Args: + url (str): URL of the object to download + model_dir (str, optional): directory in which to save the object + map_location (optional): a function or a dict specifying how to + remap storage locations (see torch.load) + progress (bool, optional): whether or not to display a progress bar + to stderr. Default: True + check_hash(bool, optional): If True, the filename part of the URL + should follow the naming convention ``filename-.ext`` + where ```` is the first eight or more digits of the + SHA256 hash of the contents of the file. The hash is used to + ensure unique names and to verify the contents of the file. + Default: False + file_name (str, optional): name for the downloaded file. Filename + from ``url`` will be used if not set. Default: None. + + Example: + >>> url = ('https://s3.amazonaws.com/pytorch/models/resnet18-5c106' + ... 'cde.pth') + >>> state_dict = torch.hub.load_state_dict_from_url(url) + """ + # Issue warning to move data if old env is set + if os.getenv('TORCH_MODEL_ZOO'): + warnings.warn( + 'TORCH_MODEL_ZOO is deprecated, please use env ' + 'TORCH_HOME instead', DeprecationWarning) + + if model_dir is None: + torch_home = _get_torch_home() + model_dir = os.path.join(torch_home, 'checkpoints') + + mkdir_or_exist(model_dir) + + parts = urlparse(url) + filename = os.path.basename(parts.path) + if file_name is not None: + filename = file_name + cached_file = os.path.join(model_dir, filename) + if not os.path.exists(cached_file): + sys.stderr.write('Downloading: "{}" to {}\n'.format( + url, cached_file)) + hash_prefix = None + if check_hash: + r = HASH_REGEX.search(filename) # r is Optional[Match[str]] + hash_prefix = r.group(1) if r else None + download_url_to_file( + url, cached_file, hash_prefix, progress=progress) + + if _is_legacy_zip_format(cached_file): + return _legacy_zip_load(cached_file, model_dir, map_location) + + try: + return torch.load(cached_file, map_location=map_location) + except RuntimeError as error: + if digit_version(TORCH_VERSION) < digit_version('1.5.0'): + warnings.warn( + f'If the error is the same as "{cached_file} is a zip ' + 'archive (did you mean to use torch.jit.load()?)", you can' + ' upgrade your torch to 1.5.0 or higher (current torch ' + f'version is {TORCH_VERSION}). The error was raised ' + ' because the checkpoint was saved in torch>=1.6.0 but ' + 'loaded in torch<1.5.') + raise error +else: + from torch.utils.model_zoo import load_url # noqa: F401 diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/logging.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/logging.py new file mode 100644 index 0000000000000000000000000000000000000000..c4c7025f0e6fa9cbb2e1ecdef7ffb20ff933769e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/logging.py @@ -0,0 +1,111 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import logging + +import torch.distributed as dist + +logger_initialized = {} + + +def get_logger(name, log_file=None, log_level=logging.INFO, file_mode='w'): + """Initialize and get a logger by name. + + If the logger has not been initialized, this method will initialize the + logger by adding one or two handlers, otherwise the initialized logger will + be directly returned. During initialization, a StreamHandler will always be + added. If `log_file` is specified and the process rank is 0, a FileHandler + will also be added. + + Args: + name (str): Logger name. + log_file (str | None): The log filename. If specified, a FileHandler + will be added to the logger. + log_level (int): The logger level. Note that only the process of + rank 0 is affected, and other processes will set the level to + "Error" thus be silent most of the time. + file_mode (str): The file mode used in opening log file. + Defaults to 'w'. + + Returns: + logging.Logger: The expected logger. + """ + logger = logging.getLogger(name) + if name in logger_initialized: + return logger + # handle hierarchical names + # e.g., logger "a" is initialized, then logger "a.b" will skip the + # initialization since it is a child of "a". + for logger_name in logger_initialized: + if name.startswith(logger_name): + return logger + + # handle duplicate logs to the console + # Starting in 1.8.0, PyTorch DDP attaches a StreamHandler (NOTSET) + # to the root logger. As logger.propagate is True by default, this root + # level handler causes logging messages from rank>0 processes to + # unexpectedly show up on the console, creating much unwanted clutter. + # To fix this issue, we set the root logger's StreamHandler, if any, to log + # at the ERROR level. + for handler in logger.root.handlers: + if type(handler) is logging.StreamHandler: + handler.setLevel(logging.ERROR) + + stream_handler = logging.StreamHandler() + handlers = [stream_handler] + + if dist.is_available() and dist.is_initialized(): + rank = dist.get_rank() + else: + rank = 0 + + # only rank 0 will add a FileHandler + if rank == 0 and log_file is not None: + # Here, the default behaviour of the official logger is 'a'. Thus, we + # provide an interface to change the file mode to the default + # behaviour. + file_handler = logging.FileHandler(log_file, file_mode) + handlers.append(file_handler) + + formatter = logging.Formatter( + '%(asctime)s - %(name)s - %(levelname)s - %(message)s') + for handler in handlers: + handler.setFormatter(formatter) + handler.setLevel(log_level) + logger.addHandler(handler) + + if rank == 0: + logger.setLevel(log_level) + else: + logger.setLevel(logging.ERROR) + + logger_initialized[name] = True + + return logger + + +def print_log(msg, logger=None, level=logging.INFO): + """Print a log message. + + Args: + msg (str): The message to be logged. + logger (logging.Logger | str | None): The logger to be used. + Some special loggers are: + + - "silent": no message will be printed. + - other str: the logger obtained with `get_root_logger(logger)`. + - None: The `print()` method will be used to print log messages. + level (int): Logging level. Only available when `logger` is a Logger + object or "root". + """ + if logger is None: + print(msg) + elif isinstance(logger, logging.Logger): + logger.log(level, msg) + elif logger == 'silent': + pass + elif isinstance(logger, str): + _logger = get_logger(logger) + _logger.log(level, msg) + else: + raise TypeError( + 'logger should be either a logging.Logger object, str, ' + f'"silent" or None, but got {type(logger)}') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/misc.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/misc.py new file mode 100644 index 0000000000000000000000000000000000000000..7957ea89b762763566139edfbf0a75401dc4e268 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/misc.py @@ -0,0 +1,377 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import collections.abc +import functools +import itertools +import subprocess +import warnings +from collections import abc +from importlib import import_module +from inspect import getfullargspec +from itertools import repeat + + +# From PyTorch internals +def _ntuple(n): + + def parse(x): + if isinstance(x, collections.abc.Iterable): + return x + return tuple(repeat(x, n)) + + return parse + + +to_1tuple = _ntuple(1) +to_2tuple = _ntuple(2) +to_3tuple = _ntuple(3) +to_4tuple = _ntuple(4) +to_ntuple = _ntuple + + +def is_str(x): + """Whether the input is an string instance. + + Note: This method is deprecated since python 2 is no longer supported. + """ + return isinstance(x, str) + + +def import_modules_from_strings(imports, allow_failed_imports=False): + """Import modules from the given list of strings. + + Args: + imports (list | str | None): The given module names to be imported. + allow_failed_imports (bool): If True, the failed imports will return + None. Otherwise, an ImportError is raise. Default: False. + + Returns: + list[module] | module | None: The imported modules. + + Examples: + >>> osp, sys = import_modules_from_strings( + ... ['os.path', 'sys']) + >>> import os.path as osp_ + >>> import sys as sys_ + >>> assert osp == osp_ + >>> assert sys == sys_ + """ + if not imports: + return + single_import = False + if isinstance(imports, str): + single_import = True + imports = [imports] + if not isinstance(imports, list): + raise TypeError( + f'custom_imports must be a list but got type {type(imports)}') + imported = [] + for imp in imports: + if not isinstance(imp, str): + raise TypeError( + f'{imp} is of type {type(imp)} and cannot be imported.') + try: + imported_tmp = import_module(imp) + except ImportError: + if allow_failed_imports: + warnings.warn(f'{imp} failed to import and is ignored.', + UserWarning) + imported_tmp = None + else: + raise ImportError + imported.append(imported_tmp) + if single_import: + imported = imported[0] + return imported + + +def iter_cast(inputs, dst_type, return_type=None): + """Cast elements of an iterable object into some type. + + Args: + inputs (Iterable): The input object. + dst_type (type): Destination type. + return_type (type, optional): If specified, the output object will be + converted to this type, otherwise an iterator. + + Returns: + iterator or specified type: The converted object. + """ + if not isinstance(inputs, abc.Iterable): + raise TypeError('inputs must be an iterable object') + if not isinstance(dst_type, type): + raise TypeError('"dst_type" must be a valid type') + + out_iterable = map(dst_type, inputs) + + if return_type is None: + return out_iterable + else: + return return_type(out_iterable) + + +def list_cast(inputs, dst_type): + """Cast elements of an iterable object into a list of some type. + + A partial method of :func:`iter_cast`. + """ + return iter_cast(inputs, dst_type, return_type=list) + + +def tuple_cast(inputs, dst_type): + """Cast elements of an iterable object into a tuple of some type. + + A partial method of :func:`iter_cast`. + """ + return iter_cast(inputs, dst_type, return_type=tuple) + + +def is_seq_of(seq, expected_type, seq_type=None): + """Check whether it is a sequence of some type. + + Args: + seq (Sequence): The sequence to be checked. + expected_type (type): Expected type of sequence items. + seq_type (type, optional): Expected sequence type. + + Returns: + bool: Whether the sequence is valid. + """ + if seq_type is None: + exp_seq_type = abc.Sequence + else: + assert isinstance(seq_type, type) + exp_seq_type = seq_type + if not isinstance(seq, exp_seq_type): + return False + for item in seq: + if not isinstance(item, expected_type): + return False + return True + + +def is_list_of(seq, expected_type): + """Check whether it is a list of some type. + + A partial method of :func:`is_seq_of`. + """ + return is_seq_of(seq, expected_type, seq_type=list) + + +def is_tuple_of(seq, expected_type): + """Check whether it is a tuple of some type. + + A partial method of :func:`is_seq_of`. + """ + return is_seq_of(seq, expected_type, seq_type=tuple) + + +def slice_list(in_list, lens): + """Slice a list into several sub lists by a list of given length. + + Args: + in_list (list): The list to be sliced. + lens(int or list): The expected length of each out list. + + Returns: + list: A list of sliced list. + """ + if isinstance(lens, int): + assert len(in_list) % lens == 0 + lens = [lens] * int(len(in_list) / lens) + if not isinstance(lens, list): + raise TypeError('"indices" must be an integer or a list of integers') + elif sum(lens) != len(in_list): + raise ValueError('sum of lens and list length does not ' + f'match: {sum(lens)} != {len(in_list)}') + out_list = [] + idx = 0 + for i in range(len(lens)): + out_list.append(in_list[idx:idx + lens[i]]) + idx += lens[i] + return out_list + + +def concat_list(in_list): + """Concatenate a list of list into a single list. + + Args: + in_list (list): The list of list to be merged. + + Returns: + list: The concatenated flat list. + """ + return list(itertools.chain(*in_list)) + + +def check_prerequisites( + prerequisites, + checker, + msg_tmpl='Prerequisites "{}" are required in method "{}" but not ' + 'found, please install them first.'): # yapf: disable + """A decorator factory to check if prerequisites are satisfied. + + Args: + prerequisites (str of list[str]): Prerequisites to be checked. + checker (callable): The checker method that returns True if a + prerequisite is meet, False otherwise. + msg_tmpl (str): The message template with two variables. + + Returns: + decorator: A specific decorator. + """ + + def wrap(func): + + @functools.wraps(func) + def wrapped_func(*args, **kwargs): + requirements = [prerequisites] if isinstance( + prerequisites, str) else prerequisites + missing = [] + for item in requirements: + if not checker(item): + missing.append(item) + if missing: + print(msg_tmpl.format(', '.join(missing), func.__name__)) + raise RuntimeError('Prerequisites not meet.') + else: + return func(*args, **kwargs) + + return wrapped_func + + return wrap + + +def _check_py_package(package): + try: + import_module(package) + except ImportError: + return False + else: + return True + + +def _check_executable(cmd): + if subprocess.call(f'which {cmd}', shell=True) != 0: + return False + else: + return True + + +def requires_package(prerequisites): + """A decorator to check if some python packages are installed. + + Example: + >>> @requires_package('numpy') + >>> func(arg1, args): + >>> return numpy.zeros(1) + array([0.]) + >>> @requires_package(['numpy', 'non_package']) + >>> func(arg1, args): + >>> return numpy.zeros(1) + ImportError + """ + return check_prerequisites(prerequisites, checker=_check_py_package) + + +def requires_executable(prerequisites): + """A decorator to check if some executable files are installed. + + Example: + >>> @requires_executable('ffmpeg') + >>> func(arg1, args): + >>> print(1) + 1 + """ + return check_prerequisites(prerequisites, checker=_check_executable) + + +def deprecated_api_warning(name_dict, cls_name=None): + """A decorator to check if some arguments are deprecate and try to replace + deprecate src_arg_name to dst_arg_name. + + Args: + name_dict(dict): + key (str): Deprecate argument names. + val (str): Expected argument names. + + Returns: + func: New function. + """ + + def api_warning_wrapper(old_func): + + @functools.wraps(old_func) + def new_func(*args, **kwargs): + # get the arg spec of the decorated method + args_info = getfullargspec(old_func) + # get name of the function + func_name = old_func.__name__ + if cls_name is not None: + func_name = f'{cls_name}.{func_name}' + if args: + arg_names = args_info.args[:len(args)] + for src_arg_name, dst_arg_name in name_dict.items(): + if src_arg_name in arg_names: + warnings.warn( + f'"{src_arg_name}" is deprecated in ' + f'`{func_name}`, please use "{dst_arg_name}" ' + 'instead', DeprecationWarning) + arg_names[arg_names.index(src_arg_name)] = dst_arg_name + if kwargs: + for src_arg_name, dst_arg_name in name_dict.items(): + if src_arg_name in kwargs: + + assert dst_arg_name not in kwargs, ( + f'The expected behavior is to replace ' + f'the deprecated key `{src_arg_name}` to ' + f'new key `{dst_arg_name}`, but got them ' + f'in the arguments at the same time, which ' + f'is confusing. `{src_arg_name} will be ' + f'deprecated in the future, please ' + f'use `{dst_arg_name}` instead.') + + warnings.warn( + f'"{src_arg_name}" is deprecated in ' + f'`{func_name}`, please use "{dst_arg_name}" ' + 'instead', DeprecationWarning) + kwargs[dst_arg_name] = kwargs.pop(src_arg_name) + + # apply converted arguments to the decorated method + output = old_func(*args, **kwargs) + return output + + return new_func + + return api_warning_wrapper + + +def is_method_overridden(method, base_class, derived_class): + """Check if a method of base class is overridden in derived class. + + Args: + method (str): the method name to check. + base_class (type): the class of the base class. + derived_class (type | Any): the class or instance of the derived class. + """ + assert isinstance(base_class, type), \ + "base_class doesn't accept instance, Please pass class instead." + + if not isinstance(derived_class, type): + derived_class = derived_class.__class__ + + base_method = getattr(base_class, method) + derived_method = getattr(derived_class, method) + return derived_method != base_method + + +def has_method(obj: object, method: str) -> bool: + """Check whether the object has a method. + + Args: + method (str): The method name to check. + obj (object): The object to check. + + Returns: + bool: True if the object has the method else False. + """ + return hasattr(obj, method) and callable(getattr(obj, method)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/parrots_jit.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/parrots_jit.py new file mode 100644 index 0000000000000000000000000000000000000000..61873f6dbb9b10ed972c90aa8faa321e3cb3249e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/parrots_jit.py @@ -0,0 +1,41 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os + +from .parrots_wrapper import TORCH_VERSION + +parrots_jit_option = os.getenv('PARROTS_JIT_OPTION') + +if TORCH_VERSION == 'parrots' and parrots_jit_option == 'ON': + from parrots.jit import pat as jit +else: + + def jit(func=None, + check_input=None, + full_shape=True, + derivate=False, + coderize=False, + optimize=False): + + def wrapper(func): + + def wrapper_inner(*args, **kargs): + return func(*args, **kargs) + + return wrapper_inner + + if func is None: + return wrapper + else: + return func + + +if TORCH_VERSION == 'parrots': + from parrots.utils.tester import skip_no_elena +else: + + def skip_no_elena(func): + + def wrapper(*args, **kargs): + return func(*args, **kargs) + + return wrapper diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/parrots_wrapper.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/parrots_wrapper.py new file mode 100644 index 0000000000000000000000000000000000000000..7e657b5616ed2debc28041f063e45113623ca879 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/parrots_wrapper.py @@ -0,0 +1,114 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from functools import partial + +import torch + +TORCH_VERSION = torch.__version__ + + +def is_cuda_available() -> bool: + return torch.cuda.is_available() + + +IS_CUDA_AVAILABLE = is_cuda_available() + + +def is_rocm_pytorch() -> bool: + is_rocm = False + if TORCH_VERSION != 'parrots': + try: + from torch.utils.cpp_extension import ROCM_HOME + is_rocm = True if ((torch.version.hip is not None) and + (ROCM_HOME is not None)) else False + except ImportError: + pass + return is_rocm + + +def _get_cuda_home(): + if TORCH_VERSION == 'parrots': + from parrots.utils.build_extension import CUDA_HOME + else: + if is_rocm_pytorch(): + from torch.utils.cpp_extension import ROCM_HOME + CUDA_HOME = ROCM_HOME + else: + from torch.utils.cpp_extension import CUDA_HOME + return CUDA_HOME + + +def get_build_config(): + if TORCH_VERSION == 'parrots': + from parrots.config import get_build_info + return get_build_info() + else: + return torch.__config__.show() + + +def _get_conv(): + if TORCH_VERSION == 'parrots': + from parrots.nn.modules.conv import _ConvNd, _ConvTransposeMixin + else: + from torch.nn.modules.conv import _ConvNd, _ConvTransposeMixin + return _ConvNd, _ConvTransposeMixin + + +def _get_dataloader(): + if TORCH_VERSION == 'parrots': + from torch.utils.data import DataLoader, PoolDataLoader + else: + from torch.utils.data import DataLoader + PoolDataLoader = DataLoader + return DataLoader, PoolDataLoader + + +def _get_extension(): + if TORCH_VERSION == 'parrots': + from parrots.utils.build_extension import BuildExtension, Extension + CppExtension = partial(Extension, cuda=False) + CUDAExtension = partial(Extension, cuda=True) + else: + from torch.utils.cpp_extension import (BuildExtension, CppExtension, + CUDAExtension) + return BuildExtension, CppExtension, CUDAExtension + + +def _get_pool(): + if TORCH_VERSION == 'parrots': + from parrots.nn.modules.pool import (_AdaptiveAvgPoolNd, + _AdaptiveMaxPoolNd, _AvgPoolNd, + _MaxPoolNd) + else: + from torch.nn.modules.pooling import (_AdaptiveAvgPoolNd, + _AdaptiveMaxPoolNd, _AvgPoolNd, + _MaxPoolNd) + return _AdaptiveAvgPoolNd, _AdaptiveMaxPoolNd, _AvgPoolNd, _MaxPoolNd + + +def _get_norm(): + if TORCH_VERSION == 'parrots': + from parrots.nn.modules.batchnorm import _BatchNorm, _InstanceNorm + SyncBatchNorm_ = torch.nn.SyncBatchNorm2d + else: + from torch.nn.modules.batchnorm import _BatchNorm + from torch.nn.modules.instancenorm import _InstanceNorm + SyncBatchNorm_ = torch.nn.SyncBatchNorm + return _BatchNorm, _InstanceNorm, SyncBatchNorm_ + + +_ConvNd, _ConvTransposeMixin = _get_conv() +DataLoader, PoolDataLoader = _get_dataloader() +BuildExtension, CppExtension, CUDAExtension = _get_extension() +_BatchNorm, _InstanceNorm, SyncBatchNorm_ = _get_norm() +_AdaptiveAvgPoolNd, _AdaptiveMaxPoolNd, _AvgPoolNd, _MaxPoolNd = _get_pool() + + +class SyncBatchNorm(SyncBatchNorm_): + + def _check_input_dim(self, input): + if TORCH_VERSION == 'parrots': + if input.dim() < 2: + raise ValueError( + f'expected at least 2D input (got {input.dim()}D input)') + else: + super()._check_input_dim(input) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/path.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/path.py new file mode 100644 index 0000000000000000000000000000000000000000..56808183777d8070a94f8c346b7929da1f56ceb4 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/path.py @@ -0,0 +1,101 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os +import os.path as osp +from pathlib import Path + +from .misc import is_str + + +def is_filepath(x): + return is_str(x) or isinstance(x, Path) + + +def fopen(filepath, *args, **kwargs): + if is_str(filepath): + return open(filepath, *args, **kwargs) + elif isinstance(filepath, Path): + return filepath.open(*args, **kwargs) + raise ValueError('`filepath` should be a string or a Path') + + +def check_file_exist(filename, msg_tmpl='file "{}" does not exist'): + if not osp.isfile(filename): + raise FileNotFoundError(msg_tmpl.format(filename)) + + +def mkdir_or_exist(dir_name, mode=0o777): + if dir_name == '': + return + dir_name = osp.expanduser(dir_name) + os.makedirs(dir_name, mode=mode, exist_ok=True) + + +def symlink(src, dst, overwrite=True, **kwargs): + if os.path.lexists(dst) and overwrite: + os.remove(dst) + os.symlink(src, dst, **kwargs) + + +def scandir(dir_path, suffix=None, recursive=False, case_sensitive=True): + """Scan a directory to find the interested files. + + Args: + dir_path (str | :obj:`Path`): Path of the directory. + suffix (str | tuple(str), optional): File suffix that we are + interested in. Default: None. + recursive (bool, optional): If set to True, recursively scan the + directory. Default: False. + case_sensitive (bool, optional) : If set to False, ignore the case of + suffix. Default: True. + + Returns: + A generator for all the interested files with relative paths. + """ + if isinstance(dir_path, (str, Path)): + dir_path = str(dir_path) + else: + raise TypeError('"dir_path" must be a string or Path object') + + if (suffix is not None) and not isinstance(suffix, (str, tuple)): + raise TypeError('"suffix" must be a string or tuple of strings') + + if suffix is not None and not case_sensitive: + suffix = suffix.lower() if isinstance(suffix, str) else tuple( + item.lower() for item in suffix) + + root = dir_path + + def _scandir(dir_path, suffix, recursive, case_sensitive): + for entry in os.scandir(dir_path): + if not entry.name.startswith('.') and entry.is_file(): + rel_path = osp.relpath(entry.path, root) + _rel_path = rel_path if case_sensitive else rel_path.lower() + if suffix is None or _rel_path.endswith(suffix): + yield rel_path + elif recursive and os.path.isdir(entry.path): + # scan recursively if entry.path is a directory + yield from _scandir(entry.path, suffix, recursive, + case_sensitive) + + return _scandir(dir_path, suffix, recursive, case_sensitive) + + +def find_vcs_root(path, markers=('.git', )): + """Finds the root directory (including itself) of specified markers. + + Args: + path (str): Path of directory or file. + markers (list[str], optional): List of file or directory names. + + Returns: + The directory contained one of the markers or None if not found. + """ + if osp.isfile(path): + path = osp.dirname(path) + + prev, cur = None, osp.abspath(osp.expanduser(path)) + while cur != prev: + if any(osp.exists(osp.join(cur, marker)) for marker in markers): + return cur + prev, cur = cur, osp.split(cur)[0] + return None diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/progressbar.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/progressbar.py new file mode 100644 index 0000000000000000000000000000000000000000..0062f670dd94fa9da559ab26ef85517dcf5211c7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/progressbar.py @@ -0,0 +1,208 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import sys +from collections.abc import Iterable +from multiprocessing import Pool +from shutil import get_terminal_size + +from .timer import Timer + + +class ProgressBar: + """A progress bar which can print the progress.""" + + def __init__(self, task_num=0, bar_width=50, start=True, file=sys.stdout): + self.task_num = task_num + self.bar_width = bar_width + self.completed = 0 + self.file = file + if start: + self.start() + + @property + def terminal_width(self): + width, _ = get_terminal_size() + return width + + def start(self): + if self.task_num > 0: + self.file.write(f'[{" " * self.bar_width}] 0/{self.task_num}, ' + 'elapsed: 0s, ETA:') + else: + self.file.write('completed: 0, elapsed: 0s') + self.file.flush() + self.timer = Timer() + + def update(self, num_tasks=1): + assert num_tasks > 0 + self.completed += num_tasks + elapsed = self.timer.since_start() + if elapsed > 0: + fps = self.completed / elapsed + else: + fps = float('inf') + if self.task_num > 0: + percentage = self.completed / float(self.task_num) + eta = int(elapsed * (1 - percentage) / percentage + 0.5) + msg = f'\r[{{}}] {self.completed}/{self.task_num}, ' \ + f'{fps:.1f} task/s, elapsed: {int(elapsed + 0.5)}s, ' \ + f'ETA: {eta:5}s' + + bar_width = min(self.bar_width, + int(self.terminal_width - len(msg)) + 2, + int(self.terminal_width * 0.6)) + bar_width = max(2, bar_width) + mark_width = int(bar_width * percentage) + bar_chars = '>' * mark_width + ' ' * (bar_width - mark_width) + self.file.write(msg.format(bar_chars)) + else: + self.file.write( + f'completed: {self.completed}, elapsed: {int(elapsed + 0.5)}s,' + f' {fps:.1f} tasks/s') + self.file.flush() + + +def track_progress(func, tasks, bar_width=50, file=sys.stdout, **kwargs): + """Track the progress of tasks execution with a progress bar. + + Tasks are done with a simple for-loop. + + Args: + func (callable): The function to be applied to each task. + tasks (list or tuple[Iterable, int]): A list of tasks or + (tasks, total num). + bar_width (int): Width of progress bar. + + Returns: + list: The task results. + """ + if isinstance(tasks, tuple): + assert len(tasks) == 2 + assert isinstance(tasks[0], Iterable) + assert isinstance(tasks[1], int) + task_num = tasks[1] + tasks = tasks[0] + elif isinstance(tasks, Iterable): + task_num = len(tasks) + else: + raise TypeError( + '"tasks" must be an iterable object or a (iterator, int) tuple') + prog_bar = ProgressBar(task_num, bar_width, file=file) + results = [] + for task in tasks: + results.append(func(task, **kwargs)) + prog_bar.update() + prog_bar.file.write('\n') + return results + + +def init_pool(process_num, initializer=None, initargs=None): + if initializer is None: + return Pool(process_num) + elif initargs is None: + return Pool(process_num, initializer) + else: + if not isinstance(initargs, tuple): + raise TypeError('"initargs" must be a tuple') + return Pool(process_num, initializer, initargs) + + +def track_parallel_progress(func, + tasks, + nproc, + initializer=None, + initargs=None, + bar_width=50, + chunksize=1, + skip_first=False, + keep_order=True, + file=sys.stdout): + """Track the progress of parallel task execution with a progress bar. + + The built-in :mod:`multiprocessing` module is used for process pools and + tasks are done with :func:`Pool.map` or :func:`Pool.imap_unordered`. + + Args: + func (callable): The function to be applied to each task. + tasks (list or tuple[Iterable, int]): A list of tasks or + (tasks, total num). + nproc (int): Process (worker) number. + initializer (None or callable): Refer to :class:`multiprocessing.Pool` + for details. + initargs (None or tuple): Refer to :class:`multiprocessing.Pool` for + details. + chunksize (int): Refer to :class:`multiprocessing.Pool` for details. + bar_width (int): Width of progress bar. + skip_first (bool): Whether to skip the first sample for each worker + when estimating fps, since the initialization step may takes + longer. + keep_order (bool): If True, :func:`Pool.imap` is used, otherwise + :func:`Pool.imap_unordered` is used. + + Returns: + list: The task results. + """ + if isinstance(tasks, tuple): + assert len(tasks) == 2 + assert isinstance(tasks[0], Iterable) + assert isinstance(tasks[1], int) + task_num = tasks[1] + tasks = tasks[0] + elif isinstance(tasks, Iterable): + task_num = len(tasks) + else: + raise TypeError( + '"tasks" must be an iterable object or a (iterator, int) tuple') + pool = init_pool(nproc, initializer, initargs) + start = not skip_first + task_num -= nproc * chunksize * int(skip_first) + prog_bar = ProgressBar(task_num, bar_width, start, file=file) + results = [] + if keep_order: + gen = pool.imap(func, tasks, chunksize) + else: + gen = pool.imap_unordered(func, tasks, chunksize) + for result in gen: + results.append(result) + if skip_first: + if len(results) < nproc * chunksize: + continue + elif len(results) == nproc * chunksize: + prog_bar.start() + continue + prog_bar.update() + prog_bar.file.write('\n') + pool.close() + pool.join() + return results + + +def track_iter_progress(tasks, bar_width=50, file=sys.stdout): + """Track the progress of tasks iteration or enumeration with a progress + bar. + + Tasks are yielded with a simple for-loop. + + Args: + tasks (list or tuple[Iterable, int]): A list of tasks or + (tasks, total num). + bar_width (int): Width of progress bar. + + Yields: + list: The task results. + """ + if isinstance(tasks, tuple): + assert len(tasks) == 2 + assert isinstance(tasks[0], Iterable) + assert isinstance(tasks[1], int) + task_num = tasks[1] + tasks = tasks[0] + elif isinstance(tasks, Iterable): + task_num = len(tasks) + else: + raise TypeError( + '"tasks" must be an iterable object or a (iterator, int) tuple') + prog_bar = ProgressBar(task_num, bar_width, file=file) + for task in tasks: + yield task + prog_bar.update() + prog_bar.file.write('\n') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/registry.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/registry.py new file mode 100644 index 0000000000000000000000000000000000000000..a6d92b68bc1cb170335a0b5efd7c23648e5bdd98 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/registry.py @@ -0,0 +1,337 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import inspect +import warnings +from functools import partial + +from .misc import deprecated_api_warning, is_seq_of + + +def build_from_cfg(cfg, registry, default_args=None): + """Build a module from config dict when it is a class configuration, or + call a function from config dict when it is a function configuration. + + Example: + >>> MODELS = Registry('models') + >>> @MODELS.register_module() + >>> class ResNet: + >>> pass + >>> resnet = build_from_cfg(dict(type='Resnet'), MODELS) + >>> # Returns an instantiated object + >>> @MODELS.register_module() + >>> def resnet50(): + >>> pass + >>> resnet = build_from_cfg(dict(type='resnet50'), MODELS) + >>> # Return a result of the calling function + + Args: + cfg (dict): Config dict. It should at least contain the key "type". + registry (:obj:`Registry`): The registry to search the type from. + default_args (dict, optional): Default initialization arguments. + + Returns: + object: The constructed object. + """ + if not isinstance(cfg, dict): + raise TypeError(f'cfg must be a dict, but got {type(cfg)}') + if 'type' not in cfg: + if default_args is None or 'type' not in default_args: + raise KeyError( + '`cfg` or `default_args` must contain the key "type", ' + f'but got {cfg}\n{default_args}') + if not isinstance(registry, Registry): + raise TypeError('registry must be an mmcv.Registry object, ' + f'but got {type(registry)}') + if not (isinstance(default_args, dict) or default_args is None): + raise TypeError('default_args must be a dict or None, ' + f'but got {type(default_args)}') + + args = cfg.copy() + + if default_args is not None: + for name, value in default_args.items(): + args.setdefault(name, value) + + obj_type = args.pop('type') + if isinstance(obj_type, str): + obj_cls = registry.get(obj_type) + if obj_cls is None: + raise KeyError( + f'{obj_type} is not in the {registry.name} registry') + elif inspect.isclass(obj_type) or inspect.isfunction(obj_type): + obj_cls = obj_type + else: + raise TypeError( + f'type must be a str or valid type, but got {type(obj_type)}') + try: + return obj_cls(**args) + except Exception as e: + # Normal TypeError does not print class name. + raise type(e)(f'{obj_cls.__name__}: {e}') + + +class Registry: + """A registry to map strings to classes or functions. + + Registered object could be built from registry. Meanwhile, registered + functions could be called from registry. + + Example: + >>> MODELS = Registry('models') + >>> @MODELS.register_module() + >>> class ResNet: + >>> pass + >>> resnet = MODELS.build(dict(type='ResNet')) + >>> @MODELS.register_module() + >>> def resnet50(): + >>> pass + >>> resnet = MODELS.build(dict(type='resnet50')) + + Please refer to + https://mmcv.readthedocs.io/en/latest/understand_mmcv/registry.html for + advanced usage. + + Args: + name (str): Registry name. + build_func(func, optional): Build function to construct instance from + Registry, func:`build_from_cfg` is used if neither ``parent`` or + ``build_func`` is specified. If ``parent`` is specified and + ``build_func`` is not given, ``build_func`` will be inherited + from ``parent``. Default: None. + parent (Registry, optional): Parent registry. The class registered in + children registry could be built from parent. Default: None. + scope (str, optional): The scope of registry. It is the key to search + for children registry. If not specified, scope will be the name of + the package where class is defined, e.g. mmdet, mmcls, mmseg. + Default: None. + """ + + def __init__(self, name, build_func=None, parent=None, scope=None): + self._name = name + self._module_dict = dict() + self._children = dict() + self._scope = self.infer_scope() if scope is None else scope + + # self.build_func will be set with the following priority: + # 1. build_func + # 2. parent.build_func + # 3. build_from_cfg + if build_func is None: + if parent is not None: + self.build_func = parent.build_func + else: + self.build_func = build_from_cfg + else: + self.build_func = build_func + if parent is not None: + assert isinstance(parent, Registry) + parent._add_children(self) + self.parent = parent + else: + self.parent = None + + def __len__(self): + return len(self._module_dict) + + def __contains__(self, key): + return self.get(key) is not None + + def __repr__(self): + format_str = self.__class__.__name__ + \ + f'(name={self._name}, ' \ + f'items={self._module_dict})' + return format_str + + @staticmethod + def infer_scope(): + """Infer the scope of registry. + + The name of the package where registry is defined will be returned. + + Example: + >>> # in mmdet/models/backbone/resnet.py + >>> MODELS = Registry('models') + >>> @MODELS.register_module() + >>> class ResNet: + >>> pass + The scope of ``ResNet`` will be ``mmdet``. + + Returns: + str: The inferred scope name. + """ + # We access the caller using inspect.currentframe() instead of + # inspect.stack() for performance reasons. See details in PR #1844 + frame = inspect.currentframe() + # get the frame where `infer_scope()` is called + infer_scope_caller = frame.f_back.f_back + filename = inspect.getmodule(infer_scope_caller).__name__ + split_filename = filename.split('.') + return split_filename[0] + + @staticmethod + def split_scope_key(key): + """Split scope and key. + + The first scope will be split from key. + + Examples: + >>> Registry.split_scope_key('mmdet.ResNet') + 'mmdet', 'ResNet' + >>> Registry.split_scope_key('ResNet') + None, 'ResNet' + + Return: + tuple[str | None, str]: The former element is the first scope of + the key, which can be ``None``. The latter is the remaining key. + """ + split_index = key.find('.') + if split_index != -1: + return key[:split_index], key[split_index + 1:] + else: + return None, key + + @property + def name(self): + return self._name + + @property + def scope(self): + return self._scope + + @property + def module_dict(self): + return self._module_dict + + @property + def children(self): + return self._children + + def get(self, key): + """Get the registry record. + + Args: + key (str): The class name in string format. + + Returns: + class: The corresponding class. + """ + scope, real_key = self.split_scope_key(key) + if scope is None or scope == self._scope: + # get from self + if real_key in self._module_dict: + return self._module_dict[real_key] + else: + # get from self._children + if scope in self._children: + return self._children[scope].get(real_key) + else: + # goto root + parent = self.parent + while parent.parent is not None: + parent = parent.parent + return parent.get(key) + + def build(self, *args, **kwargs): + return self.build_func(*args, **kwargs, registry=self) + + def _add_children(self, registry): + """Add children for a registry. + + The ``registry`` will be added as children based on its scope. + The parent registry could build objects from children registry. + + Example: + >>> models = Registry('models') + >>> mmdet_models = Registry('models', parent=models) + >>> @mmdet_models.register_module() + >>> class ResNet: + >>> pass + >>> resnet = models.build(dict(type='mmdet.ResNet')) + """ + + assert isinstance(registry, Registry) + assert registry.scope is not None + assert registry.scope not in self.children, \ + f'scope {registry.scope} exists in {self.name} registry' + self.children[registry.scope] = registry + + @deprecated_api_warning(name_dict=dict(module_class='module')) + def _register_module(self, module, module_name=None, force=False): + if not inspect.isclass(module) and not inspect.isfunction(module): + raise TypeError('module must be a class or a function, ' + f'but got {type(module)}') + + if module_name is None: + module_name = module.__name__ + if isinstance(module_name, str): + module_name = [module_name] + for name in module_name: + if not force and name in self._module_dict: + raise KeyError(f'{name} is already registered ' + f'in {self.name}') + self._module_dict[name] = module + + def deprecated_register_module(self, cls=None, force=False): + warnings.warn( + 'The old API of register_module(module, force=False) ' + 'is deprecated and will be removed, please use the new API ' + 'register_module(name=None, force=False, module=None) instead.', + DeprecationWarning) + if cls is None: + return partial(self.deprecated_register_module, force=force) + self._register_module(cls, force=force) + return cls + + def register_module(self, name=None, force=False, module=None): + """Register a module. + + A record will be added to `self._module_dict`, whose key is the class + name or the specified name, and value is the class itself. + It can be used as a decorator or a normal function. + + Example: + >>> backbones = Registry('backbone') + >>> @backbones.register_module() + >>> class ResNet: + >>> pass + + >>> backbones = Registry('backbone') + >>> @backbones.register_module(name='mnet') + >>> class MobileNet: + >>> pass + + >>> backbones = Registry('backbone') + >>> class ResNet: + >>> pass + >>> backbones.register_module(ResNet) + + Args: + name (str | None): The module name to be registered. If not + specified, the class name will be used. + force (bool, optional): Whether to override an existing class with + the same name. Default: False. + module (type): Module class or function to be registered. + """ + if not isinstance(force, bool): + raise TypeError(f'force must be a boolean, but got {type(force)}') + # NOTE: This is a walkaround to be compatible with the old api, + # while it may introduce unexpected bugs. + if isinstance(name, type): + return self.deprecated_register_module(name, force=force) + + # raise the error ahead of time + if not (name is None or isinstance(name, str) or is_seq_of(name, str)): + raise TypeError( + 'name must be either of None, an instance of str or a sequence' + f' of str, but got {type(name)}') + + # use it as a normal method: x.register_module(module=SomeClass) + if module is not None: + self._register_module(module=module, module_name=name, force=force) + return module + + # use it as a decorator: @x.register_module() + def _register(module): + self._register_module(module=module, module_name=name, force=force) + return module + + return _register diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/seed.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/seed.py new file mode 100644 index 0000000000000000000000000000000000000000..003f9236774165af2de921af3c06f9fe057725dc --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/seed.py @@ -0,0 +1,23 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import random + +import numpy as np +import torch + + +def worker_init_fn(worker_id: int, num_workers: int, rank: int, seed: int): + """Function to initialize each worker. + + The seed of each worker equals to + ``num_worker * rank + worker_id + user_seed``. + + Args: + worker_id (int): Id for each worker. + num_workers (int): Number of workers. + rank (int): Rank in distributed training. + seed (int): Random seed. + """ + worker_seed = num_workers * rank + worker_id + seed + np.random.seed(worker_seed) + random.seed(worker_seed) + torch.manual_seed(worker_seed) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/testing.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/testing.py new file mode 100644 index 0000000000000000000000000000000000000000..7b64e8fae39022fece6f5910cc6656598f31bff5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/testing.py @@ -0,0 +1,141 @@ +# Copyright (c) Open-MMLab. +import sys +from collections.abc import Iterable +from runpy import run_path +from shlex import split +from typing import Any, Dict, List +from unittest.mock import patch + + +def check_python_script(cmd): + """Run the python cmd script with `__main__`. The difference between + `os.system` is that, this function exectues code in the current process, so + that it can be tracked by coverage tools. Currently it supports two forms: + + - ./tests/data/scripts/hello.py zz + - python tests/data/scripts/hello.py zz + """ + args = split(cmd) + if args[0] == 'python': + args = args[1:] + with patch.object(sys, 'argv', args): + run_path(args[0], run_name='__main__') + + +def _any(judge_result): + """Since built-in ``any`` works only when the element of iterable is not + iterable, implement the function.""" + if not isinstance(judge_result, Iterable): + return judge_result + + try: + for element in judge_result: + if _any(element): + return True + except TypeError: + # Maybe encounter the case: torch.tensor(True) | torch.tensor(False) + if judge_result: + return True + return False + + +def assert_dict_contains_subset(dict_obj: Dict[Any, Any], + expected_subset: Dict[Any, Any]) -> bool: + """Check if the dict_obj contains the expected_subset. + + Args: + dict_obj (Dict[Any, Any]): Dict object to be checked. + expected_subset (Dict[Any, Any]): Subset expected to be contained in + dict_obj. + + Returns: + bool: Whether the dict_obj contains the expected_subset. + """ + + for key, value in expected_subset.items(): + if key not in dict_obj.keys() or _any(dict_obj[key] != value): + return False + return True + + +def assert_attrs_equal(obj: Any, expected_attrs: Dict[str, Any]) -> bool: + """Check if attribute of class object is correct. + + Args: + obj (object): Class object to be checked. + expected_attrs (Dict[str, Any]): Dict of the expected attrs. + + Returns: + bool: Whether the attribute of class object is correct. + """ + for attr, value in expected_attrs.items(): + if not hasattr(obj, attr) or _any(getattr(obj, attr) != value): + return False + return True + + +def assert_dict_has_keys(obj: Dict[str, Any], + expected_keys: List[str]) -> bool: + """Check if the obj has all the expected_keys. + + Args: + obj (Dict[str, Any]): Object to be checked. + expected_keys (List[str]): Keys expected to contained in the keys of + the obj. + + Returns: + bool: Whether the obj has the expected keys. + """ + return set(expected_keys).issubset(set(obj.keys())) + + +def assert_keys_equal(result_keys: List[str], target_keys: List[str]) -> bool: + """Check if target_keys is equal to result_keys. + + Args: + result_keys (List[str]): Result keys to be checked. + target_keys (List[str]): Target keys to be checked. + + Returns: + bool: Whether target_keys is equal to result_keys. + """ + return set(result_keys) == set(target_keys) + + +def assert_is_norm_layer(module) -> bool: + """Check if the module is a norm layer. + + Args: + module (nn.Module): The module to be checked. + + Returns: + bool: Whether the module is a norm layer. + """ + from torch.nn import GroupNorm, LayerNorm + + from .parrots_wrapper import _BatchNorm, _InstanceNorm + norm_layer_candidates = (_BatchNorm, _InstanceNorm, GroupNorm, LayerNorm) + return isinstance(module, norm_layer_candidates) + + +def assert_params_all_zeros(module) -> bool: + """Check if the parameters of the module is all zeros. + + Args: + module (nn.Module): The module to be checked. + + Returns: + bool: Whether the parameters of the module is all zeros. + """ + weight_data = module.weight.data + is_weight_zero = weight_data.allclose( + weight_data.new_zeros(weight_data.size())) + + if hasattr(module, 'bias') and module.bias is not None: + bias_data = module.bias.data + is_bias_zero = bias_data.allclose( + bias_data.new_zeros(bias_data.size())) + else: + is_bias_zero = True + + return is_weight_zero and is_bias_zero diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/timer.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/timer.py new file mode 100644 index 0000000000000000000000000000000000000000..02e96e5379c3b9e0019d9e34173eb8628587f902 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/timer.py @@ -0,0 +1,118 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from time import time + + +class TimerError(Exception): + + def __init__(self, message): + self.message = message + super(TimerError, self).__init__(message) + + +class Timer: + """A flexible Timer class. + + Examples: + >>> import time + >>> import mmcv + >>> with mmcv.Timer(): + >>> # simulate a code block that will run for 1s + >>> time.sleep(1) + 1.000 + >>> with mmcv.Timer(print_tmpl='it takes {:.1f} seconds'): + >>> # simulate a code block that will run for 1s + >>> time.sleep(1) + it takes 1.0 seconds + >>> timer = mmcv.Timer() + >>> time.sleep(0.5) + >>> print(timer.since_start()) + 0.500 + >>> time.sleep(0.5) + >>> print(timer.since_last_check()) + 0.500 + >>> print(timer.since_start()) + 1.000 + """ + + def __init__(self, start=True, print_tmpl=None): + self._is_running = False + self.print_tmpl = print_tmpl if print_tmpl else '{:.3f}' + if start: + self.start() + + @property + def is_running(self): + """bool: indicate whether the timer is running""" + return self._is_running + + def __enter__(self): + self.start() + return self + + def __exit__(self, type, value, traceback): + print(self.print_tmpl.format(self.since_last_check())) + self._is_running = False + + def start(self): + """Start the timer.""" + if not self._is_running: + self._t_start = time() + self._is_running = True + self._t_last = time() + + def since_start(self): + """Total time since the timer is started. + + Returns: + float: Time in seconds. + """ + if not self._is_running: + raise TimerError('timer is not running') + self._t_last = time() + return self._t_last - self._t_start + + def since_last_check(self): + """Time since the last checking. + + Either :func:`since_start` or :func:`since_last_check` is a checking + operation. + + Returns: + float: Time in seconds. + """ + if not self._is_running: + raise TimerError('timer is not running') + dur = time() - self._t_last + self._t_last = time() + return dur + + +_g_timers = {} # global timers + + +def check_time(timer_id): + """Add check points in a single line. + + This method is suitable for running a task on a list of items. A timer will + be registered when the method is called for the first time. + + Examples: + >>> import time + >>> import mmcv + >>> for i in range(1, 6): + >>> # simulate a code block + >>> time.sleep(i) + >>> mmcv.check_time('task1') + 2.000 + 3.000 + 4.000 + 5.000 + + Args: + str: Timer identifier. + """ + if timer_id not in _g_timers: + _g_timers[timer_id] = Timer() + return 0 + else: + return _g_timers[timer_id].since_last_check() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/trace.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/trace.py new file mode 100644 index 0000000000000000000000000000000000000000..45423bd0551b8c4824193110546d5328ea4253d1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/trace.py @@ -0,0 +1,24 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch + +from mmcv.utils import digit_version + + +def is_jit_tracing() -> bool: + if (torch.__version__ != 'parrots' + and digit_version(torch.__version__) >= digit_version('1.6.0')): + on_trace = torch.jit.is_tracing() + # In PyTorch 1.6, torch.jit.is_tracing has a bug. + # Refers to https://github.com/pytorch/pytorch/issues/42448 + if isinstance(on_trace, bool): + return on_trace + else: + return torch._C._is_tracing() + else: + warnings.warn( + 'torch.jit.is_tracing is only supported after v1.6.0. ' + 'Therefore is_tracing returns False automatically. Please ' + 'set on_trace manually if you are using trace.', UserWarning) + return False diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/version_utils.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/version_utils.py new file mode 100644 index 0000000000000000000000000000000000000000..963c45a2e8a86a88413ab6c18c22481fb9831985 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/utils/version_utils.py @@ -0,0 +1,90 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os +import subprocess +import warnings + +from packaging.version import parse + + +def digit_version(version_str: str, length: int = 4): + """Convert a version string into a tuple of integers. + + This method is usually used for comparing two versions. For pre-release + versions: alpha < beta < rc. + + Args: + version_str (str): The version string. + length (int): The maximum number of version levels. Default: 4. + + Returns: + tuple[int]: The version info in digits (integers). + """ + assert 'parrots' not in version_str + version = parse(version_str) + assert version.release, f'failed to parse version {version_str}' + release = list(version.release) + release = release[:length] + if len(release) < length: + release = release + [0] * (length - len(release)) + if version.is_prerelease: + mapping = {'a': -3, 'b': -2, 'rc': -1} + val = -4 + # version.pre can be None + if version.pre: + if version.pre[0] not in mapping: + warnings.warn(f'unknown prerelease version {version.pre[0]}, ' + 'version checking may go wrong') + else: + val = mapping[version.pre[0]] + release.extend([val, version.pre[-1]]) + else: + release.extend([val, 0]) + + elif version.is_postrelease: + release.extend([1, version.post]) + else: + release.extend([0, 0]) + return tuple(release) + + +def _minimal_ext_cmd(cmd): + # construct minimal environment + env = {} + for k in ['SYSTEMROOT', 'PATH', 'HOME']: + v = os.environ.get(k) + if v is not None: + env[k] = v + # LANGUAGE is used on win32 + env['LANGUAGE'] = 'C' + env['LANG'] = 'C' + env['LC_ALL'] = 'C' + out = subprocess.Popen( + cmd, stdout=subprocess.PIPE, env=env).communicate()[0] + return out + + +def get_git_hash(fallback='unknown', digits=None): + """Get the git hash of the current repo. + + Args: + fallback (str, optional): The fallback string when git hash is + unavailable. Defaults to 'unknown'. + digits (int, optional): kept digits of the hash. Defaults to None, + meaning all digits are kept. + + Returns: + str: Git commit hash. + """ + + if digits is not None and not isinstance(digits, int): + raise TypeError('digits must be None or an integer') + + try: + out = _minimal_ext_cmd(['git', 'rev-parse', 'HEAD']) + sha = out.strip().decode('ascii') + if digits is not None: + sha = sha[:digits] + except OSError: + sha = fallback + + return sha diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmcv/version.py b/cv/semantic_segmentation/att_unet/pytorch/mmcv/version.py new file mode 100644 index 0000000000000000000000000000000000000000..a97ffc2dd280f931f1e72825aaef160cdc1ba155 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmcv/version.py @@ -0,0 +1,35 @@ +# Copyright (c) OpenMMLab. All rights reserved. +__version__ = '1.5.0' + + +def parse_version_info(version_str: str, length: int = 4) -> tuple: + """Parse a version string into a tuple. + + Args: + version_str (str): The version string. + length (int): The maximum number of version levels. Default: 4. + + Returns: + tuple[int | str]: The version info, e.g., "1.3.0" is parsed into + (1, 3, 0, 0, 0, 0), and "2.0.0rc1" is parsed into + (2, 0, 0, 0, 'rc', 1) (when length is set to 4). + """ + from packaging.version import parse + version = parse(version_str) + assert version.release, f'failed to parse version {version_str}' + release = list(version.release) + release = release[:length] + if len(release) < length: + release = release + [0] * (length - len(release)) + if version.is_prerelease: + release.extend(list(version.pre)) + elif version.is_postrelease: + release.extend(list(version.post)) + else: + release.extend([0, 0]) + return tuple(release) + + +version_info = tuple(int(x) for x in __version__.split('.')[:3]) + +__all__ = ['__version__', 'version_info', 'parse_version_info'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..360abfc85761f3144de88422a45c858d602f38d2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/__init__.py @@ -0,0 +1,62 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import mmcv +from packaging.version import parse + +from .version import __version__, version_info + +MMCV_MIN = '1.3.13' +MMCV_MAX = '1.6.0' + + +def digit_version(version_str: str, length: int = 4): + """Convert a version string into a tuple of integers. + + This method is usually used for comparing two versions. For pre-release + versions: alpha < beta < rc. + + Args: + version_str (str): The version string. + length (int): The maximum number of version levels. Default: 4. + + Returns: + tuple[int]: The version info in digits (integers). + """ + version = parse(version_str) + assert version.release, f'failed to parse version {version_str}' + release = list(version.release) + release = release[:length] + if len(release) < length: + release = release + [0] * (length - len(release)) + if version.is_prerelease: + mapping = {'a': -3, 'b': -2, 'rc': -1} + val = -4 + # version.pre can be None + if version.pre: + if version.pre[0] not in mapping: + warnings.warn(f'unknown prerelease version {version.pre[0]}, ' + 'version checking may go wrong') + else: + val = mapping[version.pre[0]] + release.extend([val, version.pre[-1]]) + else: + release.extend([val, 0]) + + elif version.is_postrelease: + release.extend([1, version.post]) + else: + release.extend([0, 0]) + return tuple(release) + + +mmcv_min_version = digit_version(MMCV_MIN) +mmcv_max_version = digit_version(MMCV_MAX) +mmcv_version = digit_version(mmcv.__version__) + + +assert (mmcv_min_version <= mmcv_version <= mmcv_max_version), \ + f'MMCV=={mmcv.__version__} is used but incompatible. ' \ + f'Please install mmcv>={mmcv_min_version}, <={mmcv_max_version}.' + +__all__ = ['__version__', 'version_info', 'digit_version'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..c6881805330b61eecc632ac7e93d94cf83dab6cc --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/__init__.py @@ -0,0 +1,11 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .inference import inference_segmentor, init_segmentor, show_result_pyplot +from .test import multi_gpu_test, single_gpu_test +from .train import (get_root_logger, init_random_seed, set_random_seed, + train_segmentor) + +__all__ = [ + 'get_root_logger', 'set_random_seed', 'train_segmentor', 'init_segmentor', + 'inference_segmentor', 'multi_gpu_test', 'single_gpu_test', + 'show_result_pyplot', 'init_random_seed' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/inference.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/inference.py new file mode 100644 index 0000000000000000000000000000000000000000..906943804d1ed7b7dce29064204d6d8a3977b23f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/inference.py @@ -0,0 +1,136 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import matplotlib.pyplot as plt +import mmcv +import torch +from mmcv.parallel import collate, scatter +from mmcv.runner import load_checkpoint + +from mmseg.datasets.pipelines import Compose +from mmseg.models import build_segmentor + + +def init_segmentor(config, checkpoint=None, device='cuda:0'): + """Initialize a segmentor from config file. + + Args: + config (str or :obj:`mmcv.Config`): Config file path or the config + object. + checkpoint (str, optional): Checkpoint path. If left as None, the model + will not load any weights. + device (str, optional) CPU/CUDA device option. Default 'cuda:0'. + Use 'cpu' for loading model on CPU. + Returns: + nn.Module: The constructed segmentor. + """ + if isinstance(config, str): + config = mmcv.Config.fromfile(config) + elif not isinstance(config, mmcv.Config): + raise TypeError('config must be a filename or Config object, ' + 'but got {}'.format(type(config))) + config.model.pretrained = None + config.model.train_cfg = None + model = build_segmentor(config.model, test_cfg=config.get('test_cfg')) + if checkpoint is not None: + checkpoint = load_checkpoint(model, checkpoint, map_location='cpu') + model.CLASSES = checkpoint['meta']['CLASSES'] + model.PALETTE = checkpoint['meta']['PALETTE'] + model.cfg = config # save the config in the model for convenience + model.to(device) + model.eval() + return model + + +class LoadImage: + """A simple pipeline to load image.""" + + def __call__(self, results): + """Call function to load images into results. + + Args: + results (dict): A result dict contains the file name + of the image to be read. + + Returns: + dict: ``results`` will be returned containing loaded image. + """ + + if isinstance(results['img'], str): + results['filename'] = results['img'] + results['ori_filename'] = results['img'] + else: + results['filename'] = None + results['ori_filename'] = None + img = mmcv.imread(results['img']) + results['img'] = img + results['img_shape'] = img.shape + results['ori_shape'] = img.shape + return results + + +def inference_segmentor(model, img): + """Inference image(s) with the segmentor. + + Args: + model (nn.Module): The loaded segmentor. + imgs (str/ndarray or list[str/ndarray]): Either image files or loaded + images. + + Returns: + (list[Tensor]): The segmentation result. + """ + cfg = model.cfg + device = next(model.parameters()).device # model device + # build the data pipeline + test_pipeline = [LoadImage()] + cfg.data.test.pipeline[1:] + test_pipeline = Compose(test_pipeline) + # prepare data + data = dict(img=img) + data = test_pipeline(data) + data = collate([data], samples_per_gpu=1) + if next(model.parameters()).is_cuda: + # scatter to specified GPU + data = scatter(data, [device])[0] + else: + data['img_metas'] = [i.data[0] for i in data['img_metas']] + + # forward the model + with torch.no_grad(): + result = model(return_loss=False, rescale=True, **data) + return result + + +def show_result_pyplot(model, + img, + result, + palette=None, + fig_size=(15, 10), + opacity=0.5, + title='', + block=True): + """Visualize the segmentation results on the image. + + Args: + model (nn.Module): The loaded segmentor. + img (str or np.ndarray): Image filename or loaded image. + result (list): The segmentation result. + palette (list[list[int]]] | None): The palette of segmentation + map. If None is given, random palette will be generated. + Default: None + fig_size (tuple): Figure size of the pyplot figure. + opacity(float): Opacity of painted segmentation map. + Default 0.5. + Must be in (0, 1] range. + title (str): The title of pyplot figure. + Default is ''. + block (bool): Whether to block the pyplot figure. + Default is True. + """ + if hasattr(model, 'module'): + model = model.module + img = model.show_result( + img, result, palette=palette, show=False, opacity=opacity) + plt.figure(figsize=fig_size) + plt.imshow(mmcv.bgr2rgb(img)) + plt.title(title) + plt.tight_layout() + plt.show(block=block) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/test.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/test.py new file mode 100644 index 0000000000000000000000000000000000000000..cc4fcc97975741697b7c99c32f66c47b6206f1a6 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/test.py @@ -0,0 +1,233 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import tempfile +import warnings + +import mmcv +import numpy as np +import torch +from mmcv.engine import collect_results_cpu, collect_results_gpu +from mmcv.image import tensor2imgs +from mmcv.runner import get_dist_info + + +def np2tmp(array, temp_file_name=None, tmpdir=None): + """Save ndarray to local numpy file. + + Args: + array (ndarray): Ndarray to save. + temp_file_name (str): Numpy file name. If 'temp_file_name=None', this + function will generate a file name with tempfile.NamedTemporaryFile + to save ndarray. Default: None. + tmpdir (str): Temporary directory to save Ndarray files. Default: None. + Returns: + str: The numpy file name. + """ + + if temp_file_name is None: + temp_file_name = tempfile.NamedTemporaryFile( + suffix='.npy', delete=False, dir=tmpdir).name + np.save(temp_file_name, array) + return temp_file_name + + +def single_gpu_test(model, + data_loader, + show=False, + out_dir=None, + efficient_test=False, + opacity=0.5, + pre_eval=False, + format_only=False, + format_args={}): + """Test with single GPU by progressive mode. + + Args: + model (nn.Module): Model to be tested. + data_loader (utils.data.Dataloader): Pytorch data loader. + show (bool): Whether show results during inference. Default: False. + out_dir (str, optional): If specified, the results will be dumped into + the directory to save output results. + efficient_test (bool): Whether save the results as local numpy files to + save CPU memory during evaluation. Mutually exclusive with + pre_eval and format_results. Default: False. + opacity(float): Opacity of painted segmentation map. + Default 0.5. + Must be in (0, 1] range. + pre_eval (bool): Use dataset.pre_eval() function to generate + pre_results for metric evaluation. Mutually exclusive with + efficient_test and format_results. Default: False. + format_only (bool): Only format result for results commit. + Mutually exclusive with pre_eval and efficient_test. + Default: False. + format_args (dict): The args for format_results. Default: {}. + Returns: + list: list of evaluation pre-results or list of save file names. + """ + if efficient_test: + warnings.warn( + 'DeprecationWarning: ``efficient_test`` will be deprecated, the ' + 'evaluation is CPU memory friendly with pre_eval=True') + mmcv.mkdir_or_exist('.efficient_test') + # when none of them is set true, return segmentation results as + # a list of np.array. + assert [efficient_test, pre_eval, format_only].count(True) <= 1, \ + '``efficient_test``, ``pre_eval`` and ``format_only`` are mutually ' \ + 'exclusive, only one of them could be true .' + + model.eval() + results = [] + dataset = data_loader.dataset + prog_bar = mmcv.ProgressBar(len(dataset)) + # The pipeline about how the data_loader retrieval samples from dataset: + # sampler -> batch_sampler -> indices + # The indices are passed to dataset_fetcher to get data from dataset. + # data_fetcher -> collate_fn(dataset[index]) -> data_sample + # we use batch_sampler to get correct data idx + loader_indices = data_loader.batch_sampler + + for batch_indices, data in zip(loader_indices, data_loader): + with torch.no_grad(): + result = model(return_loss=False, **data) + + if show or out_dir: + img_tensor = data['img'][0] + img_metas = data['img_metas'][0].data[0] + imgs = tensor2imgs(img_tensor, **img_metas[0]['img_norm_cfg']) + assert len(imgs) == len(img_metas) + + for img, img_meta in zip(imgs, img_metas): + h, w, _ = img_meta['img_shape'] + img_show = img[:h, :w, :] + + ori_h, ori_w = img_meta['ori_shape'][:-1] + img_show = mmcv.imresize(img_show, (ori_w, ori_h)) + + if out_dir: + out_file = osp.join(out_dir, img_meta['ori_filename']) + else: + out_file = None + + model.module.show_result( + img_show, + result, + palette=dataset.PALETTE, + show=show, + out_file=out_file, + opacity=opacity) + + if efficient_test: + result = [np2tmp(_, tmpdir='.efficient_test') for _ in result] + + if format_only: + result = dataset.format_results( + result, indices=batch_indices, **format_args) + if pre_eval: + # TODO: adapt samples_per_gpu > 1. + # only samples_per_gpu=1 valid now + result = dataset.pre_eval(result, indices=batch_indices) + results.extend(result) + else: + results.extend(result) + + batch_size = len(result) + for _ in range(batch_size): + prog_bar.update() + + return results + + +def multi_gpu_test(model, + data_loader, + tmpdir=None, + gpu_collect=False, + efficient_test=False, + pre_eval=False, + format_only=False, + format_args={}): + """Test model with multiple gpus by progressive mode. + + This method tests model with multiple gpus and collects the results + under two different modes: gpu and cpu modes. By setting 'gpu_collect=True' + it encodes results to gpu tensors and use gpu communication for results + collection. On cpu mode it saves the results on different gpus to 'tmpdir' + and collects them by the rank 0 worker. + + Args: + model (nn.Module): Model to be tested. + data_loader (utils.data.Dataloader): Pytorch data loader. + tmpdir (str): Path of directory to save the temporary results from + different gpus under cpu mode. The same path is used for efficient + test. Default: None. + gpu_collect (bool): Option to use either gpu or cpu to collect results. + Default: False. + efficient_test (bool): Whether save the results as local numpy files to + save CPU memory during evaluation. Mutually exclusive with + pre_eval and format_results. Default: False. + pre_eval (bool): Use dataset.pre_eval() function to generate + pre_results for metric evaluation. Mutually exclusive with + efficient_test and format_results. Default: False. + format_only (bool): Only format result for results commit. + Mutually exclusive with pre_eval and efficient_test. + Default: False. + format_args (dict): The args for format_results. Default: {}. + + Returns: + list: list of evaluation pre-results or list of save file names. + """ + if efficient_test: + warnings.warn( + 'DeprecationWarning: ``efficient_test`` will be deprecated, the ' + 'evaluation is CPU memory friendly with pre_eval=True') + mmcv.mkdir_or_exist('.efficient_test') + # when none of them is set true, return segmentation results as + # a list of np.array. + assert [efficient_test, pre_eval, format_only].count(True) <= 1, \ + '``efficient_test``, ``pre_eval`` and ``format_only`` are mutually ' \ + 'exclusive, only one of them could be true .' + + model.eval() + results = [] + dataset = data_loader.dataset + # The pipeline about how the data_loader retrieval samples from dataset: + # sampler -> batch_sampler -> indices + # The indices are passed to dataset_fetcher to get data from dataset. + # data_fetcher -> collate_fn(dataset[index]) -> data_sample + # we use batch_sampler to get correct data idx + + # batch_sampler based on DistributedSampler, the indices only point to data + # samples of related machine. + loader_indices = data_loader.batch_sampler + + rank, world_size = get_dist_info() + if rank == 0: + prog_bar = mmcv.ProgressBar(len(dataset)) + + for batch_indices, data in zip(loader_indices, data_loader): + with torch.no_grad(): + result = model(return_loss=False, rescale=True, **data) + + if efficient_test: + result = [np2tmp(_, tmpdir='.efficient_test') for _ in result] + + if format_only: + result = dataset.format_results( + result, indices=batch_indices, **format_args) + if pre_eval: + # TODO: adapt samples_per_gpu > 1. + # only samples_per_gpu=1 valid now + result = dataset.pre_eval(result, indices=batch_indices) + + results.extend(result) + + if rank == 0: + batch_size = len(result) * world_size + for _ in range(batch_size): + prog_bar.update() + + # collect results from all ranks + if gpu_collect: + results = collect_results_gpu(results, len(dataset)) + else: + results = collect_results_cpu(results, len(dataset), tmpdir) + return results diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/train.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/train.py new file mode 100644 index 0000000000000000000000000000000000000000..473adef285519c1c07e341082b7ccab584929cd4 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/apis/train.py @@ -0,0 +1,195 @@ +# Copyright (c) 2022, Shanghai Iluvatar CoreX Semiconductor Co., Ltd. +# Copyright (c) OpenMMLab. All rights reserved. + +import random +import warnings + +import mmcv +import numpy as np +import torch +import torch.distributed as dist +from mmcv.parallel import MMDataParallel, MMDistributedDataParallel +from mmcv.runner import (HOOKS, DistSamplerSeedHook, EpochBasedRunner, + build_runner, get_dist_info) +from mmcv.utils import build_from_cfg + +from mmseg import digit_version +from mmseg.core import DistEvalHook, EvalHook, build_optimizer +from mmseg.datasets import build_dataloader, build_dataset +from mmseg.utils import find_latest_checkpoint, get_root_logger + + +def init_random_seed(seed=None, device='cuda'): + """Initialize random seed. + + If the seed is not set, the seed will be automatically randomized, + and then broadcast to all processes to prevent some potential bugs. + Args: + seed (int, Optional): The seed. Default to None. + device (str): The device where the seed will be put on. + Default to 'cuda'. + Returns: + int: Seed to be used. + """ + if seed is not None: + return seed + + # Make sure all ranks share the same random seed to prevent + # some potential bugs. Please refer to + # https://github.com/open-mmlab/mmdetection/issues/6339 + rank, world_size = get_dist_info() + seed = np.random.randint(2**31) + if world_size == 1: + return seed + + if rank == 0: + random_num = torch.tensor(seed, dtype=torch.int32, device=device) + else: + random_num = torch.tensor(0, dtype=torch.int32, device=device) + dist.broadcast(random_num, src=0) + return random_num.item() + + +def set_random_seed(seed, deterministic=False): + """Set random seed. + + Args: + seed (int): Seed to be used. + deterministic (bool): Whether to set the deterministic option for + CUDNN backend, i.e., set `torch.backends.cudnn.deterministic` + to True and `torch.backends.cudnn.benchmark` to False. + Default: False. + """ + random.seed(seed) + np.random.seed(seed) + torch.manual_seed(seed) + torch.cuda.manual_seed_all(seed) + if deterministic: + torch.backends.cudnn.deterministic = True + torch.backends.cudnn.benchmark = False + + +def train_segmentor(model, + dataset, + cfg, + distributed=False, + validate=False, + timestamp=None, + meta=None): + """Launch segmentor training.""" + logger = get_root_logger(cfg.log_level) + + # prepare data loaders + dataset = dataset if isinstance(dataset, (list, tuple)) else [dataset] + # The default loader config + loader_cfg = dict( + # cfg.gpus will be ignored if distributed + num_gpus=len(cfg.gpu_ids), + dist=distributed, + seed=cfg.seed, + drop_last=True) + # The overall dataloader settings + loader_cfg.update({ + k: v + for k, v in cfg.data.items() if k not in [ + 'train', 'val', 'test', 'train_dataloader', 'val_dataloader', + 'test_dataloader' + ] + }) + + # The specific dataloader settings + train_loader_cfg = {**loader_cfg, **cfg.data.get('train_dataloader', {})} + data_loaders = [build_dataloader(ds, **train_loader_cfg) for ds in dataset] + + # put model on gpus + if distributed: + find_unused_parameters = cfg.get('find_unused_parameters', False) + # Sets the `find_unused_parameters` parameter in + # torch.nn.parallel.DistributedDataParallel + model = MMDistributedDataParallel( + model.cuda(), + device_ids=[torch.cuda.current_device()], + broadcast_buffers=False, + find_unused_parameters=find_unused_parameters) + else: + if not torch.cuda.is_available(): + assert digit_version(mmcv.__version__) >= digit_version('1.4.4'), \ + 'Please use MMCV >= 1.4.4 for CPU training!' + model = MMDataParallel(model, device_ids=cfg.gpu_ids) + # build runner + optimizer = build_optimizer(model, cfg.optimizer) + + if cfg.get('runner') is None: + cfg.runner = {'type': 'IterBasedRunner', 'max_iters': cfg.total_iters} + warnings.warn( + 'config is now expected to have a `runner` section, ' + 'please set `runner` in your config.', UserWarning) + + runner = build_runner( + cfg.runner, + default_args=dict( + model=model, + batch_processor=None, + optimizer=optimizer, + work_dir=cfg.work_dir, + logger=logger, + meta=meta)) + + runner.cfg = cfg + + # register hooks + runner.register_training_hooks(cfg.lr_config, cfg.optimizer_config, + cfg.checkpoint_config, cfg.log_config, + cfg.get('momentum_config', None)) + if distributed: + # when distributed training by epoch, using`DistSamplerSeedHook` to set + # the different seed to distributed sampler for each epoch, it will + # shuffle dataset at each epoch and avoid overfitting. + if isinstance(runner, EpochBasedRunner): + runner.register_hook(DistSamplerSeedHook()) + + # an ugly walkaround to make the .log and .log.json filenames the same + runner.timestamp = timestamp + + # register eval hooks + if validate: + val_dataset = build_dataset(cfg.data.val, dict(test_mode=True)) + # The specific dataloader settings + val_loader_cfg = { + **loader_cfg, + 'samples_per_gpu': 1, + 'shuffle': False, # Not shuffle by default + **cfg.data.get('val_dataloader', {}), + } + val_dataloader = build_dataloader(val_dataset, **val_loader_cfg) + eval_cfg = cfg.get('evaluation', {}) + eval_cfg['by_epoch'] = cfg.runner['type'] != 'IterBasedRunner' + eval_hook = DistEvalHook if distributed else EvalHook + # In this PR (https://github.com/open-mmlab/mmcv/pull/1193), the + # priority of IterTimerHook has been modified from 'NORMAL' to 'LOW'. + runner.register_hook( + eval_hook(val_dataloader, **eval_cfg), priority='LOW') + + # user-defined hooks + if cfg.get('custom_hooks', None): + custom_hooks = cfg.custom_hooks + assert isinstance(custom_hooks, list), \ + f'custom_hooks expect list type, but got {type(custom_hooks)}' + for hook_cfg in cfg.custom_hooks: + assert isinstance(hook_cfg, dict), \ + 'Each item in custom_hooks expects dict type, but got ' \ + f'{type(hook_cfg)}' + hook_cfg = hook_cfg.copy() + priority = hook_cfg.pop('priority', 'NORMAL') + hook = build_from_cfg(hook_cfg, HOOKS) + runner.register_hook(hook, priority=priority) + + if cfg.resume_from is None and cfg.get('auto_resume'): + resume_from = find_latest_checkpoint(cfg.work_dir) + if resume_from is not None: + cfg.resume_from = resume_from + if cfg.resume_from: + runner.resume(cfg.resume_from) + elif cfg.load_from: + runner.load_checkpoint(cfg.load_from) + runner.run(data_loaders, cfg.workflow) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..1a077d2f1fd8889b8f60851cb7940ec8abc28567 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/__init__.py @@ -0,0 +1,11 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .builder import (OPTIMIZER_BUILDERS, build_optimizer, + build_optimizer_constructor) +from .evaluation import * # noqa: F401, F403 +from .optimizers import * # noqa: F401, F403 +from .seg import * # noqa: F401, F403 +from .utils import * # noqa: F401, F403 + +__all__ = [ + 'OPTIMIZER_BUILDERS', 'build_optimizer', 'build_optimizer_constructor' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/builder.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/builder.py new file mode 100644 index 0000000000000000000000000000000000000000..406dd9b4b7027e9c2254b0d18cf0c80a7161912b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/builder.py @@ -0,0 +1,33 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy + +from mmcv.runner.optimizer import OPTIMIZER_BUILDERS as MMCV_OPTIMIZER_BUILDERS +from mmcv.utils import Registry, build_from_cfg + +OPTIMIZER_BUILDERS = Registry( + 'optimizer builder', parent=MMCV_OPTIMIZER_BUILDERS) + + +def build_optimizer_constructor(cfg): + constructor_type = cfg.get('type') + if constructor_type in OPTIMIZER_BUILDERS: + return build_from_cfg(cfg, OPTIMIZER_BUILDERS) + elif constructor_type in MMCV_OPTIMIZER_BUILDERS: + return build_from_cfg(cfg, MMCV_OPTIMIZER_BUILDERS) + else: + raise KeyError(f'{constructor_type} is not registered ' + 'in the optimizer builder registry.') + + +def build_optimizer(model, cfg): + optimizer_cfg = copy.deepcopy(cfg) + constructor_type = optimizer_cfg.pop('constructor', + 'DefaultOptimizerConstructor') + paramwise_cfg = optimizer_cfg.pop('paramwise_cfg', None) + optim_constructor = build_optimizer_constructor( + dict( + type=constructor_type, + optimizer_cfg=optimizer_cfg, + paramwise_cfg=paramwise_cfg)) + optimizer = optim_constructor(model) + return optimizer diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..3d16d17e54222f006e32cd6b9e6ca323e3738f03 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/__init__.py @@ -0,0 +1,11 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .class_names import get_classes, get_palette +from .eval_hooks import DistEvalHook, EvalHook +from .metrics import (eval_metrics, intersect_and_union, mean_dice, + mean_fscore, mean_iou, pre_eval_to_metrics) + +__all__ = [ + 'EvalHook', 'DistEvalHook', 'mean_dice', 'mean_iou', 'mean_fscore', + 'eval_metrics', 'get_classes', 'get_palette', 'pre_eval_to_metrics', + 'intersect_and_union' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/class_names.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/class_names.py new file mode 100644 index 0000000000000000000000000000000000000000..e3bff6231435412852d87b22f5b897207c1f56af --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/class_names.py @@ -0,0 +1,316 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import mmcv + + +def cityscapes_classes(): + """Cityscapes class names for external use.""" + return [ + 'road', 'sidewalk', 'building', 'wall', 'fence', 'pole', + 'traffic light', 'traffic sign', 'vegetation', 'terrain', 'sky', + 'person', 'rider', 'car', 'truck', 'bus', 'train', 'motorcycle', + 'bicycle' + ] + + +def ade_classes(): + """ADE20K class names for external use.""" + return [ + 'wall', 'building', 'sky', 'floor', 'tree', 'ceiling', 'road', 'bed ', + 'windowpane', 'grass', 'cabinet', 'sidewalk', 'person', 'earth', + 'door', 'table', 'mountain', 'plant', 'curtain', 'chair', 'car', + 'water', 'painting', 'sofa', 'shelf', 'house', 'sea', 'mirror', 'rug', + 'field', 'armchair', 'seat', 'fence', 'desk', 'rock', 'wardrobe', + 'lamp', 'bathtub', 'railing', 'cushion', 'base', 'box', 'column', + 'signboard', 'chest of drawers', 'counter', 'sand', 'sink', + 'skyscraper', 'fireplace', 'refrigerator', 'grandstand', 'path', + 'stairs', 'runway', 'case', 'pool table', 'pillow', 'screen door', + 'stairway', 'river', 'bridge', 'bookcase', 'blind', 'coffee table', + 'toilet', 'flower', 'book', 'hill', 'bench', 'countertop', 'stove', + 'palm', 'kitchen island', 'computer', 'swivel chair', 'boat', 'bar', + 'arcade machine', 'hovel', 'bus', 'towel', 'light', 'truck', 'tower', + 'chandelier', 'awning', 'streetlight', 'booth', 'television receiver', + 'airplane', 'dirt track', 'apparel', 'pole', 'land', 'bannister', + 'escalator', 'ottoman', 'bottle', 'buffet', 'poster', 'stage', 'van', + 'ship', 'fountain', 'conveyer belt', 'canopy', 'washer', 'plaything', + 'swimming pool', 'stool', 'barrel', 'basket', 'waterfall', 'tent', + 'bag', 'minibike', 'cradle', 'oven', 'ball', 'food', 'step', 'tank', + 'trade name', 'microwave', 'pot', 'animal', 'bicycle', 'lake', + 'dishwasher', 'screen', 'blanket', 'sculpture', 'hood', 'sconce', + 'vase', 'traffic light', 'tray', 'ashcan', 'fan', 'pier', 'crt screen', + 'plate', 'monitor', 'bulletin board', 'shower', 'radiator', 'glass', + 'clock', 'flag' + ] + + +def voc_classes(): + """Pascal VOC class names for external use.""" + return [ + 'background', 'aeroplane', 'bicycle', 'bird', 'boat', 'bottle', 'bus', + 'car', 'cat', 'chair', 'cow', 'diningtable', 'dog', 'horse', + 'motorbike', 'person', 'pottedplant', 'sheep', 'sofa', 'train', + 'tvmonitor' + ] + + +def cocostuff_classes(): + """CocoStuff class names for external use.""" + return [ + 'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', + 'truck', 'boat', 'traffic light', 'fire hydrant', 'stop sign', + 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', + 'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', + 'handbag', 'tie', 'suitcase', 'frisbee', 'skis', 'snowboard', + 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', + 'surfboard', 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', + 'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange', + 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', + 'couch', 'potted plant', 'bed', 'dining table', 'toilet', 'tv', + 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', + 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', + 'scissors', 'teddy bear', 'hair drier', 'toothbrush', 'banner', + 'blanket', 'branch', 'bridge', 'building-other', 'bush', 'cabinet', + 'cage', 'cardboard', 'carpet', 'ceiling-other', 'ceiling-tile', + 'cloth', 'clothes', 'clouds', 'counter', 'cupboard', 'curtain', + 'desk-stuff', 'dirt', 'door-stuff', 'fence', 'floor-marble', + 'floor-other', 'floor-stone', 'floor-tile', 'floor-wood', 'flower', + 'fog', 'food-other', 'fruit', 'furniture-other', 'grass', 'gravel', + 'ground-other', 'hill', 'house', 'leaves', 'light', 'mat', 'metal', + 'mirror-stuff', 'moss', 'mountain', 'mud', 'napkin', 'net', 'paper', + 'pavement', 'pillow', 'plant-other', 'plastic', 'platform', + 'playingfield', 'railing', 'railroad', 'river', 'road', 'rock', 'roof', + 'rug', 'salad', 'sand', 'sea', 'shelf', 'sky-other', 'skyscraper', + 'snow', 'solid-other', 'stairs', 'stone', 'straw', 'structural-other', + 'table', 'tent', 'textile-other', 'towel', 'tree', 'vegetable', + 'wall-brick', 'wall-concrete', 'wall-other', 'wall-panel', + 'wall-stone', 'wall-tile', 'wall-wood', 'water-other', 'waterdrops', + 'window-blind', 'window-other', 'wood' + ] + + +def loveda_classes(): + """LoveDA class names for external use.""" + return [ + 'background', 'building', 'road', 'water', 'barren', 'forest', + 'agricultural' + ] + + +def potsdam_classes(): + """Potsdam class names for external use.""" + return [ + 'impervious_surface', 'building', 'low_vegetation', 'tree', 'car', + 'clutter' + ] + + +def vaihingen_classes(): + """Vaihingen class names for external use.""" + return [ + 'impervious_surface', 'building', 'low_vegetation', 'tree', 'car', + 'clutter' + ] + + +def isaid_classes(): + """iSAID class names for external use.""" + return [ + 'background', 'ship', 'store_tank', 'baseball_diamond', 'tennis_court', + 'basketball_court', 'Ground_Track_Field', 'Bridge', 'Large_Vehicle', + 'Small_Vehicle', 'Helicopter', 'Swimming_pool', 'Roundabout', + 'Soccer_ball_field', 'plane', 'Harbor' + ] + + +def stare_classes(): + """stare class names for external use.""" + return ['background', 'vessel'] + + +def cityscapes_palette(): + """Cityscapes palette for external use.""" + return [[128, 64, 128], [244, 35, 232], [70, 70, 70], [102, 102, 156], + [190, 153, 153], [153, 153, 153], [250, 170, 30], [220, 220, 0], + [107, 142, 35], [152, 251, 152], [70, 130, 180], [220, 20, 60], + [255, 0, 0], [0, 0, 142], [0, 0, 70], [0, 60, 100], [0, 80, 100], + [0, 0, 230], [119, 11, 32]] + + +def ade_palette(): + """ADE20K palette for external use.""" + return [[120, 120, 120], [180, 120, 120], [6, 230, 230], [80, 50, 50], + [4, 200, 3], [120, 120, 80], [140, 140, 140], [204, 5, 255], + [230, 230, 230], [4, 250, 7], [224, 5, 255], [235, 255, 7], + [150, 5, 61], [120, 120, 70], [8, 255, 51], [255, 6, 82], + [143, 255, 140], [204, 255, 4], [255, 51, 7], [204, 70, 3], + [0, 102, 200], [61, 230, 250], [255, 6, 51], [11, 102, 255], + [255, 7, 71], [255, 9, 224], [9, 7, 230], [220, 220, 220], + [255, 9, 92], [112, 9, 255], [8, 255, 214], [7, 255, 224], + [255, 184, 6], [10, 255, 71], [255, 41, 10], [7, 255, 255], + [224, 255, 8], [102, 8, 255], [255, 61, 6], [255, 194, 7], + [255, 122, 8], [0, 255, 20], [255, 8, 41], [255, 5, 153], + [6, 51, 255], [235, 12, 255], [160, 150, 20], [0, 163, 255], + [140, 140, 140], [250, 10, 15], [20, 255, 0], [31, 255, 0], + [255, 31, 0], [255, 224, 0], [153, 255, 0], [0, 0, 255], + [255, 71, 0], [0, 235, 255], [0, 173, 255], [31, 0, 255], + [11, 200, 200], [255, 82, 0], [0, 255, 245], [0, 61, 255], + [0, 255, 112], [0, 255, 133], [255, 0, 0], [255, 163, 0], + [255, 102, 0], [194, 255, 0], [0, 143, 255], [51, 255, 0], + [0, 82, 255], [0, 255, 41], [0, 255, 173], [10, 0, 255], + [173, 255, 0], [0, 255, 153], [255, 92, 0], [255, 0, 255], + [255, 0, 245], [255, 0, 102], [255, 173, 0], [255, 0, 20], + [255, 184, 184], [0, 31, 255], [0, 255, 61], [0, 71, 255], + [255, 0, 204], [0, 255, 194], [0, 255, 82], [0, 10, 255], + [0, 112, 255], [51, 0, 255], [0, 194, 255], [0, 122, 255], + [0, 255, 163], [255, 153, 0], [0, 255, 10], [255, 112, 0], + [143, 255, 0], [82, 0, 255], [163, 255, 0], [255, 235, 0], + [8, 184, 170], [133, 0, 255], [0, 255, 92], [184, 0, 255], + [255, 0, 31], [0, 184, 255], [0, 214, 255], [255, 0, 112], + [92, 255, 0], [0, 224, 255], [112, 224, 255], [70, 184, 160], + [163, 0, 255], [153, 0, 255], [71, 255, 0], [255, 0, 163], + [255, 204, 0], [255, 0, 143], [0, 255, 235], [133, 255, 0], + [255, 0, 235], [245, 0, 255], [255, 0, 122], [255, 245, 0], + [10, 190, 212], [214, 255, 0], [0, 204, 255], [20, 0, 255], + [255, 255, 0], [0, 153, 255], [0, 41, 255], [0, 255, 204], + [41, 0, 255], [41, 255, 0], [173, 0, 255], [0, 245, 255], + [71, 0, 255], [122, 0, 255], [0, 255, 184], [0, 92, 255], + [184, 255, 0], [0, 133, 255], [255, 214, 0], [25, 194, 194], + [102, 255, 0], [92, 0, 255]] + + +def voc_palette(): + """Pascal VOC palette for external use.""" + return [[0, 0, 0], [128, 0, 0], [0, 128, 0], [128, 128, 0], [0, 0, 128], + [128, 0, 128], [0, 128, 128], [128, 128, 128], [64, 0, 0], + [192, 0, 0], [64, 128, 0], [192, 128, 0], [64, 0, 128], + [192, 0, 128], [64, 128, 128], [192, 128, 128], [0, 64, 0], + [128, 64, 0], [0, 192, 0], [128, 192, 0], [0, 64, 128]] + + +def cocostuff_palette(): + """CocoStuff palette for external use.""" + return [[0, 192, 64], [0, 192, 64], [0, 64, 96], [128, 192, 192], + [0, 64, 64], [0, 192, 224], [0, 192, 192], [128, 192, 64], + [0, 192, 96], [128, 192, 64], [128, 32, 192], [0, 0, 224], + [0, 0, 64], [0, 160, 192], [128, 0, 96], [128, 0, 192], + [0, 32, 192], [128, 128, 224], [0, 0, 192], [128, 160, 192], + [128, 128, 0], [128, 0, 32], [128, 32, 0], [128, 0, 128], + [64, 128, 32], [0, 160, 0], [0, 0, 0], [192, 128, 160], [0, 32, 0], + [0, 128, 128], [64, 128, 160], [128, 160, 0], [0, 128, 0], + [192, 128, 32], [128, 96, 128], [0, 0, 128], [64, 0, 32], + [0, 224, 128], [128, 0, 0], [192, 0, 160], [0, 96, 128], + [128, 128, 128], [64, 0, 160], [128, 224, 128], [128, 128, 64], + [192, 0, 32], [128, 96, 0], [128, 0, 192], [0, 128, 32], + [64, 224, 0], [0, 0, 64], [128, 128, 160], [64, 96, 0], + [0, 128, 192], [0, 128, 160], [192, 224, 0], [0, 128, 64], + [128, 128, 32], [192, 32, 128], [0, 64, 192], [0, 0, 32], + [64, 160, 128], [128, 64, 64], [128, 0, 160], [64, 32, 128], + [128, 192, 192], [0, 0, 160], [192, 160, 128], [128, 192, 0], + [128, 0, 96], [192, 32, 0], [128, 64, 128], [64, 128, 96], + [64, 160, 0], [0, 64, 0], [192, 128, 224], [64, 32, 0], + [0, 192, 128], [64, 128, 224], [192, 160, 0], [0, 192, 0], + [192, 128, 96], [192, 96, 128], [0, 64, 128], [64, 0, 96], + [64, 224, 128], [128, 64, 0], [192, 0, 224], [64, 96, 128], + [128, 192, 128], [64, 0, 224], [192, 224, 128], [128, 192, 64], + [192, 0, 96], [192, 96, 0], [128, 64, 192], [0, 128, 96], + [0, 224, 0], [64, 64, 64], [128, 128, 224], [0, 96, 0], + [64, 192, 192], [0, 128, 224], [128, 224, 0], [64, 192, 64], + [128, 128, 96], [128, 32, 128], [64, 0, 192], [0, 64, 96], + [0, 160, 128], [192, 0, 64], [128, 64, 224], [0, 32, 128], + [192, 128, 192], [0, 64, 224], [128, 160, 128], [192, 128, 0], + [128, 64, 32], [128, 32, 64], [192, 0, 128], [64, 192, 32], + [0, 160, 64], [64, 0, 0], [192, 192, 160], [0, 32, 64], + [64, 128, 128], [64, 192, 160], [128, 160, 64], [64, 128, 0], + [192, 192, 32], [128, 96, 192], [64, 0, 128], [64, 64, 32], + [0, 224, 192], [192, 0, 0], [192, 64, 160], [0, 96, 192], + [192, 128, 128], [64, 64, 160], [128, 224, 192], [192, 128, 64], + [192, 64, 32], [128, 96, 64], [192, 0, 192], [0, 192, 32], + [64, 224, 64], [64, 0, 64], [128, 192, 160], [64, 96, 64], + [64, 128, 192], [0, 192, 160], [192, 224, 64], [64, 128, 64], + [128, 192, 32], [192, 32, 192], [64, 64, 192], [0, 64, 32], + [64, 160, 192], [192, 64, 64], [128, 64, 160], [64, 32, 192], + [192, 192, 192], [0, 64, 160], [192, 160, 192], [192, 192, 0], + [128, 64, 96], [192, 32, 64], [192, 64, 128], [64, 192, 96], + [64, 160, 64], [64, 64, 0]] + + +def loveda_palette(): + """LoveDA palette for external use.""" + return [[255, 255, 255], [255, 0, 0], [255, 255, 0], [0, 0, 255], + [159, 129, 183], [0, 255, 0], [255, 195, 128]] + + +def potsdam_palette(): + """Potsdam palette for external use.""" + return [[255, 255, 255], [0, 0, 255], [0, 255, 255], [0, 255, 0], + [255, 255, 0], [255, 0, 0]] + + +def vaihingen_palette(): + """Vaihingen palette for external use.""" + return [[255, 255, 255], [0, 0, 255], [0, 255, 255], [0, 255, 0], + [255, 255, 0], [255, 0, 0]] + + +def isaid_palette(): + """iSAID palette for external use.""" + return [[0, 0, 0], [0, 0, 63], [0, 63, 63], [0, 63, 0], [0, 63, 127], + [0, 63, 191], [0, 63, 255], [0, 127, 63], [0, 127, + 127], [0, 0, 127], + [0, 0, 191], [0, 0, 255], [0, 191, 127], [0, 127, 191], + [0, 127, 255], [0, 100, 155]] + + +def stare_palette(): + """STARE palette for external use.""" + return [[120, 120, 120], [6, 230, 230]] + + +dataset_aliases = { + 'cityscapes': ['cityscapes'], + 'ade': ['ade', 'ade20k'], + 'voc': ['voc', 'pascal_voc', 'voc12', 'voc12aug'], + 'loveda': ['loveda'], + 'potsdam': ['potsdam'], + 'vaihingen': ['vaihingen'], + 'cocostuff': [ + 'cocostuff', 'cocostuff10k', 'cocostuff164k', 'coco-stuff', + 'coco-stuff10k', 'coco-stuff164k', 'coco_stuff', 'coco_stuff10k', + 'coco_stuff164k' + ], + 'isaid': ['isaid', 'iSAID'], + 'stare': ['stare', 'STARE'] +} + + +def get_classes(dataset): + """Get class names of a dataset.""" + alias2name = {} + for name, aliases in dataset_aliases.items(): + for alias in aliases: + alias2name[alias] = name + + if mmcv.is_str(dataset): + if dataset in alias2name: + labels = eval(alias2name[dataset] + '_classes()') + else: + raise ValueError(f'Unrecognized dataset: {dataset}') + else: + raise TypeError(f'dataset must a str, but got {type(dataset)}') + return labels + + +def get_palette(dataset): + """Get class palette (RGB) of a dataset.""" + alias2name = {} + for name, aliases in dataset_aliases.items(): + for alias in aliases: + alias2name[alias] = name + + if mmcv.is_str(dataset): + if dataset in alias2name: + labels = eval(alias2name[dataset] + '_palette()') + else: + raise ValueError(f'Unrecognized dataset: {dataset}') + else: + raise TypeError(f'dataset must a str, but got {type(dataset)}') + return labels diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/eval_hooks.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/eval_hooks.py new file mode 100644 index 0000000000000000000000000000000000000000..952db3b0b410ecbe1999435c3ca8722034381ef7 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/eval_hooks.py @@ -0,0 +1,128 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import warnings + +import torch.distributed as dist +from mmcv.runner import DistEvalHook as _DistEvalHook +from mmcv.runner import EvalHook as _EvalHook +from torch.nn.modules.batchnorm import _BatchNorm + + +class EvalHook(_EvalHook): + """Single GPU EvalHook, with efficient test support. + + Args: + by_epoch (bool): Determine perform evaluation by epoch or by iteration. + If set to True, it will perform by epoch. Otherwise, by iteration. + Default: False. + efficient_test (bool): Whether save the results as local numpy files to + save CPU memory during evaluation. Default: False. + pre_eval (bool): Whether to use progressive mode to evaluate model. + Default: False. + Returns: + list: The prediction results. + """ + + greater_keys = ['mIoU', 'mAcc', 'aAcc'] + + def __init__(self, + *args, + by_epoch=False, + efficient_test=False, + pre_eval=False, + **kwargs): + super().__init__(*args, by_epoch=by_epoch, **kwargs) + self.pre_eval = pre_eval + if efficient_test: + warnings.warn( + 'DeprecationWarning: ``efficient_test`` for evaluation hook ' + 'is deprecated, the evaluation hook is CPU memory friendly ' + 'with ``pre_eval=True`` as argument for ``single_gpu_test()`` ' + 'function') + + def _do_evaluate(self, runner): + """perform evaluation and save ckpt.""" + if not self._should_evaluate(runner): + return + + from mmseg.apis import single_gpu_test + results = single_gpu_test( + runner.model, self.dataloader, show=False, pre_eval=self.pre_eval) + runner.log_buffer.clear() + runner.log_buffer.output['eval_iter_num'] = len(self.dataloader) + key_score = self.evaluate(runner, results) + if self.save_best: + self._save_ckpt(runner, key_score) + + +class DistEvalHook(_DistEvalHook): + """Distributed EvalHook, with efficient test support. + + Args: + by_epoch (bool): Determine perform evaluation by epoch or by iteration. + If set to True, it will perform by epoch. Otherwise, by iteration. + Default: False. + efficient_test (bool): Whether save the results as local numpy files to + save CPU memory during evaluation. Default: False. + pre_eval (bool): Whether to use progressive mode to evaluate model. + Default: False. + Returns: + list: The prediction results. + """ + + greater_keys = ['mIoU', 'mAcc', 'aAcc'] + + def __init__(self, + *args, + by_epoch=False, + efficient_test=False, + pre_eval=False, + **kwargs): + super().__init__(*args, by_epoch=by_epoch, **kwargs) + self.pre_eval = pre_eval + if efficient_test: + warnings.warn( + 'DeprecationWarning: ``efficient_test`` for evaluation hook ' + 'is deprecated, the evaluation hook is CPU memory friendly ' + 'with ``pre_eval=True`` as argument for ``multi_gpu_test()`` ' + 'function') + + def _do_evaluate(self, runner): + """perform evaluation and save ckpt.""" + # Synchronization of BatchNorm's buffer (running_mean + # and running_var) is not supported in the DDP of pytorch, + # which may cause the inconsistent performance of models in + # different ranks, so we broadcast BatchNorm's buffers + # of rank 0 to other ranks to avoid this. + if self.broadcast_bn_buffer: + model = runner.model + for name, module in model.named_modules(): + if isinstance(module, + _BatchNorm) and module.track_running_stats: + dist.broadcast(module.running_var, 0) + dist.broadcast(module.running_mean, 0) + + if not self._should_evaluate(runner): + return + + tmpdir = self.tmpdir + if tmpdir is None: + tmpdir = osp.join(runner.work_dir, '.eval_hook') + + from mmseg.apis import multi_gpu_test + results = multi_gpu_test( + runner.model, + self.dataloader, + tmpdir=tmpdir, + gpu_collect=self.gpu_collect, + pre_eval=self.pre_eval) + + runner.log_buffer.clear() + + if runner.rank == 0: + print('\n') + runner.log_buffer.output['eval_iter_num'] = len(self.dataloader) + key_score = self.evaluate(runner, results) + + if self.save_best: + self._save_ckpt(runner, key_score) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/metrics.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/metrics.py new file mode 100644 index 0000000000000000000000000000000000000000..a1c0908e155cf06f51c72f4cec27ad15a482d4eb --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/evaluation/metrics.py @@ -0,0 +1,395 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from collections import OrderedDict + +import mmcv +import numpy as np +import torch + + +def f_score(precision, recall, beta=1): + """calculate the f-score value. + + Args: + precision (float | torch.Tensor): The precision value. + recall (float | torch.Tensor): The recall value. + beta (int): Determines the weight of recall in the combined score. + Default: False. + + Returns: + [torch.tensor]: The f-score value. + """ + score = (1 + beta**2) * (precision * recall) / ( + (beta**2 * precision) + recall) + return score + + +def intersect_and_union(pred_label, + label, + num_classes, + ignore_index, + label_map=dict(), + reduce_zero_label=False): + """Calculate intersection and Union. + + Args: + pred_label (ndarray | str): Prediction segmentation map + or predict result filename. + label (ndarray | str): Ground truth segmentation map + or label filename. + num_classes (int): Number of categories. + ignore_index (int): Index that will be ignored in evaluation. + label_map (dict): Mapping old labels to new labels. The parameter will + work only when label is str. Default: dict(). + reduce_zero_label (bool): Whether ignore zero label. The parameter will + work only when label is str. Default: False. + + Returns: + torch.Tensor: The intersection of prediction and ground truth + histogram on all classes. + torch.Tensor: The union of prediction and ground truth histogram on + all classes. + torch.Tensor: The prediction histogram on all classes. + torch.Tensor: The ground truth histogram on all classes. + """ + + if isinstance(pred_label, str): + pred_label = torch.from_numpy(np.load(pred_label)) + else: + pred_label = torch.from_numpy((pred_label)) + + if isinstance(label, str): + label = torch.from_numpy( + mmcv.imread(label, flag='unchanged', backend='pillow')) + else: + label = torch.from_numpy(label) + + if label_map is not None: + for old_id, new_id in label_map.items(): + label[label == old_id] = new_id + if reduce_zero_label: + label[label == 0] = 255 + label = label - 1 + label[label == 254] = 255 + + mask = (label != ignore_index) + pred_label = pred_label[mask] + label = label[mask] + + intersect = pred_label[pred_label == label] + area_intersect = torch.histc( + intersect.float(), bins=(num_classes), min=0, max=num_classes - 1) + area_pred_label = torch.histc( + pred_label.float(), bins=(num_classes), min=0, max=num_classes - 1) + area_label = torch.histc( + label.float(), bins=(num_classes), min=0, max=num_classes - 1) + area_union = area_pred_label + area_label - area_intersect + return area_intersect, area_union, area_pred_label, area_label + + +def total_intersect_and_union(results, + gt_seg_maps, + num_classes, + ignore_index, + label_map=dict(), + reduce_zero_label=False): + """Calculate Total Intersection and Union. + + Args: + results (list[ndarray] | list[str]): List of prediction segmentation + maps or list of prediction result filenames. + gt_seg_maps (list[ndarray] | list[str] | Iterables): list of ground + truth segmentation maps or list of label filenames. + num_classes (int): Number of categories. + ignore_index (int): Index that will be ignored in evaluation. + label_map (dict): Mapping old labels to new labels. Default: dict(). + reduce_zero_label (bool): Whether ignore zero label. Default: False. + + Returns: + ndarray: The intersection of prediction and ground truth histogram + on all classes. + ndarray: The union of prediction and ground truth histogram on all + classes. + ndarray: The prediction histogram on all classes. + ndarray: The ground truth histogram on all classes. + """ + total_area_intersect = torch.zeros((num_classes, ), dtype=torch.float64) + total_area_union = torch.zeros((num_classes, ), dtype=torch.float64) + total_area_pred_label = torch.zeros((num_classes, ), dtype=torch.float64) + total_area_label = torch.zeros((num_classes, ), dtype=torch.float64) + for result, gt_seg_map in zip(results, gt_seg_maps): + area_intersect, area_union, area_pred_label, area_label = \ + intersect_and_union( + result, gt_seg_map, num_classes, ignore_index, + label_map, reduce_zero_label) + total_area_intersect += area_intersect + total_area_union += area_union + total_area_pred_label += area_pred_label + total_area_label += area_label + return total_area_intersect, total_area_union, total_area_pred_label, \ + total_area_label + + +def mean_iou(results, + gt_seg_maps, + num_classes, + ignore_index, + nan_to_num=None, + label_map=dict(), + reduce_zero_label=False): + """Calculate Mean Intersection and Union (mIoU) + + Args: + results (list[ndarray] | list[str]): List of prediction segmentation + maps or list of prediction result filenames. + gt_seg_maps (list[ndarray] | list[str]): list of ground truth + segmentation maps or list of label filenames. + num_classes (int): Number of categories. + ignore_index (int): Index that will be ignored in evaluation. + nan_to_num (int, optional): If specified, NaN values will be replaced + by the numbers defined by the user. Default: None. + label_map (dict): Mapping old labels to new labels. Default: dict(). + reduce_zero_label (bool): Whether ignore zero label. Default: False. + + Returns: + dict[str, float | ndarray]: + float: Overall accuracy on all images. + ndarray: Per category accuracy, shape (num_classes, ). + ndarray: Per category IoU, shape (num_classes, ). + """ + iou_result = eval_metrics( + results=results, + gt_seg_maps=gt_seg_maps, + num_classes=num_classes, + ignore_index=ignore_index, + metrics=['mIoU'], + nan_to_num=nan_to_num, + label_map=label_map, + reduce_zero_label=reduce_zero_label) + return iou_result + + +def mean_dice(results, + gt_seg_maps, + num_classes, + ignore_index, + nan_to_num=None, + label_map=dict(), + reduce_zero_label=False): + """Calculate Mean Dice (mDice) + + Args: + results (list[ndarray] | list[str]): List of prediction segmentation + maps or list of prediction result filenames. + gt_seg_maps (list[ndarray] | list[str]): list of ground truth + segmentation maps or list of label filenames. + num_classes (int): Number of categories. + ignore_index (int): Index that will be ignored in evaluation. + nan_to_num (int, optional): If specified, NaN values will be replaced + by the numbers defined by the user. Default: None. + label_map (dict): Mapping old labels to new labels. Default: dict(). + reduce_zero_label (bool): Whether ignore zero label. Default: False. + + Returns: + dict[str, float | ndarray]: Default metrics. + float: Overall accuracy on all images. + ndarray: Per category accuracy, shape (num_classes, ). + ndarray: Per category dice, shape (num_classes, ). + """ + + dice_result = eval_metrics( + results=results, + gt_seg_maps=gt_seg_maps, + num_classes=num_classes, + ignore_index=ignore_index, + metrics=['mDice'], + nan_to_num=nan_to_num, + label_map=label_map, + reduce_zero_label=reduce_zero_label) + return dice_result + + +def mean_fscore(results, + gt_seg_maps, + num_classes, + ignore_index, + nan_to_num=None, + label_map=dict(), + reduce_zero_label=False, + beta=1): + """Calculate Mean Intersection and Union (mIoU) + + Args: + results (list[ndarray] | list[str]): List of prediction segmentation + maps or list of prediction result filenames. + gt_seg_maps (list[ndarray] | list[str]): list of ground truth + segmentation maps or list of label filenames. + num_classes (int): Number of categories. + ignore_index (int): Index that will be ignored in evaluation. + nan_to_num (int, optional): If specified, NaN values will be replaced + by the numbers defined by the user. Default: None. + label_map (dict): Mapping old labels to new labels. Default: dict(). + reduce_zero_label (bool): Whether ignore zero label. Default: False. + beta (int): Determines the weight of recall in the combined score. + Default: False. + + + Returns: + dict[str, float | ndarray]: Default metrics. + float: Overall accuracy on all images. + ndarray: Per category recall, shape (num_classes, ). + ndarray: Per category precision, shape (num_classes, ). + ndarray: Per category f-score, shape (num_classes, ). + """ + fscore_result = eval_metrics( + results=results, + gt_seg_maps=gt_seg_maps, + num_classes=num_classes, + ignore_index=ignore_index, + metrics=['mFscore'], + nan_to_num=nan_to_num, + label_map=label_map, + reduce_zero_label=reduce_zero_label, + beta=beta) + return fscore_result + + +def eval_metrics(results, + gt_seg_maps, + num_classes, + ignore_index, + metrics=['mIoU'], + nan_to_num=None, + label_map=dict(), + reduce_zero_label=False, + beta=1): + """Calculate evaluation metrics + Args: + results (list[ndarray] | list[str]): List of prediction segmentation + maps or list of prediction result filenames. + gt_seg_maps (list[ndarray] | list[str] | Iterables): list of ground + truth segmentation maps or list of label filenames. + num_classes (int): Number of categories. + ignore_index (int): Index that will be ignored in evaluation. + metrics (list[str] | str): Metrics to be evaluated, 'mIoU' and 'mDice'. + nan_to_num (int, optional): If specified, NaN values will be replaced + by the numbers defined by the user. Default: None. + label_map (dict): Mapping old labels to new labels. Default: dict(). + reduce_zero_label (bool): Whether ignore zero label. Default: False. + Returns: + float: Overall accuracy on all images. + ndarray: Per category accuracy, shape (num_classes, ). + ndarray: Per category evaluation metrics, shape (num_classes, ). + """ + + total_area_intersect, total_area_union, total_area_pred_label, \ + total_area_label = total_intersect_and_union( + results, gt_seg_maps, num_classes, ignore_index, label_map, + reduce_zero_label) + ret_metrics = total_area_to_metrics(total_area_intersect, total_area_union, + total_area_pred_label, + total_area_label, metrics, nan_to_num, + beta) + + return ret_metrics + + +def pre_eval_to_metrics(pre_eval_results, + metrics=['mIoU'], + nan_to_num=None, + beta=1): + """Convert pre-eval results to metrics. + + Args: + pre_eval_results (list[tuple[torch.Tensor]]): per image eval results + for computing evaluation metric + metrics (list[str] | str): Metrics to be evaluated, 'mIoU' and 'mDice'. + nan_to_num (int, optional): If specified, NaN values will be replaced + by the numbers defined by the user. Default: None. + Returns: + float: Overall accuracy on all images. + ndarray: Per category accuracy, shape (num_classes, ). + ndarray: Per category evaluation metrics, shape (num_classes, ). + """ + + # convert list of tuples to tuple of lists, e.g. + # [(A_1, B_1, C_1, D_1), ..., (A_n, B_n, C_n, D_n)] to + # ([A_1, ..., A_n], ..., [D_1, ..., D_n]) + pre_eval_results = tuple(zip(*pre_eval_results)) + assert len(pre_eval_results) == 4 + + total_area_intersect = sum(pre_eval_results[0]) + total_area_union = sum(pre_eval_results[1]) + total_area_pred_label = sum(pre_eval_results[2]) + total_area_label = sum(pre_eval_results[3]) + + ret_metrics = total_area_to_metrics(total_area_intersect, total_area_union, + total_area_pred_label, + total_area_label, metrics, nan_to_num, + beta) + + return ret_metrics + + +def total_area_to_metrics(total_area_intersect, + total_area_union, + total_area_pred_label, + total_area_label, + metrics=['mIoU'], + nan_to_num=None, + beta=1): + """Calculate evaluation metrics + Args: + total_area_intersect (ndarray): The intersection of prediction and + ground truth histogram on all classes. + total_area_union (ndarray): The union of prediction and ground truth + histogram on all classes. + total_area_pred_label (ndarray): The prediction histogram on all + classes. + total_area_label (ndarray): The ground truth histogram on all classes. + metrics (list[str] | str): Metrics to be evaluated, 'mIoU' and 'mDice'. + nan_to_num (int, optional): If specified, NaN values will be replaced + by the numbers defined by the user. Default: None. + Returns: + float: Overall accuracy on all images. + ndarray: Per category accuracy, shape (num_classes, ). + ndarray: Per category evaluation metrics, shape (num_classes, ). + """ + if isinstance(metrics, str): + metrics = [metrics] + allowed_metrics = ['mIoU', 'mDice', 'mFscore'] + if not set(metrics).issubset(set(allowed_metrics)): + raise KeyError('metrics {} is not supported'.format(metrics)) + + all_acc = total_area_intersect.sum() / total_area_label.sum() + ret_metrics = OrderedDict({'aAcc': all_acc}) + for metric in metrics: + if metric == 'mIoU': + iou = total_area_intersect / total_area_union + acc = total_area_intersect / total_area_label + ret_metrics['IoU'] = iou + ret_metrics['Acc'] = acc + elif metric == 'mDice': + dice = 2 * total_area_intersect / ( + total_area_pred_label + total_area_label) + acc = total_area_intersect / total_area_label + ret_metrics['Dice'] = dice + ret_metrics['Acc'] = acc + elif metric == 'mFscore': + precision = total_area_intersect / total_area_pred_label + recall = total_area_intersect / total_area_label + f_value = torch.tensor( + [f_score(x[0], x[1], beta) for x in zip(precision, recall)]) + ret_metrics['Fscore'] = f_value + ret_metrics['Precision'] = precision + ret_metrics['Recall'] = recall + + ret_metrics = { + metric: value.numpy() + for metric, value in ret_metrics.items() + } + if nan_to_num is not None: + ret_metrics = OrderedDict({ + metric: np.nan_to_num(metric_value, nan=nan_to_num) + for metric, metric_value in ret_metrics.items() + }) + return ret_metrics diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/optimizers/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/optimizers/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..4fbf4ecfcd4d1f0834322e2964b55d9637c844ba --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/optimizers/__init__.py @@ -0,0 +1,7 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .layer_decay_optimizer_constructor import ( + LayerDecayOptimizerConstructor, LearningRateDecayOptimizerConstructor) + +__all__ = [ + 'LearningRateDecayOptimizerConstructor', 'LayerDecayOptimizerConstructor' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/optimizers/layer_decay_optimizer_constructor.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/optimizers/layer_decay_optimizer_constructor.py new file mode 100644 index 0000000000000000000000000000000000000000..2b6b8ff9c97f4c65ac713bde8031454f3ba7c074 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/optimizers/layer_decay_optimizer_constructor.py @@ -0,0 +1,208 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import json +import warnings + +from mmcv.runner import DefaultOptimizerConstructor, get_dist_info + +from mmseg.utils import get_root_logger +from ..builder import OPTIMIZER_BUILDERS + + +def get_layer_id_for_convnext(var_name, max_layer_id): + """Get the layer id to set the different learning rates in ``layer_wise`` + decay_type. + + Args: + var_name (str): The key of the model. + max_layer_id (int): Maximum number of backbone layers. + + Returns: + int: The id number corresponding to different learning rate in + ``LearningRateDecayOptimizerConstructor``. + """ + + if var_name in ('backbone.cls_token', 'backbone.mask_token', + 'backbone.pos_embed'): + return 0 + elif var_name.startswith('backbone.downsample_layers'): + stage_id = int(var_name.split('.')[2]) + if stage_id == 0: + layer_id = 0 + elif stage_id == 1: + layer_id = 2 + elif stage_id == 2: + layer_id = 3 + elif stage_id == 3: + layer_id = max_layer_id + return layer_id + elif var_name.startswith('backbone.stages'): + stage_id = int(var_name.split('.')[2]) + block_id = int(var_name.split('.')[3]) + if stage_id == 0: + layer_id = 1 + elif stage_id == 1: + layer_id = 2 + elif stage_id == 2: + layer_id = 3 + block_id // 3 + elif stage_id == 3: + layer_id = max_layer_id + return layer_id + else: + return max_layer_id + 1 + + +def get_stage_id_for_convnext(var_name, max_stage_id): + """Get the stage id to set the different learning rates in ``stage_wise`` + decay_type. + + Args: + var_name (str): The key of the model. + max_stage_id (int): Maximum number of backbone layers. + + Returns: + int: The id number corresponding to different learning rate in + ``LearningRateDecayOptimizerConstructor``. + """ + + if var_name in ('backbone.cls_token', 'backbone.mask_token', + 'backbone.pos_embed'): + return 0 + elif var_name.startswith('backbone.downsample_layers'): + return 0 + elif var_name.startswith('backbone.stages'): + stage_id = int(var_name.split('.')[2]) + return stage_id + 1 + else: + return max_stage_id - 1 + + +def get_layer_id_for_vit(var_name, max_layer_id): + """Get the layer id to set the different learning rates. + + Args: + var_name (str): The key of the model. + num_max_layer (int): Maximum number of backbone layers. + + Returns: + int: Returns the layer id of the key. + """ + + if var_name in ('backbone.cls_token', 'backbone.mask_token', + 'backbone.pos_embed'): + return 0 + elif var_name.startswith('backbone.patch_embed'): + return 0 + elif var_name.startswith('backbone.layers'): + layer_id = int(var_name.split('.')[2]) + return layer_id + 1 + else: + return max_layer_id - 1 + + +@OPTIMIZER_BUILDERS.register_module() +class LearningRateDecayOptimizerConstructor(DefaultOptimizerConstructor): + """Different learning rates are set for different layers of backbone. + + Note: Currently, this optimizer constructor is built for ConvNeXt, + BEiT and MAE. + """ + + def add_params(self, params, module, **kwargs): + """Add all parameters of module to the params list. + + The parameters of the given module will be added to the list of param + groups, with specific rules defined by paramwise_cfg. + + Args: + params (list[dict]): A list of param groups, it will be modified + in place. + module (nn.Module): The module to be added. + """ + logger = get_root_logger() + + parameter_groups = {} + logger.info(f'self.paramwise_cfg is {self.paramwise_cfg}') + num_layers = self.paramwise_cfg.get('num_layers') + 2 + decay_rate = self.paramwise_cfg.get('decay_rate') + decay_type = self.paramwise_cfg.get('decay_type', 'layer_wise') + logger.info('Build LearningRateDecayOptimizerConstructor ' + f'{decay_type} {decay_rate} - {num_layers}') + weight_decay = self.base_wd + for name, param in module.named_parameters(): + if not param.requires_grad: + continue # frozen weights + if len(param.shape) == 1 or name.endswith('.bias') or name in ( + 'pos_embed', 'cls_token'): + group_name = 'no_decay' + this_weight_decay = 0. + else: + group_name = 'decay' + this_weight_decay = weight_decay + if 'layer_wise' in decay_type: + if 'ConvNeXt' in module.backbone.__class__.__name__: + layer_id = get_layer_id_for_convnext( + name, self.paramwise_cfg.get('num_layers')) + logger.info(f'set param {name} as id {layer_id}') + elif 'BEiT' in module.backbone.__class__.__name__ or \ + 'MAE' in module.backbone.__class__.__name__: + layer_id = get_layer_id_for_vit(name, num_layers) + logger.info(f'set param {name} as id {layer_id}') + else: + raise NotImplementedError() + elif decay_type == 'stage_wise': + if 'ConvNeXt' in module.backbone.__class__.__name__: + layer_id = get_stage_id_for_convnext(name, num_layers) + logger.info(f'set param {name} as id {layer_id}') + else: + raise NotImplementedError() + group_name = f'layer_{layer_id}_{group_name}' + + if group_name not in parameter_groups: + scale = decay_rate**(num_layers - layer_id - 1) + + parameter_groups[group_name] = { + 'weight_decay': this_weight_decay, + 'params': [], + 'param_names': [], + 'lr_scale': scale, + 'group_name': group_name, + 'lr': scale * self.base_lr, + } + + parameter_groups[group_name]['params'].append(param) + parameter_groups[group_name]['param_names'].append(name) + rank, _ = get_dist_info() + if rank == 0: + to_display = {} + for key in parameter_groups: + to_display[key] = { + 'param_names': parameter_groups[key]['param_names'], + 'lr_scale': parameter_groups[key]['lr_scale'], + 'lr': parameter_groups[key]['lr'], + 'weight_decay': parameter_groups[key]['weight_decay'], + } + logger.info(f'Param groups = {json.dumps(to_display, indent=2)}') + params.extend(parameter_groups.values()) + + +@OPTIMIZER_BUILDERS.register_module() +class LayerDecayOptimizerConstructor(LearningRateDecayOptimizerConstructor): + """Different learning rates are set for different layers of backbone. + + Note: Currently, this optimizer constructor is built for BEiT, + and it will be deprecated. + Please use ``LearningRateDecayOptimizerConstructor`` instead. + """ + + def __init__(self, optimizer_cfg, paramwise_cfg): + warnings.warn('DeprecationWarning: Original ' + 'LayerDecayOptimizerConstructor of BEiT ' + 'will be deprecated. Please use ' + 'LearningRateDecayOptimizerConstructor instead, ' + 'and set decay_type = layer_wise_vit in paramwise_cfg.') + paramwise_cfg.update({'decay_type': 'layer_wise_vit'}) + warnings.warn('DeprecationWarning: Layer_decay_rate will ' + 'be deleted, please use decay_rate instead.') + paramwise_cfg['decay_rate'] = paramwise_cfg.pop('layer_decay_rate') + super(LayerDecayOptimizerConstructor, + self).__init__(optimizer_cfg, paramwise_cfg) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..5206b96be6f87e99e8ae820bdd788444f4d255d9 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/__init__.py @@ -0,0 +1,5 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .builder import build_pixel_sampler +from .sampler import BasePixelSampler, OHEMPixelSampler + +__all__ = ['build_pixel_sampler', 'BasePixelSampler', 'OHEMPixelSampler'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/builder.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/builder.py new file mode 100644 index 0000000000000000000000000000000000000000..1cecd347bffb6ab289f27e0f9bbab91c3a5d4bd8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/builder.py @@ -0,0 +1,9 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.utils import Registry, build_from_cfg + +PIXEL_SAMPLERS = Registry('pixel sampler') + + +def build_pixel_sampler(cfg, **default_args): + """Build pixel sampler for segmentation map.""" + return build_from_cfg(cfg, PIXEL_SAMPLERS, default_args) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..5a7648564adbfcdb3e66f640e4f9c61de6e215e1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/__init__.py @@ -0,0 +1,5 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .base_pixel_sampler import BasePixelSampler +from .ohem_pixel_sampler import OHEMPixelSampler + +__all__ = ['BasePixelSampler', 'OHEMPixelSampler'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/base_pixel_sampler.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/base_pixel_sampler.py new file mode 100644 index 0000000000000000000000000000000000000000..03672cd478a2e464cc734ae92686c86f219da0a9 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/base_pixel_sampler.py @@ -0,0 +1,13 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from abc import ABCMeta, abstractmethod + + +class BasePixelSampler(metaclass=ABCMeta): + """Base class of pixel sampler.""" + + def __init__(self, **kwargs): + pass + + @abstractmethod + def sample(self, seg_logit, seg_label): + """Placeholder for sample function.""" diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/ohem_pixel_sampler.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/ohem_pixel_sampler.py new file mode 100644 index 0000000000000000000000000000000000000000..833a28768cd0bfddfc7ab59d3ba3cbe892b2fbb5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/seg/sampler/ohem_pixel_sampler.py @@ -0,0 +1,85 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +import torch.nn.functional as F + +from ..builder import PIXEL_SAMPLERS +from .base_pixel_sampler import BasePixelSampler + + +@PIXEL_SAMPLERS.register_module() +class OHEMPixelSampler(BasePixelSampler): + """Online Hard Example Mining Sampler for segmentation. + + Args: + context (nn.Module): The context of sampler, subclass of + :obj:`BaseDecodeHead`. + thresh (float, optional): The threshold for hard example selection. + Below which, are prediction with low confidence. If not + specified, the hard examples will be pixels of top ``min_kept`` + loss. Default: None. + min_kept (int, optional): The minimum number of predictions to keep. + Default: 100000. + """ + + def __init__(self, context, thresh=None, min_kept=100000): + super(OHEMPixelSampler, self).__init__() + self.context = context + assert min_kept > 1 + self.thresh = thresh + self.min_kept = min_kept + + def sample(self, seg_logit, seg_label): + """Sample pixels that have high loss or with low prediction confidence. + + Args: + seg_logit (torch.Tensor): segmentation logits, shape (N, C, H, W) + seg_label (torch.Tensor): segmentation label, shape (N, 1, H, W) + + Returns: + torch.Tensor: segmentation weight, shape (N, H, W) + """ + with torch.no_grad(): + assert seg_logit.shape[2:] == seg_label.shape[2:] + assert seg_label.shape[1] == 1 + seg_label = seg_label.squeeze(1).long() + batch_kept = self.min_kept * seg_label.size(0) + valid_mask = seg_label != self.context.ignore_index + seg_weight = seg_logit.new_zeros(size=seg_label.size()) + valid_seg_weight = seg_weight[valid_mask] + if self.thresh is not None: + seg_prob = F.softmax(seg_logit, dim=1) + + tmp_seg_label = seg_label.clone().unsqueeze(1) + tmp_seg_label[tmp_seg_label == self.context.ignore_index] = 0 + seg_prob = seg_prob.gather(1, tmp_seg_label).squeeze(1) + sort_prob, sort_indices = seg_prob[valid_mask].sort() + + if sort_prob.numel() > 0: + min_threshold = sort_prob[min(batch_kept, + sort_prob.numel() - 1)] + else: + min_threshold = 0.0 + threshold = max(min_threshold, self.thresh) + valid_seg_weight[seg_prob[valid_mask] < threshold] = 1. + else: + if not isinstance(self.context.loss_decode, nn.ModuleList): + losses_decode = [self.context.loss_decode] + else: + losses_decode = self.context.loss_decode + losses = 0.0 + for loss_module in losses_decode: + losses += loss_module( + seg_logit, + seg_label, + weight=None, + ignore_index=self.context.ignore_index, + reduction_override='none') + + # faster than topk according to https://github.com/pytorch/pytorch/issues/22812 # noqa + _, sort_indices = losses[valid_mask].sort(descending=True) + valid_seg_weight[sort_indices[:batch_kept]] = 1. + + seg_weight[valid_mask] = valid_seg_weight + + return seg_weight diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..28882893a53a78dcb7063e51b07273d30dd1c19f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/__init__.py @@ -0,0 +1,5 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .dist_util import check_dist_init, sync_random_seed +from .misc import add_prefix + +__all__ = ['add_prefix', 'check_dist_init', 'sync_random_seed'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/dist_util.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/dist_util.py new file mode 100644 index 0000000000000000000000000000000000000000..b3288519d0a8785db12c00da9d48e51de5ce3ba1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/dist_util.py @@ -0,0 +1,46 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import numpy as np +import torch +import torch.distributed as dist +from mmcv.runner import get_dist_info + + +def check_dist_init(): + return dist.is_available() and dist.is_initialized() + + +def sync_random_seed(seed=None, device='cuda'): + """Make sure different ranks share the same seed. All workers must call + this function, otherwise it will deadlock. This method is generally used in + `DistributedSampler`, because the seed should be identical across all + processes in the distributed group. + + In distributed sampling, different ranks should sample non-overlapped + data in the dataset. Therefore, this function is used to make sure that + each rank shuffles the data indices in the same order based + on the same seed. Then different ranks could use different indices + to select non-overlapped data from the same data list. + + Args: + seed (int, Optional): The seed. Default to None. + device (str): The device where the seed will be put on. + Default to 'cuda'. + Returns: + int: Seed to be used. + """ + + if seed is None: + seed = np.random.randint(2**31) + assert isinstance(seed, int) + + rank, world_size = get_dist_info() + + if world_size == 1: + return seed + + if rank == 0: + random_num = torch.tensor(seed, dtype=torch.int32, device=device) + else: + random_num = torch.tensor(0, dtype=torch.int32, device=device) + dist.broadcast(random_num, src=0) + return random_num.item() diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/misc.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/misc.py new file mode 100644 index 0000000000000000000000000000000000000000..282bb8d9698ebd19876d849f1e3fc2ee23e2d40d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/core/utils/misc.py @@ -0,0 +1,18 @@ +# Copyright (c) OpenMMLab. All rights reserved. +def add_prefix(inputs, prefix): + """Add prefix for dict. + + Args: + inputs (dict): The input dict with str keys. + prefix (str): The prefix to add. + + Returns: + + dict: The dict with keys updated with ``prefix``. + """ + + outputs = dict() + for name, value in inputs.items(): + outputs[f'{prefix}.{name}'] = value + + return outputs diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..3366f0aeccad857634c2702ab7df1bb4f2c684ae --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/__init__.py @@ -0,0 +1,10 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .ade import ADE20KDataset +from .builder import DATASETS, PIPELINES, build_dataloader, build_dataset +from .cityscapes import CityscapesDataset +from .coco_stuff import COCOStuffDataset +from .custom import CustomDataset +from .dataset_wrappers import (ConcatDataset, MultiImageMixDataset, + RepeatDataset) +from .pascal_context import PascalContextDataset, PascalContextDataset59 +from .voc import PascalVOCDataset diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/ade.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/ade.py new file mode 100644 index 0000000000000000000000000000000000000000..db94cebd3bbaed1dfee0f9a80f5a164a862de84f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/ade.py @@ -0,0 +1,167 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp + +import mmcv +import numpy as np +from PIL import Image + +from .builder import DATASETS +from .custom import CustomDataset + + +@DATASETS.register_module() +class ADE20KDataset(CustomDataset): + """ADE20K dataset. + + In segmentation map annotation for ADE20K, 0 stands for background, which + is not included in 150 categories. ``reduce_zero_label`` is fixed to True. + The ``img_suffix`` is fixed to '.jpg' and ``seg_map_suffix`` is fixed to + '.png'. + """ + CLASSES = ( + 'wall', 'building', 'sky', 'floor', 'tree', 'ceiling', 'road', 'bed ', + 'windowpane', 'grass', 'cabinet', 'sidewalk', 'person', 'earth', + 'door', 'table', 'mountain', 'plant', 'curtain', 'chair', 'car', + 'water', 'painting', 'sofa', 'shelf', 'house', 'sea', 'mirror', 'rug', + 'field', 'armchair', 'seat', 'fence', 'desk', 'rock', 'wardrobe', + 'lamp', 'bathtub', 'railing', 'cushion', 'base', 'box', 'column', + 'signboard', 'chest of drawers', 'counter', 'sand', 'sink', + 'skyscraper', 'fireplace', 'refrigerator', 'grandstand', 'path', + 'stairs', 'runway', 'case', 'pool table', 'pillow', 'screen door', + 'stairway', 'river', 'bridge', 'bookcase', 'blind', 'coffee table', + 'toilet', 'flower', 'book', 'hill', 'bench', 'countertop', 'stove', + 'palm', 'kitchen island', 'computer', 'swivel chair', 'boat', 'bar', + 'arcade machine', 'hovel', 'bus', 'towel', 'light', 'truck', 'tower', + 'chandelier', 'awning', 'streetlight', 'booth', 'television receiver', + 'airplane', 'dirt track', 'apparel', 'pole', 'land', 'bannister', + 'escalator', 'ottoman', 'bottle', 'buffet', 'poster', 'stage', 'van', + 'ship', 'fountain', 'conveyer belt', 'canopy', 'washer', 'plaything', + 'swimming pool', 'stool', 'barrel', 'basket', 'waterfall', 'tent', + 'bag', 'minibike', 'cradle', 'oven', 'ball', 'food', 'step', 'tank', + 'trade name', 'microwave', 'pot', 'animal', 'bicycle', 'lake', + 'dishwasher', 'screen', 'blanket', 'sculpture', 'hood', 'sconce', + 'vase', 'traffic light', 'tray', 'ashcan', 'fan', 'pier', 'crt screen', + 'plate', 'monitor', 'bulletin board', 'shower', 'radiator', 'glass', + 'clock', 'flag') + + PALETTE = [[120, 120, 120], [180, 120, 120], [6, 230, 230], [80, 50, 50], + [4, 200, 3], [120, 120, 80], [140, 140, 140], [204, 5, 255], + [230, 230, 230], [4, 250, 7], [224, 5, 255], [235, 255, 7], + [150, 5, 61], [120, 120, 70], [8, 255, 51], [255, 6, 82], + [143, 255, 140], [204, 255, 4], [255, 51, 7], [204, 70, 3], + [0, 102, 200], [61, 230, 250], [255, 6, 51], [11, 102, 255], + [255, 7, 71], [255, 9, 224], [9, 7, 230], [220, 220, 220], + [255, 9, 92], [112, 9, 255], [8, 255, 214], [7, 255, 224], + [255, 184, 6], [10, 255, 71], [255, 41, 10], [7, 255, 255], + [224, 255, 8], [102, 8, 255], [255, 61, 6], [255, 194, 7], + [255, 122, 8], [0, 255, 20], [255, 8, 41], [255, 5, 153], + [6, 51, 255], [235, 12, 255], [160, 150, 20], [0, 163, 255], + [140, 140, 140], [250, 10, 15], [20, 255, 0], [31, 255, 0], + [255, 31, 0], [255, 224, 0], [153, 255, 0], [0, 0, 255], + [255, 71, 0], [0, 235, 255], [0, 173, 255], [31, 0, 255], + [11, 200, 200], [255, 82, 0], [0, 255, 245], [0, 61, 255], + [0, 255, 112], [0, 255, 133], [255, 0, 0], [255, 163, 0], + [255, 102, 0], [194, 255, 0], [0, 143, 255], [51, 255, 0], + [0, 82, 255], [0, 255, 41], [0, 255, 173], [10, 0, 255], + [173, 255, 0], [0, 255, 153], [255, 92, 0], [255, 0, 255], + [255, 0, 245], [255, 0, 102], [255, 173, 0], [255, 0, 20], + [255, 184, 184], [0, 31, 255], [0, 255, 61], [0, 71, 255], + [255, 0, 204], [0, 255, 194], [0, 255, 82], [0, 10, 255], + [0, 112, 255], [51, 0, 255], [0, 194, 255], [0, 122, 255], + [0, 255, 163], [255, 153, 0], [0, 255, 10], [255, 112, 0], + [143, 255, 0], [82, 0, 255], [163, 255, 0], [255, 235, 0], + [8, 184, 170], [133, 0, 255], [0, 255, 92], [184, 0, 255], + [255, 0, 31], [0, 184, 255], [0, 214, 255], [255, 0, 112], + [92, 255, 0], [0, 224, 255], [112, 224, 255], [70, 184, 160], + [163, 0, 255], [153, 0, 255], [71, 255, 0], [255, 0, 163], + [255, 204, 0], [255, 0, 143], [0, 255, 235], [133, 255, 0], + [255, 0, 235], [245, 0, 255], [255, 0, 122], [255, 245, 0], + [10, 190, 212], [214, 255, 0], [0, 204, 255], [20, 0, 255], + [255, 255, 0], [0, 153, 255], [0, 41, 255], [0, 255, 204], + [41, 0, 255], [41, 255, 0], [173, 0, 255], [0, 245, 255], + [71, 0, 255], [122, 0, 255], [0, 255, 184], [0, 92, 255], + [184, 255, 0], [0, 133, 255], [255, 214, 0], [25, 194, 194], + [102, 255, 0], [92, 0, 255]] + + def __init__(self, **kwargs): + super(ADE20KDataset, self).__init__( + img_suffix='.jpg', + seg_map_suffix='.png', + reduce_zero_label=True, + **kwargs) + + def results2img(self, results, imgfile_prefix, to_label_id, indices=None): + """Write the segmentation results to images. + + Args: + results (list[ndarray]): Testing results of the + dataset. + imgfile_prefix (str): The filename prefix of the png files. + If the prefix is "somepath/xxx", + the png files will be named "somepath/xxx.png". + to_label_id (bool): whether convert output to label_id for + submission. + indices (list[int], optional): Indices of input results, if not + set, all the indices of the dataset will be used. + Default: None. + + Returns: + list[str: str]: result txt files which contains corresponding + semantic segmentation images. + """ + if indices is None: + indices = list(range(len(self))) + + mmcv.mkdir_or_exist(imgfile_prefix) + result_files = [] + for result, idx in zip(results, indices): + + filename = self.img_infos[idx]['filename'] + basename = osp.splitext(osp.basename(filename))[0] + + png_filename = osp.join(imgfile_prefix, f'{basename}.png') + + # The index range of official requirement is from 0 to 150. + # But the index range of output is from 0 to 149. + # That is because we set reduce_zero_label=True. + result = result + 1 + + output = Image.fromarray(result.astype(np.uint8)) + output.save(png_filename) + result_files.append(png_filename) + + return result_files + + def format_results(self, + results, + imgfile_prefix, + to_label_id=True, + indices=None): + """Format the results into dir (standard format for ade20k evaluation). + + Args: + results (list): Testing results of the dataset. + imgfile_prefix (str | None): The prefix of images files. It + includes the file path and the prefix of filename, e.g., + "a/b/prefix". + to_label_id (bool): whether convert output to label_id for + submission. Default: False + indices (list[int], optional): Indices of input results, if not + set, all the indices of the dataset will be used. + Default: None. + + Returns: + tuple: (result_files, tmp_dir), result_files is a list containing + the image paths, tmp_dir is the temporal directory created + for saving json/png files when img_prefix is not specified. + """ + + if indices is None: + indices = list(range(len(self))) + + assert isinstance(results, list), 'results must be a list.' + assert isinstance(indices, list), 'indices must be a list.' + + result_files = self.results2img(results, imgfile_prefix, to_label_id, + indices) + return result_files diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/builder.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/builder.py new file mode 100644 index 0000000000000000000000000000000000000000..4d852d3653ca4df4fbe2495c096775e6a42c7ac6 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/builder.py @@ -0,0 +1,191 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy +import platform +import random +from functools import partial + +import numpy as np +import torch +from mmcv.parallel import collate +from mmcv.runner import get_dist_info +from mmcv.utils import Registry, build_from_cfg, digit_version +from torch.utils.data import DataLoader + +from .samplers import DistributedSampler + +if platform.system() != 'Windows': + # https://github.com/pytorch/pytorch/issues/973 + import resource + rlimit = resource.getrlimit(resource.RLIMIT_NOFILE) + base_soft_limit = rlimit[0] + hard_limit = rlimit[1] + soft_limit = min(max(4096, base_soft_limit), hard_limit) + resource.setrlimit(resource.RLIMIT_NOFILE, (soft_limit, hard_limit)) + +DATASETS = Registry('dataset') +PIPELINES = Registry('pipeline') + + +def _concat_dataset(cfg, default_args=None): + """Build :obj:`ConcatDataset by.""" + from .dataset_wrappers import ConcatDataset + img_dir = cfg['img_dir'] + ann_dir = cfg.get('ann_dir', None) + split = cfg.get('split', None) + # pop 'separate_eval' since it is not a valid key for common datasets. + separate_eval = cfg.pop('separate_eval', True) + num_img_dir = len(img_dir) if isinstance(img_dir, (list, tuple)) else 1 + if ann_dir is not None: + num_ann_dir = len(ann_dir) if isinstance(ann_dir, (list, tuple)) else 1 + else: + num_ann_dir = 0 + if split is not None: + num_split = len(split) if isinstance(split, (list, tuple)) else 1 + else: + num_split = 0 + if num_img_dir > 1: + assert num_img_dir == num_ann_dir or num_ann_dir == 0 + assert num_img_dir == num_split or num_split == 0 + else: + assert num_split == num_ann_dir or num_ann_dir <= 1 + num_dset = max(num_split, num_img_dir) + + datasets = [] + for i in range(num_dset): + data_cfg = copy.deepcopy(cfg) + if isinstance(img_dir, (list, tuple)): + data_cfg['img_dir'] = img_dir[i] + if isinstance(ann_dir, (list, tuple)): + data_cfg['ann_dir'] = ann_dir[i] + if isinstance(split, (list, tuple)): + data_cfg['split'] = split[i] + datasets.append(build_dataset(data_cfg, default_args)) + + return ConcatDataset(datasets, separate_eval) + + +def build_dataset(cfg, default_args=None): + """Build datasets.""" + from .dataset_wrappers import (ConcatDataset, MultiImageMixDataset, + RepeatDataset) + if isinstance(cfg, (list, tuple)): + dataset = ConcatDataset([build_dataset(c, default_args) for c in cfg]) + elif cfg['type'] == 'RepeatDataset': + dataset = RepeatDataset( + build_dataset(cfg['dataset'], default_args), cfg['times']) + elif cfg['type'] == 'MultiImageMixDataset': + cp_cfg = copy.deepcopy(cfg) + cp_cfg['dataset'] = build_dataset(cp_cfg['dataset']) + cp_cfg.pop('type') + dataset = MultiImageMixDataset(**cp_cfg) + elif isinstance(cfg.get('img_dir'), (list, tuple)) or isinstance( + cfg.get('split', None), (list, tuple)): + dataset = _concat_dataset(cfg, default_args) + else: + dataset = build_from_cfg(cfg, DATASETS, default_args) + + return dataset + + +def build_dataloader(dataset, + samples_per_gpu, + workers_per_gpu, + num_gpus=1, + dist=True, + shuffle=True, + seed=None, + drop_last=False, + pin_memory=True, + persistent_workers=True, + **kwargs): + """Build PyTorch DataLoader. + + In distributed training, each GPU/process has a dataloader. + In non-distributed training, there is only one dataloader for all GPUs. + + Args: + dataset (Dataset): A PyTorch dataset. + samples_per_gpu (int): Number of training samples on each GPU, i.e., + batch size of each GPU. + workers_per_gpu (int): How many subprocesses to use for data loading + for each GPU. + num_gpus (int): Number of GPUs. Only used in non-distributed training. + dist (bool): Distributed training/test or not. Default: True. + shuffle (bool): Whether to shuffle the data at every epoch. + Default: True. + seed (int | None): Seed to be used. Default: None. + drop_last (bool): Whether to drop the last incomplete batch in epoch. + Default: False + pin_memory (bool): Whether to use pin_memory in DataLoader. + Default: True + persistent_workers (bool): If True, the data loader will not shutdown + the worker processes after a dataset has been consumed once. + This allows to maintain the workers Dataset instances alive. + The argument also has effect in PyTorch>=1.7.0. + Default: True + kwargs: any keyword argument to be used to initialize DataLoader + + Returns: + DataLoader: A PyTorch dataloader. + """ + rank, world_size = get_dist_info() + if dist: + sampler = DistributedSampler( + dataset, world_size, rank, shuffle=shuffle, seed=seed) + shuffle = False + batch_size = samples_per_gpu + num_workers = workers_per_gpu + else: + sampler = None + batch_size = num_gpus * samples_per_gpu + num_workers = num_gpus * workers_per_gpu + + init_fn = partial( + worker_init_fn, num_workers=num_workers, rank=rank, + seed=seed) if seed is not None else None + + if digit_version(torch.__version__) >= digit_version('1.8.0'): + data_loader = DataLoader( + dataset, + batch_size=batch_size, + sampler=sampler, + num_workers=num_workers, + collate_fn=partial(collate, samples_per_gpu=samples_per_gpu), + pin_memory=pin_memory, + shuffle=shuffle, + worker_init_fn=init_fn, + drop_last=drop_last, + persistent_workers=persistent_workers, + **kwargs) + else: + data_loader = DataLoader( + dataset, + batch_size=batch_size, + sampler=sampler, + num_workers=num_workers, + collate_fn=partial(collate, samples_per_gpu=samples_per_gpu), + pin_memory=pin_memory, + shuffle=shuffle, + worker_init_fn=init_fn, + drop_last=drop_last, + **kwargs) + + return data_loader + + +def worker_init_fn(worker_id, num_workers, rank, seed): + """Worker init func for dataloader. + + The seed of each worker equals to num_worker * rank + worker_id + user_seed + + Args: + worker_id (int): Worker id. + num_workers (int): Number of workers. + rank (int): The rank of current process. + seed (int): The random seed to use. + """ + + worker_seed = num_workers * rank + worker_id + seed + np.random.seed(worker_seed) + random.seed(worker_seed) + torch.manual_seed(worker_seed) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/cityscapes.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/cityscapes.py new file mode 100644 index 0000000000000000000000000000000000000000..ed633d00db33789541284df0d2ec3187d4dd01a3 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/cityscapes.py @@ -0,0 +1,214 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp + +import mmcv +import numpy as np +from mmcv.utils import print_log +from PIL import Image + +from .builder import DATASETS +from .custom import CustomDataset + + +@DATASETS.register_module() +class CityscapesDataset(CustomDataset): + """Cityscapes dataset. + + The ``img_suffix`` is fixed to '_leftImg8bit.png' and ``seg_map_suffix`` is + fixed to '_gtFine_labelTrainIds.png' for Cityscapes dataset. + """ + + CLASSES = ('road', 'sidewalk', 'building', 'wall', 'fence', 'pole', + 'traffic light', 'traffic sign', 'vegetation', 'terrain', 'sky', + 'person', 'rider', 'car', 'truck', 'bus', 'train', 'motorcycle', + 'bicycle') + + PALETTE = [[128, 64, 128], [244, 35, 232], [70, 70, 70], [102, 102, 156], + [190, 153, 153], [153, 153, 153], [250, 170, 30], [220, 220, 0], + [107, 142, 35], [152, 251, 152], [70, 130, 180], [220, 20, 60], + [255, 0, 0], [0, 0, 142], [0, 0, 70], [0, 60, 100], + [0, 80, 100], [0, 0, 230], [119, 11, 32]] + + def __init__(self, + img_suffix='_leftImg8bit.png', + seg_map_suffix='_gtFine_labelTrainIds.png', + **kwargs): + super(CityscapesDataset, self).__init__( + img_suffix=img_suffix, seg_map_suffix=seg_map_suffix, **kwargs) + + @staticmethod + def _convert_to_label_id(result): + """Convert trainId to id for cityscapes.""" + if isinstance(result, str): + result = np.load(result) + import cityscapesscripts.helpers.labels as CSLabels + result_copy = result.copy() + for trainId, label in CSLabels.trainId2label.items(): + result_copy[result == trainId] = label.id + + return result_copy + + def results2img(self, results, imgfile_prefix, to_label_id, indices=None): + """Write the segmentation results to images. + + Args: + results (list[ndarray]): Testing results of the + dataset. + imgfile_prefix (str): The filename prefix of the png files. + If the prefix is "somepath/xxx", + the png files will be named "somepath/xxx.png". + to_label_id (bool): whether convert output to label_id for + submission. + indices (list[int], optional): Indices of input results, + if not set, all the indices of the dataset will be used. + Default: None. + + Returns: + list[str: str]: result txt files which contains corresponding + semantic segmentation images. + """ + if indices is None: + indices = list(range(len(self))) + + mmcv.mkdir_or_exist(imgfile_prefix) + result_files = [] + for result, idx in zip(results, indices): + if to_label_id: + result = self._convert_to_label_id(result) + filename = self.img_infos[idx]['filename'] + basename = osp.splitext(osp.basename(filename))[0] + + png_filename = osp.join(imgfile_prefix, f'{basename}.png') + + output = Image.fromarray(result.astype(np.uint8)).convert('P') + import cityscapesscripts.helpers.labels as CSLabels + palette = np.zeros((len(CSLabels.id2label), 3), dtype=np.uint8) + for label_id, label in CSLabels.id2label.items(): + palette[label_id] = label.color + + output.putpalette(palette) + output.save(png_filename) + result_files.append(png_filename) + + return result_files + + def format_results(self, + results, + imgfile_prefix, + to_label_id=True, + indices=None): + """Format the results into dir (standard format for Cityscapes + evaluation). + + Args: + results (list): Testing results of the dataset. + imgfile_prefix (str): The prefix of images files. It + includes the file path and the prefix of filename, e.g., + "a/b/prefix". + to_label_id (bool): whether convert output to label_id for + submission. Default: False + indices (list[int], optional): Indices of input results, + if not set, all the indices of the dataset will be used. + Default: None. + + Returns: + tuple: (result_files, tmp_dir), result_files is a list containing + the image paths, tmp_dir is the temporal directory created + for saving json/png files when img_prefix is not specified. + """ + if indices is None: + indices = list(range(len(self))) + + assert isinstance(results, list), 'results must be a list.' + assert isinstance(indices, list), 'indices must be a list.' + + result_files = self.results2img(results, imgfile_prefix, to_label_id, + indices) + + return result_files + + def evaluate(self, + results, + metric='mIoU', + logger=None, + imgfile_prefix=None): + """Evaluation in Cityscapes/default protocol. + + Args: + results (list): Testing results of the dataset. + metric (str | list[str]): Metrics to be evaluated. + logger (logging.Logger | None | str): Logger used for printing + related information during evaluation. Default: None. + imgfile_prefix (str | None): The prefix of output image file, + for cityscapes evaluation only. It includes the file path and + the prefix of filename, e.g., "a/b/prefix". + If results are evaluated with cityscapes protocol, it would be + the prefix of output png files. The output files would be + png images under folder "a/b/prefix/xxx.png", where "xxx" is + the image name of cityscapes. If not specified, a temp file + will be created for evaluation. + Default: None. + + Returns: + dict[str, float]: Cityscapes/default metrics. + """ + + eval_results = dict() + metrics = metric.copy() if isinstance(metric, list) else [metric] + if 'cityscapes' in metrics: + eval_results.update( + self._evaluate_cityscapes(results, logger, imgfile_prefix)) + metrics.remove('cityscapes') + if len(metrics) > 0: + eval_results.update( + super(CityscapesDataset, + self).evaluate(results, metrics, logger)) + + return eval_results + + def _evaluate_cityscapes(self, results, logger, imgfile_prefix): + """Evaluation in Cityscapes protocol. + + Args: + results (list): Testing results of the dataset. + logger (logging.Logger | str | None): Logger used for printing + related information during evaluation. Default: None. + imgfile_prefix (str | None): The prefix of output image file + + Returns: + dict[str: float]: Cityscapes evaluation results. + """ + try: + import cityscapesscripts.evaluation.evalPixelLevelSemanticLabeling as CSEval # noqa + except ImportError: + raise ImportError('Please run "pip install cityscapesscripts" to ' + 'install cityscapesscripts first.') + msg = 'Evaluating in Cityscapes style' + if logger is None: + msg = '\n' + msg + print_log(msg, logger=logger) + + result_dir = imgfile_prefix + + eval_results = dict() + print_log(f'Evaluating results under {result_dir} ...', logger=logger) + + CSEval.args.evalInstLevelScore = True + CSEval.args.predictionPath = osp.abspath(result_dir) + CSEval.args.evalPixelAccuracy = True + CSEval.args.JSONOutput = False + + seg_map_list = [] + pred_list = [] + + # when evaluating with official cityscapesscripts, + # **_gtFine_labelIds.png is used + for seg_map in mmcv.scandir( + self.ann_dir, 'gtFine_labelIds.png', recursive=True): + seg_map_list.append(osp.join(self.ann_dir, seg_map)) + pred_list.append(CSEval.getPrediction(CSEval.args, seg_map)) + + eval_results.update( + CSEval.evaluateImgLists(pred_list, seg_map_list, CSEval.args)) + + return eval_results diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/coco_stuff.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/coco_stuff.py new file mode 100644 index 0000000000000000000000000000000000000000..24d089556599a5696c50fd0115077fbc83413061 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/coco_stuff.py @@ -0,0 +1,94 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .builder import DATASETS +from .custom import CustomDataset + + +@DATASETS.register_module() +class COCOStuffDataset(CustomDataset): + """COCO-Stuff dataset. + + In segmentation map annotation for COCO-Stuff, Train-IDs of the 10k version + are from 1 to 171, where 0 is the ignore index, and Train-ID of COCO Stuff + 164k is from 0 to 170, where 255 is the ignore index. So, they are all 171 + semantic categories. ``reduce_zero_label`` is set to True and False for the + 10k and 164k versions, respectively. The ``img_suffix`` is fixed to '.jpg', + and ``seg_map_suffix`` is fixed to '.png'. + """ + CLASSES = ( + 'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', + 'truck', 'boat', 'traffic light', 'fire hydrant', 'stop sign', + 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', + 'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', + 'handbag', 'tie', 'suitcase', 'frisbee', 'skis', 'snowboard', + 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', + 'surfboard', 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', + 'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange', + 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', + 'couch', 'potted plant', 'bed', 'dining table', 'toilet', 'tv', + 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', + 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', + 'scissors', 'teddy bear', 'hair drier', 'toothbrush', 'banner', + 'blanket', 'branch', 'bridge', 'building-other', 'bush', 'cabinet', + 'cage', 'cardboard', 'carpet', 'ceiling-other', 'ceiling-tile', + 'cloth', 'clothes', 'clouds', 'counter', 'cupboard', 'curtain', + 'desk-stuff', 'dirt', 'door-stuff', 'fence', 'floor-marble', + 'floor-other', 'floor-stone', 'floor-tile', 'floor-wood', + 'flower', 'fog', 'food-other', 'fruit', 'furniture-other', 'grass', + 'gravel', 'ground-other', 'hill', 'house', 'leaves', 'light', 'mat', + 'metal', 'mirror-stuff', 'moss', 'mountain', 'mud', 'napkin', 'net', + 'paper', 'pavement', 'pillow', 'plant-other', 'plastic', 'platform', + 'playingfield', 'railing', 'railroad', 'river', 'road', 'rock', 'roof', + 'rug', 'salad', 'sand', 'sea', 'shelf', 'sky-other', 'skyscraper', + 'snow', 'solid-other', 'stairs', 'stone', 'straw', 'structural-other', + 'table', 'tent', 'textile-other', 'towel', 'tree', 'vegetable', + 'wall-brick', 'wall-concrete', 'wall-other', 'wall-panel', + 'wall-stone', 'wall-tile', 'wall-wood', 'water-other', 'waterdrops', + 'window-blind', 'window-other', 'wood') + + PALETTE = [[0, 192, 64], [0, 192, 64], [0, 64, 96], [128, 192, 192], + [0, 64, 64], [0, 192, 224], [0, 192, 192], [128, 192, 64], + [0, 192, 96], [128, 192, 64], [128, 32, 192], [0, 0, 224], + [0, 0, 64], [0, 160, 192], [128, 0, 96], [128, 0, 192], + [0, 32, 192], [128, 128, 224], [0, 0, 192], [128, 160, 192], + [128, 128, 0], [128, 0, 32], [128, 32, 0], [128, 0, 128], + [64, 128, 32], [0, 160, 0], [0, 0, 0], [192, 128, 160], + [0, 32, 0], [0, 128, 128], [64, 128, 160], [128, 160, 0], + [0, 128, 0], [192, 128, 32], [128, 96, 128], [0, 0, 128], + [64, 0, 32], [0, 224, 128], [128, 0, 0], [192, 0, 160], + [0, 96, 128], [128, 128, 128], [64, 0, 160], [128, 224, 128], + [128, 128, 64], [192, 0, 32], [128, 96, 0], [128, 0, 192], + [0, 128, 32], [64, 224, 0], [0, 0, 64], [128, 128, 160], + [64, 96, 0], [0, 128, 192], [0, 128, 160], [192, 224, 0], + [0, 128, 64], [128, 128, 32], [192, 32, 128], [0, 64, 192], + [0, 0, 32], [64, 160, 128], [128, 64, 64], [128, 0, 160], + [64, 32, 128], [128, 192, 192], [0, 0, 160], [192, 160, 128], + [128, 192, 0], [128, 0, 96], [192, 32, 0], [128, 64, 128], + [64, 128, 96], [64, 160, 0], [0, 64, 0], [192, 128, 224], + [64, 32, 0], [0, 192, 128], [64, 128, 224], [192, 160, 0], + [0, 192, 0], [192, 128, 96], [192, 96, 128], [0, 64, 128], + [64, 0, 96], [64, 224, 128], [128, 64, 0], [192, 0, 224], + [64, 96, 128], [128, 192, 128], [64, 0, 224], [192, 224, 128], + [128, 192, 64], [192, 0, 96], [192, 96, 0], [128, 64, 192], + [0, 128, 96], [0, 224, 0], [64, 64, 64], [128, 128, 224], + [0, 96, 0], [64, 192, 192], [0, 128, 224], [128, 224, 0], + [64, 192, 64], [128, 128, 96], [128, 32, 128], [64, 0, 192], + [0, 64, 96], [0, 160, 128], [192, 0, 64], [128, 64, 224], + [0, 32, 128], [192, 128, 192], [0, 64, 224], [128, 160, 128], + [192, 128, 0], [128, 64, 32], [128, 32, 64], [192, 0, 128], + [64, 192, 32], [0, 160, 64], [64, 0, 0], [192, 192, 160], + [0, 32, 64], [64, 128, 128], [64, 192, 160], [128, 160, 64], + [64, 128, 0], [192, 192, 32], [128, 96, 192], [64, 0, 128], + [64, 64, 32], [0, 224, 192], [192, 0, 0], [192, 64, 160], + [0, 96, 192], [192, 128, 128], [64, 64, 160], [128, 224, 192], + [192, 128, 64], [192, 64, 32], [128, 96, 64], [192, 0, 192], + [0, 192, 32], [64, 224, 64], [64, 0, 64], [128, 192, 160], + [64, 96, 64], [64, 128, 192], [0, 192, 160], [192, 224, 64], + [64, 128, 64], [128, 192, 32], [192, 32, 192], [64, 64, 192], + [0, 64, 32], [64, 160, 192], [192, 64, 64], [128, 64, 160], + [64, 32, 192], [192, 192, 192], [0, 64, 160], [192, 160, 192], + [192, 192, 0], [128, 64, 96], [192, 32, 64], [192, 64, 128], + [64, 192, 96], [64, 160, 64], [64, 64, 0]] + + def __init__(self, **kwargs): + super(COCOStuffDataset, self).__init__( + img_suffix='.jpg', seg_map_suffix='_labelTrainIds.png', **kwargs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/custom.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/custom.py new file mode 100644 index 0000000000000000000000000000000000000000..4615d4114ec0a1ff8a2771999ef8c08fd95c9743 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/custom.py @@ -0,0 +1,487 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp +import warnings +from collections import OrderedDict + +import mmcv +import numpy as np +from mmcv.utils import print_log +from prettytable import PrettyTable +from torch.utils.data import Dataset + +from mmseg.core import eval_metrics, intersect_and_union, pre_eval_to_metrics +from mmseg.utils import get_root_logger +from .builder import DATASETS +from .pipelines import Compose, LoadAnnotations + + +@DATASETS.register_module() +class CustomDataset(Dataset): + """Custom dataset for semantic segmentation. An example of file structure + is as followed. + + .. code-block:: none + + ├── data + │ ├── my_dataset + │ │ ├── img_dir + │ │ │ ├── train + │ │ │ │ ├── xxx{img_suffix} + │ │ │ │ ├── yyy{img_suffix} + │ │ │ │ ├── zzz{img_suffix} + │ │ │ ├── val + │ │ ├── ann_dir + │ │ │ ├── train + │ │ │ │ ├── xxx{seg_map_suffix} + │ │ │ │ ├── yyy{seg_map_suffix} + │ │ │ │ ├── zzz{seg_map_suffix} + │ │ │ ├── val + + The img/gt_semantic_seg pair of CustomDataset should be of the same + except suffix. A valid img/gt_semantic_seg filename pair should be like + ``xxx{img_suffix}`` and ``xxx{seg_map_suffix}`` (extension is also included + in the suffix). If split is given, then ``xxx`` is specified in txt file. + Otherwise, all files in ``img_dir/``and ``ann_dir`` will be loaded. + Please refer to ``docs/en/tutorials/new_dataset.md`` for more details. + + + Args: + pipeline (list[dict]): Processing pipeline + img_dir (str): Path to image directory + img_suffix (str): Suffix of images. Default: '.jpg' + ann_dir (str, optional): Path to annotation directory. Default: None + seg_map_suffix (str): Suffix of segmentation maps. Default: '.png' + split (str, optional): Split txt file. If split is specified, only + file with suffix in the splits will be loaded. Otherwise, all + images in img_dir/ann_dir will be loaded. Default: None + data_root (str, optional): Data root for img_dir/ann_dir. Default: + None. + test_mode (bool): If test_mode=True, gt wouldn't be loaded. + ignore_index (int): The label index to be ignored. Default: 255 + reduce_zero_label (bool): Whether to mark label zero as ignored. + Default: False + classes (str | Sequence[str], optional): Specify classes to load. + If is None, ``cls.CLASSES`` will be used. Default: None. + palette (Sequence[Sequence[int]]] | np.ndarray | None): + The palette of segmentation map. If None is given, and + self.PALETTE is None, random palette will be generated. + Default: None + gt_seg_map_loader_cfg (dict, optional): build LoadAnnotations to + load gt for evaluation, load from disk by default. Default: None. + file_client_args (dict): Arguments to instantiate a FileClient. + See :class:`mmcv.fileio.FileClient` for details. + Defaults to ``dict(backend='disk')``. + """ + + CLASSES = None + + PALETTE = None + + def __init__(self, + pipeline, + img_dir, + img_suffix='.jpg', + ann_dir=None, + seg_map_suffix='.png', + split=None, + data_root=None, + test_mode=False, + ignore_index=255, + reduce_zero_label=False, + classes=None, + palette=None, + gt_seg_map_loader_cfg=None, + file_client_args=dict(backend='disk')): + self.pipeline = Compose(pipeline) + self.img_dir = img_dir + self.img_suffix = img_suffix + self.ann_dir = ann_dir + self.seg_map_suffix = seg_map_suffix + self.split = split + self.data_root = data_root + self.test_mode = test_mode + self.ignore_index = ignore_index + self.reduce_zero_label = reduce_zero_label + self.label_map = None + self.CLASSES, self.PALETTE = self.get_classes_and_palette( + classes, palette) + self.gt_seg_map_loader = LoadAnnotations( + ) if gt_seg_map_loader_cfg is None else LoadAnnotations( + **gt_seg_map_loader_cfg) + + self.file_client_args = file_client_args + self.file_client = mmcv.FileClient.infer_client(self.file_client_args) + + if test_mode: + assert self.CLASSES is not None, \ + '`cls.CLASSES` or `classes` should be specified when testing' + + # join paths if data_root is specified + if self.data_root is not None: + if not osp.isabs(self.img_dir): + self.img_dir = osp.join(self.data_root, self.img_dir) + if not (self.ann_dir is None or osp.isabs(self.ann_dir)): + self.ann_dir = osp.join(self.data_root, self.ann_dir) + if not (self.split is None or osp.isabs(self.split)): + self.split = osp.join(self.data_root, self.split) + + # load annotations + self.img_infos = self.load_annotations(self.img_dir, self.img_suffix, + self.ann_dir, + self.seg_map_suffix, self.split) + + def __len__(self): + """Total number of samples of data.""" + return len(self.img_infos) + + def load_annotations(self, img_dir, img_suffix, ann_dir, seg_map_suffix, + split): + """Load annotation from directory. + + Args: + img_dir (str): Path to image directory + img_suffix (str): Suffix of images. + ann_dir (str|None): Path to annotation directory. + seg_map_suffix (str|None): Suffix of segmentation maps. + split (str|None): Split txt file. If split is specified, only file + with suffix in the splits will be loaded. Otherwise, all images + in img_dir/ann_dir will be loaded. Default: None + + Returns: + list[dict]: All image info of dataset. + """ + + img_infos = [] + if split is not None: + lines = mmcv.list_from_file( + split, file_client_args=self.file_client_args) + for line in lines: + img_name = line.strip() + img_info = dict(filename=img_name + img_suffix) + if ann_dir is not None: + seg_map = img_name + seg_map_suffix + img_info['ann'] = dict(seg_map=seg_map) + img_infos.append(img_info) + else: + for img in self.file_client.list_dir_or_file( + dir_path=img_dir, + list_dir=False, + suffix=img_suffix, + recursive=True): + img_info = dict(filename=img) + if ann_dir is not None: + seg_map = img.replace(img_suffix, seg_map_suffix) + img_info['ann'] = dict(seg_map=seg_map) + img_infos.append(img_info) + img_infos = sorted(img_infos, key=lambda x: x['filename']) + + print_log(f'Loaded {len(img_infos)} images', logger=get_root_logger()) + return img_infos + + def get_ann_info(self, idx): + """Get annotation by index. + + Args: + idx (int): Index of data. + + Returns: + dict: Annotation info of specified index. + """ + + return self.img_infos[idx]['ann'] + + def pre_pipeline(self, results): + """Prepare results dict for pipeline.""" + results['seg_fields'] = [] + results['img_prefix'] = self.img_dir + results['seg_prefix'] = self.ann_dir + if self.custom_classes: + results['label_map'] = self.label_map + + def __getitem__(self, idx): + """Get training/test data after pipeline. + + Args: + idx (int): Index of data. + + Returns: + dict: Training/test data (with annotation if `test_mode` is set + False). + """ + + if self.test_mode: + return self.prepare_test_img(idx) + else: + return self.prepare_train_img(idx) + + def prepare_train_img(self, idx): + """Get training data and annotations after pipeline. + + Args: + idx (int): Index of data. + + Returns: + dict: Training data and annotation after pipeline with new keys + introduced by pipeline. + """ + + img_info = self.img_infos[idx] + ann_info = self.get_ann_info(idx) + results = dict(img_info=img_info, ann_info=ann_info) + self.pre_pipeline(results) + return self.pipeline(results) + + def prepare_test_img(self, idx): + """Get testing data after pipeline. + + Args: + idx (int): Index of data. + + Returns: + dict: Testing data after pipeline with new keys introduced by + pipeline. + """ + + img_info = self.img_infos[idx] + results = dict(img_info=img_info) + self.pre_pipeline(results) + return self.pipeline(results) + + def format_results(self, results, imgfile_prefix, indices=None, **kwargs): + """Place holder to format result to dataset specific output.""" + raise NotImplementedError + + def get_gt_seg_map_by_idx(self, index): + """Get one ground truth segmentation map for evaluation.""" + ann_info = self.get_ann_info(index) + results = dict(ann_info=ann_info) + self.pre_pipeline(results) + self.gt_seg_map_loader(results) + return results['gt_semantic_seg'] + + def get_gt_seg_maps(self, efficient_test=None): + """Get ground truth segmentation maps for evaluation.""" + if efficient_test is not None: + warnings.warn( + 'DeprecationWarning: ``efficient_test`` has been deprecated ' + 'since MMSeg v0.16, the ``get_gt_seg_maps()`` is CPU memory ' + 'friendly by default. ') + + for idx in range(len(self)): + ann_info = self.get_ann_info(idx) + results = dict(ann_info=ann_info) + self.pre_pipeline(results) + self.gt_seg_map_loader(results) + yield results['gt_semantic_seg'] + + def pre_eval(self, preds, indices): + """Collect eval result from each iteration. + + Args: + preds (list[torch.Tensor] | torch.Tensor): the segmentation logit + after argmax, shape (N, H, W). + indices (list[int] | int): the prediction related ground truth + indices. + + Returns: + list[torch.Tensor]: (area_intersect, area_union, area_prediction, + area_ground_truth). + """ + # In order to compat with batch inference + if not isinstance(indices, list): + indices = [indices] + if not isinstance(preds, list): + preds = [preds] + + pre_eval_results = [] + + for pred, index in zip(preds, indices): + seg_map = self.get_gt_seg_map_by_idx(index) + pre_eval_results.append( + intersect_and_union( + pred, + seg_map, + len(self.CLASSES), + self.ignore_index, + # as the labels has been converted when dataset initialized + # in `get_palette_for_custom_classes ` this `label_map` + # should be `dict()`, see + # https://github.com/open-mmlab/mmsegmentation/issues/1415 + # for more ditails + label_map=dict(), + reduce_zero_label=self.reduce_zero_label)) + + return pre_eval_results + + def get_classes_and_palette(self, classes=None, palette=None): + """Get class names of current dataset. + + Args: + classes (Sequence[str] | str | None): If classes is None, use + default CLASSES defined by builtin dataset. If classes is a + string, take it as a file name. The file contains the name of + classes where each line contains one class name. If classes is + a tuple or list, override the CLASSES defined by the dataset. + palette (Sequence[Sequence[int]]] | np.ndarray | None): + The palette of segmentation map. If None is given, random + palette will be generated. Default: None + """ + if classes is None: + self.custom_classes = False + return self.CLASSES, self.PALETTE + + self.custom_classes = True + if isinstance(classes, str): + # take it as a file path + class_names = mmcv.list_from_file(classes) + elif isinstance(classes, (tuple, list)): + class_names = classes + else: + raise ValueError(f'Unsupported type {type(classes)} of classes.') + + if self.CLASSES: + if not set(class_names).issubset(self.CLASSES): + raise ValueError('classes is not a subset of CLASSES.') + + # dictionary, its keys are the old label ids and its values + # are the new label ids. + # used for changing pixel labels in load_annotations. + self.label_map = {} + for i, c in enumerate(self.CLASSES): + if c not in class_names: + self.label_map[i] = -1 + else: + self.label_map[i] = class_names.index(c) + + palette = self.get_palette_for_custom_classes(class_names, palette) + + return class_names, palette + + def get_palette_for_custom_classes(self, class_names, palette=None): + + if self.label_map is not None: + # return subset of palette + palette = [] + for old_id, new_id in sorted( + self.label_map.items(), key=lambda x: x[1]): + if new_id != -1: + palette.append(self.PALETTE[old_id]) + palette = type(self.PALETTE)(palette) + + elif palette is None: + if self.PALETTE is None: + # Get random state before set seed, and restore + # random state later. + # It will prevent loss of randomness, as the palette + # may be different in each iteration if not specified. + # See: https://github.com/open-mmlab/mmdetection/issues/5844 + state = np.random.get_state() + np.random.seed(42) + # random palette + palette = np.random.randint(0, 255, size=(len(class_names), 3)) + np.random.set_state(state) + else: + palette = self.PALETTE + + return palette + + def evaluate(self, + results, + metric='mIoU', + logger=None, + gt_seg_maps=None, + **kwargs): + """Evaluate the dataset. + + Args: + results (list[tuple[torch.Tensor]] | list[str]): per image pre_eval + results or predict segmentation map for computing evaluation + metric. + metric (str | list[str]): Metrics to be evaluated. 'mIoU', + 'mDice' and 'mFscore' are supported. + logger (logging.Logger | None | str): Logger used for printing + related information during evaluation. Default: None. + gt_seg_maps (generator[ndarray]): Custom gt seg maps as input, + used in ConcatDataset + + Returns: + dict[str, float]: Default metrics. + """ + if isinstance(metric, str): + metric = [metric] + allowed_metrics = ['mIoU', 'mDice', 'mFscore'] + if not set(metric).issubset(set(allowed_metrics)): + raise KeyError('metric {} is not supported'.format(metric)) + + eval_results = {} + # test a list of files + if mmcv.is_list_of(results, np.ndarray) or mmcv.is_list_of( + results, str): + if gt_seg_maps is None: + gt_seg_maps = self.get_gt_seg_maps() + num_classes = len(self.CLASSES) + ret_metrics = eval_metrics( + results, + gt_seg_maps, + num_classes, + self.ignore_index, + metric, + label_map=dict(), + reduce_zero_label=self.reduce_zero_label) + # test a list of pre_eval_results + else: + ret_metrics = pre_eval_to_metrics(results, metric) + + # Because dataset.CLASSES is required for per-eval. + if self.CLASSES is None: + class_names = tuple(range(num_classes)) + else: + class_names = self.CLASSES + + # summary table + ret_metrics_summary = OrderedDict({ + ret_metric: np.round(np.nanmean(ret_metric_value) * 100, 2) + for ret_metric, ret_metric_value in ret_metrics.items() + }) + + # each class table + ret_metrics.pop('aAcc', None) + ret_metrics_class = OrderedDict({ + ret_metric: np.round(ret_metric_value * 100, 2) + for ret_metric, ret_metric_value in ret_metrics.items() + }) + ret_metrics_class.update({'Class': class_names}) + ret_metrics_class.move_to_end('Class', last=False) + + # for logger + class_table_data = PrettyTable() + for key, val in ret_metrics_class.items(): + class_table_data.add_column(key, val) + + summary_table_data = PrettyTable() + for key, val in ret_metrics_summary.items(): + if key == 'aAcc': + summary_table_data.add_column(key, [val]) + else: + summary_table_data.add_column('m' + key, [val]) + + print_log('per class results:', logger) + print_log('\n' + class_table_data.get_string(), logger=logger) + print_log('Summary:', logger) + print_log('\n' + summary_table_data.get_string(), logger=logger) + + # each metric dict + for key, value in ret_metrics_summary.items(): + if key == 'aAcc': + eval_results[key] = value / 100.0 + else: + eval_results['m' + key] = value / 100.0 + + ret_metrics_class.pop('Class', None) + for key, value in ret_metrics_class.items(): + eval_results.update({ + key + '.' + str(name): value[idx] / 100.0 + for idx, name in enumerate(class_names) + }) + + return eval_results diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/dataset_wrappers.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/dataset_wrappers.py new file mode 100644 index 0000000000000000000000000000000000000000..1fb089f9f287f841d0a99f67ab840f28175c87ec --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/dataset_wrappers.py @@ -0,0 +1,277 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import bisect +import collections +import copy +from itertools import chain + +import mmcv +import numpy as np +from mmcv.utils import build_from_cfg, print_log +from torch.utils.data.dataset import ConcatDataset as _ConcatDataset + +from .builder import DATASETS, PIPELINES +from .cityscapes import CityscapesDataset + + +@DATASETS.register_module() +class ConcatDataset(_ConcatDataset): + """A wrapper of concatenated dataset. + + Same as :obj:`torch.utils.data.dataset.ConcatDataset`, but + support evaluation and formatting results + + Args: + datasets (list[:obj:`Dataset`]): A list of datasets. + separate_eval (bool): Whether to evaluate the concatenated + dataset results separately, Defaults to True. + """ + + def __init__(self, datasets, separate_eval=True): + super(ConcatDataset, self).__init__(datasets) + self.CLASSES = datasets[0].CLASSES + self.PALETTE = datasets[0].PALETTE + self.separate_eval = separate_eval + assert separate_eval in [True, False], \ + f'separate_eval can only be True or False,' \ + f'but get {separate_eval}' + if any([isinstance(ds, CityscapesDataset) for ds in datasets]): + raise NotImplementedError( + 'Evaluating ConcatDataset containing CityscapesDataset' + 'is not supported!') + + def evaluate(self, results, logger=None, **kwargs): + """Evaluate the results. + + Args: + results (list[tuple[torch.Tensor]] | list[str]]): per image + pre_eval results or predict segmentation map for + computing evaluation metric. + logger (logging.Logger | str | None): Logger used for printing + related information during evaluation. Default: None. + + Returns: + dict[str: float]: evaluate results of the total dataset + or each separate + dataset if `self.separate_eval=True`. + """ + assert len(results) == self.cumulative_sizes[-1], \ + ('Dataset and results have different sizes: ' + f'{self.cumulative_sizes[-1]} v.s. {len(results)}') + + # Check whether all the datasets support evaluation + for dataset in self.datasets: + assert hasattr(dataset, 'evaluate'), \ + f'{type(dataset)} does not implement evaluate function' + + if self.separate_eval: + dataset_idx = -1 + total_eval_results = dict() + for size, dataset in zip(self.cumulative_sizes, self.datasets): + start_idx = 0 if dataset_idx == -1 else \ + self.cumulative_sizes[dataset_idx] + end_idx = self.cumulative_sizes[dataset_idx + 1] + + results_per_dataset = results[start_idx:end_idx] + print_log( + f'\nEvaluateing {dataset.img_dir} with ' + f'{len(results_per_dataset)} images now', + logger=logger) + + eval_results_per_dataset = dataset.evaluate( + results_per_dataset, logger=logger, **kwargs) + dataset_idx += 1 + for k, v in eval_results_per_dataset.items(): + total_eval_results.update({f'{dataset_idx}_{k}': v}) + + return total_eval_results + + if len(set([type(ds) for ds in self.datasets])) != 1: + raise NotImplementedError( + 'All the datasets should have same types when ' + 'self.separate_eval=False') + else: + if mmcv.is_list_of(results, np.ndarray) or mmcv.is_list_of( + results, str): + # merge the generators of gt_seg_maps + gt_seg_maps = chain( + *[dataset.get_gt_seg_maps() for dataset in self.datasets]) + else: + # if the results are `pre_eval` results, + # we do not need gt_seg_maps to evaluate + gt_seg_maps = None + eval_results = self.datasets[0].evaluate( + results, gt_seg_maps=gt_seg_maps, logger=logger, **kwargs) + return eval_results + + def get_dataset_idx_and_sample_idx(self, indice): + """Return dataset and sample index when given an indice of + ConcatDataset. + + Args: + indice (int): indice of sample in ConcatDataset + + Returns: + int: the index of sub dataset the sample belong to + int: the index of sample in its corresponding subset + """ + if indice < 0: + if -indice > len(self): + raise ValueError( + 'absolute value of index should not exceed dataset length') + indice = len(self) + indice + dataset_idx = bisect.bisect_right(self.cumulative_sizes, indice) + if dataset_idx == 0: + sample_idx = indice + else: + sample_idx = indice - self.cumulative_sizes[dataset_idx - 1] + return dataset_idx, sample_idx + + def format_results(self, results, imgfile_prefix, indices=None, **kwargs): + """format result for every sample of ConcatDataset.""" + if indices is None: + indices = list(range(len(self))) + + assert isinstance(results, list), 'results must be a list.' + assert isinstance(indices, list), 'indices must be a list.' + + ret_res = [] + for i, indice in enumerate(indices): + dataset_idx, sample_idx = self.get_dataset_idx_and_sample_idx( + indice) + res = self.datasets[dataset_idx].format_results( + [results[i]], + imgfile_prefix + f'/{dataset_idx}', + indices=[sample_idx], + **kwargs) + ret_res.append(res) + return sum(ret_res, []) + + def pre_eval(self, preds, indices): + """do pre eval for every sample of ConcatDataset.""" + # In order to compat with batch inference + if not isinstance(indices, list): + indices = [indices] + if not isinstance(preds, list): + preds = [preds] + ret_res = [] + for i, indice in enumerate(indices): + dataset_idx, sample_idx = self.get_dataset_idx_and_sample_idx( + indice) + res = self.datasets[dataset_idx].pre_eval(preds[i], sample_idx) + ret_res.append(res) + return sum(ret_res, []) + + +@DATASETS.register_module() +class RepeatDataset(object): + """A wrapper of repeated dataset. + + The length of repeated dataset will be `times` larger than the original + dataset. This is useful when the data loading time is long but the dataset + is small. Using RepeatDataset can reduce the data loading time between + epochs. + + Args: + dataset (:obj:`Dataset`): The dataset to be repeated. + times (int): Repeat times. + """ + + def __init__(self, dataset, times): + self.dataset = dataset + self.times = times + self.CLASSES = dataset.CLASSES + self.PALETTE = dataset.PALETTE + self._ori_len = len(self.dataset) + + def __getitem__(self, idx): + """Get item from original dataset.""" + return self.dataset[idx % self._ori_len] + + def __len__(self): + """The length is multiplied by ``times``""" + return self.times * self._ori_len + + +@DATASETS.register_module() +class MultiImageMixDataset: + """A wrapper of multiple images mixed dataset. + + Suitable for training on multiple images mixed data augmentation like + mosaic and mixup. For the augmentation pipeline of mixed image data, + the `get_indexes` method needs to be provided to obtain the image + indexes, and you can set `skip_flags` to change the pipeline running + process. + + + Args: + dataset (:obj:`CustomDataset`): The dataset to be mixed. + pipeline (Sequence[dict]): Sequence of transform object or + config dict to be composed. + skip_type_keys (list[str], optional): Sequence of type string to + be skip pipeline. Default to None. + """ + + def __init__(self, dataset, pipeline, skip_type_keys=None): + assert isinstance(pipeline, collections.abc.Sequence) + if skip_type_keys is not None: + assert all([ + isinstance(skip_type_key, str) + for skip_type_key in skip_type_keys + ]) + self._skip_type_keys = skip_type_keys + + self.pipeline = [] + self.pipeline_types = [] + for transform in pipeline: + if isinstance(transform, dict): + self.pipeline_types.append(transform['type']) + transform = build_from_cfg(transform, PIPELINES) + self.pipeline.append(transform) + else: + raise TypeError('pipeline must be a dict') + + self.dataset = dataset + self.CLASSES = dataset.CLASSES + self.PALETTE = dataset.PALETTE + self.num_samples = len(dataset) + + def __len__(self): + return self.num_samples + + def __getitem__(self, idx): + results = copy.deepcopy(self.dataset[idx]) + for (transform, transform_type) in zip(self.pipeline, + self.pipeline_types): + if self._skip_type_keys is not None and \ + transform_type in self._skip_type_keys: + continue + + if hasattr(transform, 'get_indexes'): + indexes = transform.get_indexes(self.dataset) + if not isinstance(indexes, collections.abc.Sequence): + indexes = [indexes] + mix_results = [ + copy.deepcopy(self.dataset[index]) for index in indexes + ] + results['mix_results'] = mix_results + + results = transform(results) + + if 'mix_results' in results: + results.pop('mix_results') + + return results + + def update_skip_type_keys(self, skip_type_keys): + """Update skip_type_keys. + + It is called by an external hook. + + Args: + skip_type_keys (list[str], optional): Sequence of type + string to be skip pipeline. + """ + assert all([ + isinstance(skip_type_key, str) for skip_type_key in skip_type_keys + ]) + self._skip_type_keys = skip_type_keys diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pascal_context.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pascal_context.py new file mode 100644 index 0000000000000000000000000000000000000000..efacee0f3fb856f10c6d589346316342a1f20c5e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pascal_context.py @@ -0,0 +1,103 @@ +# Copyright (c) OpenMMLab. All rights reserved. + +from .builder import DATASETS +from .custom import CustomDataset + + +@DATASETS.register_module() +class PascalContextDataset(CustomDataset): + """PascalContext dataset. + + In segmentation map annotation for PascalContext, 0 stands for background, + which is included in 60 categories. ``reduce_zero_label`` is fixed to + False. The ``img_suffix`` is fixed to '.jpg' and ``seg_map_suffix`` is + fixed to '.png'. + + Args: + split (str): Split txt file for PascalContext. + """ + + CLASSES = ('background', 'aeroplane', 'bag', 'bed', 'bedclothes', 'bench', + 'bicycle', 'bird', 'boat', 'book', 'bottle', 'building', 'bus', + 'cabinet', 'car', 'cat', 'ceiling', 'chair', 'cloth', + 'computer', 'cow', 'cup', 'curtain', 'dog', 'door', 'fence', + 'floor', 'flower', 'food', 'grass', 'ground', 'horse', + 'keyboard', 'light', 'motorbike', 'mountain', 'mouse', 'person', + 'plate', 'platform', 'pottedplant', 'road', 'rock', 'sheep', + 'shelves', 'sidewalk', 'sign', 'sky', 'snow', 'sofa', 'table', + 'track', 'train', 'tree', 'truck', 'tvmonitor', 'wall', 'water', + 'window', 'wood') + + PALETTE = [[120, 120, 120], [180, 120, 120], [6, 230, 230], [80, 50, 50], + [4, 200, 3], [120, 120, 80], [140, 140, 140], [204, 5, 255], + [230, 230, 230], [4, 250, 7], [224, 5, 255], [235, 255, 7], + [150, 5, 61], [120, 120, 70], [8, 255, 51], [255, 6, 82], + [143, 255, 140], [204, 255, 4], [255, 51, 7], [204, 70, 3], + [0, 102, 200], [61, 230, 250], [255, 6, 51], [11, 102, 255], + [255, 7, 71], [255, 9, 224], [9, 7, 230], [220, 220, 220], + [255, 9, 92], [112, 9, 255], [8, 255, 214], [7, 255, 224], + [255, 184, 6], [10, 255, 71], [255, 41, 10], [7, 255, 255], + [224, 255, 8], [102, 8, 255], [255, 61, 6], [255, 194, 7], + [255, 122, 8], [0, 255, 20], [255, 8, 41], [255, 5, 153], + [6, 51, 255], [235, 12, 255], [160, 150, 20], [0, 163, 255], + [140, 140, 140], [250, 10, 15], [20, 255, 0], [31, 255, 0], + [255, 31, 0], [255, 224, 0], [153, 255, 0], [0, 0, 255], + [255, 71, 0], [0, 235, 255], [0, 173, 255], [31, 0, 255]] + + def __init__(self, split, **kwargs): + super(PascalContextDataset, self).__init__( + img_suffix='.jpg', + seg_map_suffix='.png', + split=split, + reduce_zero_label=False, + **kwargs) + assert self.file_client.exists(self.img_dir) and self.split is not None + + +@DATASETS.register_module() +class PascalContextDataset59(CustomDataset): + """PascalContext dataset. + + In segmentation map annotation for PascalContext, 0 stands for background, + which is included in 60 categories. ``reduce_zero_label`` is fixed to + False. The ``img_suffix`` is fixed to '.jpg' and ``seg_map_suffix`` is + fixed to '.png'. + + Args: + split (str): Split txt file for PascalContext. + """ + + CLASSES = ('aeroplane', 'bag', 'bed', 'bedclothes', 'bench', 'bicycle', + 'bird', 'boat', 'book', 'bottle', 'building', 'bus', 'cabinet', + 'car', 'cat', 'ceiling', 'chair', 'cloth', 'computer', 'cow', + 'cup', 'curtain', 'dog', 'door', 'fence', 'floor', 'flower', + 'food', 'grass', 'ground', 'horse', 'keyboard', 'light', + 'motorbike', 'mountain', 'mouse', 'person', 'plate', 'platform', + 'pottedplant', 'road', 'rock', 'sheep', 'shelves', 'sidewalk', + 'sign', 'sky', 'snow', 'sofa', 'table', 'track', 'train', + 'tree', 'truck', 'tvmonitor', 'wall', 'water', 'window', 'wood') + + PALETTE = [[180, 120, 120], [6, 230, 230], [80, 50, 50], [4, 200, 3], + [120, 120, 80], [140, 140, 140], [204, 5, 255], [230, 230, 230], + [4, 250, 7], [224, 5, 255], [235, 255, 7], [150, 5, 61], + [120, 120, 70], [8, 255, 51], [255, 6, 82], [143, 255, 140], + [204, 255, 4], [255, 51, 7], [204, 70, 3], [0, 102, 200], + [61, 230, 250], [255, 6, 51], [11, 102, 255], [255, 7, 71], + [255, 9, 224], [9, 7, 230], [220, 220, 220], [255, 9, 92], + [112, 9, 255], [8, 255, 214], [7, 255, 224], [255, 184, 6], + [10, 255, 71], [255, 41, 10], [7, 255, 255], [224, 255, 8], + [102, 8, 255], [255, 61, 6], [255, 194, 7], [255, 122, 8], + [0, 255, 20], [255, 8, 41], [255, 5, 153], [6, 51, 255], + [235, 12, 255], [160, 150, 20], [0, 163, 255], [140, 140, 140], + [250, 10, 15], [20, 255, 0], [31, 255, 0], [255, 31, 0], + [255, 224, 0], [153, 255, 0], [0, 0, 255], [255, 71, 0], + [0, 235, 255], [0, 173, 255], [31, 0, 255]] + + def __init__(self, split, **kwargs): + super(PascalContextDataset59, self).__init__( + img_suffix='.jpg', + seg_map_suffix='.png', + split=split, + reduce_zero_label=True, + **kwargs) + assert self.file_client.exists(self.img_dir) and self.split is not None diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..8256a6fe2f03a381ee62a0271411d5102caf8c43 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/__init__.py @@ -0,0 +1,19 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .compose import Compose +from .formatting import (Collect, ImageToTensor, ToDataContainer, ToTensor, + Transpose, to_tensor) +from .loading import LoadAnnotations, LoadImageFromFile +from .test_time_aug import MultiScaleFlipAug +from .transforms import (CLAHE, AdjustGamma, Normalize, Pad, + PhotoMetricDistortion, RandomCrop, RandomCutOut, + RandomFlip, RandomMosaic, RandomRotate, Rerange, + Resize, RGB2Gray, SegRescale) + +__all__ = [ + 'Compose', 'to_tensor', 'ToTensor', 'ImageToTensor', 'ToDataContainer', + 'Transpose', 'Collect', 'LoadAnnotations', 'LoadImageFromFile', + 'MultiScaleFlipAug', 'Resize', 'RandomFlip', 'Pad', 'RandomCrop', + 'Normalize', 'SegRescale', 'PhotoMetricDistortion', 'RandomRotate', + 'AdjustGamma', 'CLAHE', 'Rerange', 'RGB2Gray', 'RandomCutOut', + 'RandomMosaic' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/compose.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/compose.py new file mode 100644 index 0000000000000000000000000000000000000000..30280c1332abc253434ae4e88271d73de2690ecb --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/compose.py @@ -0,0 +1,52 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import collections + +from mmcv.utils import build_from_cfg + +from ..builder import PIPELINES + + +@PIPELINES.register_module() +class Compose(object): + """Compose multiple transforms sequentially. + + Args: + transforms (Sequence[dict | callable]): Sequence of transform object or + config dict to be composed. + """ + + def __init__(self, transforms): + assert isinstance(transforms, collections.abc.Sequence) + self.transforms = [] + for transform in transforms: + if isinstance(transform, dict): + transform = build_from_cfg(transform, PIPELINES) + self.transforms.append(transform) + elif callable(transform): + self.transforms.append(transform) + else: + raise TypeError('transform must be callable or a dict') + + def __call__(self, data): + """Call function to apply transforms sequentially. + + Args: + data (dict): A result dict contains the data to transform. + + Returns: + dict: Transformed data. + """ + + for t in self.transforms: + data = t(data) + if data is None: + return None + return data + + def __repr__(self): + format_string = self.__class__.__name__ + '(' + for t in self.transforms: + format_string += '\n' + format_string += f' {t}' + format_string += '\n)' + return format_string diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/formating.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/formating.py new file mode 100644 index 0000000000000000000000000000000000000000..f6e53bfebe3e76412600361da01c36cb440bafd8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/formating.py @@ -0,0 +1,9 @@ +# Copyright (c) OpenMMLab. All rights reserved. +# flake8: noqa +import warnings + +from .formatting import * + +warnings.warn('DeprecationWarning: mmseg.datasets.pipelines.formating will be ' + 'deprecated in 2021, please replace it with ' + 'mmseg.datasets.pipelines.formatting.') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/formatting.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/formatting.py new file mode 100644 index 0000000000000000000000000000000000000000..4e057c1b8161148166d17cf339b4c04e3f31bc5f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/formatting.py @@ -0,0 +1,289 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from collections.abc import Sequence + +import mmcv +import numpy as np +import torch +from mmcv.parallel import DataContainer as DC + +from ..builder import PIPELINES + + +def to_tensor(data): + """Convert objects of various python types to :obj:`torch.Tensor`. + + Supported types are: :class:`numpy.ndarray`, :class:`torch.Tensor`, + :class:`Sequence`, :class:`int` and :class:`float`. + + Args: + data (torch.Tensor | numpy.ndarray | Sequence | int | float): Data to + be converted. + """ + + if isinstance(data, torch.Tensor): + return data + elif isinstance(data, np.ndarray): + return torch.from_numpy(data) + elif isinstance(data, Sequence) and not mmcv.is_str(data): + return torch.tensor(data) + elif isinstance(data, int): + return torch.LongTensor([data]) + elif isinstance(data, float): + return torch.FloatTensor([data]) + else: + raise TypeError(f'type {type(data)} cannot be converted to tensor.') + + +@PIPELINES.register_module() +class ToTensor(object): + """Convert some results to :obj:`torch.Tensor` by given keys. + + Args: + keys (Sequence[str]): Keys that need to be converted to Tensor. + """ + + def __init__(self, keys): + self.keys = keys + + def __call__(self, results): + """Call function to convert data in results to :obj:`torch.Tensor`. + + Args: + results (dict): Result dict contains the data to convert. + + Returns: + dict: The result dict contains the data converted + to :obj:`torch.Tensor`. + """ + + for key in self.keys: + results[key] = to_tensor(results[key]) + return results + + def __repr__(self): + return self.__class__.__name__ + f'(keys={self.keys})' + + +@PIPELINES.register_module() +class ImageToTensor(object): + """Convert image to :obj:`torch.Tensor` by given keys. + + The dimension order of input image is (H, W, C). The pipeline will convert + it to (C, H, W). If only 2 dimension (H, W) is given, the output would be + (1, H, W). + + Args: + keys (Sequence[str]): Key of images to be converted to Tensor. + """ + + def __init__(self, keys): + self.keys = keys + + def __call__(self, results): + """Call function to convert image in results to :obj:`torch.Tensor` and + transpose the channel order. + + Args: + results (dict): Result dict contains the image data to convert. + + Returns: + dict: The result dict contains the image converted + to :obj:`torch.Tensor` and transposed to (C, H, W) order. + """ + + for key in self.keys: + img = results[key] + if len(img.shape) < 3: + img = np.expand_dims(img, -1) + results[key] = to_tensor(img.transpose(2, 0, 1)) + return results + + def __repr__(self): + return self.__class__.__name__ + f'(keys={self.keys})' + + +@PIPELINES.register_module() +class Transpose(object): + """Transpose some results by given keys. + + Args: + keys (Sequence[str]): Keys of results to be transposed. + order (Sequence[int]): Order of transpose. + """ + + def __init__(self, keys, order): + self.keys = keys + self.order = order + + def __call__(self, results): + """Call function to convert image in results to :obj:`torch.Tensor` and + transpose the channel order. + + Args: + results (dict): Result dict contains the image data to convert. + + Returns: + dict: The result dict contains the image converted + to :obj:`torch.Tensor` and transposed to (C, H, W) order. + """ + + for key in self.keys: + results[key] = results[key].transpose(self.order) + return results + + def __repr__(self): + return self.__class__.__name__ + \ + f'(keys={self.keys}, order={self.order})' + + +@PIPELINES.register_module() +class ToDataContainer(object): + """Convert results to :obj:`mmcv.DataContainer` by given fields. + + Args: + fields (Sequence[dict]): Each field is a dict like + ``dict(key='xxx', **kwargs)``. The ``key`` in result will + be converted to :obj:`mmcv.DataContainer` with ``**kwargs``. + Default: ``(dict(key='img', stack=True), + dict(key='gt_semantic_seg'))``. + """ + + def __init__(self, + fields=(dict(key='img', + stack=True), dict(key='gt_semantic_seg'))): + self.fields = fields + + def __call__(self, results): + """Call function to convert data in results to + :obj:`mmcv.DataContainer`. + + Args: + results (dict): Result dict contains the data to convert. + + Returns: + dict: The result dict contains the data converted to + :obj:`mmcv.DataContainer`. + """ + + for field in self.fields: + field = field.copy() + key = field.pop('key') + results[key] = DC(results[key], **field) + return results + + def __repr__(self): + return self.__class__.__name__ + f'(fields={self.fields})' + + +@PIPELINES.register_module() +class DefaultFormatBundle(object): + """Default formatting bundle. + + It simplifies the pipeline of formatting common fields, including "img" + and "gt_semantic_seg". These fields are formatted as follows. + + - img: (1)transpose, (2)to tensor, (3)to DataContainer (stack=True) + - gt_semantic_seg: (1)unsqueeze dim-0 (2)to tensor, + (3)to DataContainer (stack=True) + """ + + def __call__(self, results): + """Call function to transform and format common fields in results. + + Args: + results (dict): Result dict contains the data to convert. + + Returns: + dict: The result dict contains the data that is formatted with + default bundle. + """ + + if 'img' in results: + img = results['img'] + if len(img.shape) < 3: + img = np.expand_dims(img, -1) + img = np.ascontiguousarray(img.transpose(2, 0, 1)) + results['img'] = DC(to_tensor(img), stack=True) + if 'gt_semantic_seg' in results: + # convert to long + results['gt_semantic_seg'] = DC( + to_tensor(results['gt_semantic_seg'][None, + ...].astype(np.int64)), + stack=True) + return results + + def __repr__(self): + return self.__class__.__name__ + + +@PIPELINES.register_module() +class Collect(object): + """Collect data from the loader relevant to the specific task. + + This is usually the last stage of the data loader pipeline. Typically keys + is set to some subset of "img", "gt_semantic_seg". + + The "img_meta" item is always populated. The contents of the "img_meta" + dictionary depends on "meta_keys". By default this includes: + + - "img_shape": shape of the image input to the network as a tuple + (h, w, c). Note that images may be zero padded on the bottom/right + if the batch tensor is larger than this shape. + + - "scale_factor": a float indicating the preprocessing scale + + - "flip": a boolean indicating if image flip transform was used + + - "filename": path to the image file + + - "ori_shape": original shape of the image as a tuple (h, w, c) + + - "pad_shape": image shape after padding + + - "img_norm_cfg": a dict of normalization information: + - mean - per channel mean subtraction + - std - per channel std divisor + - to_rgb - bool indicating if bgr was converted to rgb + + Args: + keys (Sequence[str]): Keys of results to be collected in ``data``. + meta_keys (Sequence[str], optional): Meta keys to be converted to + ``mmcv.DataContainer`` and collected in ``data[img_metas]``. + Default: (``filename``, ``ori_filename``, ``ori_shape``, + ``img_shape``, ``pad_shape``, ``scale_factor``, ``flip``, + ``flip_direction``, ``img_norm_cfg``) + """ + + def __init__(self, + keys, + meta_keys=('filename', 'ori_filename', 'ori_shape', + 'img_shape', 'pad_shape', 'scale_factor', 'flip', + 'flip_direction', 'img_norm_cfg')): + self.keys = keys + self.meta_keys = meta_keys + + def __call__(self, results): + """Call function to collect keys in results. The keys in ``meta_keys`` + will be converted to :obj:mmcv.DataContainer. + + Args: + results (dict): Result dict contains the data to collect. + + Returns: + dict: The result dict contains the following keys + - keys in``self.keys`` + - ``img_metas`` + """ + + data = {} + img_meta = {} + for key in self.meta_keys: + img_meta[key] = results[key] + data['img_metas'] = DC(img_meta, cpu_only=True) + for key in self.keys: + data[key] = results[key] + return data + + def __repr__(self): + return self.__class__.__name__ + \ + f'(keys={self.keys}, meta_keys={self.meta_keys})' diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/loading.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/loading.py new file mode 100644 index 0000000000000000000000000000000000000000..572e43431832454b31fb8ed0305c995b404328c4 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/loading.py @@ -0,0 +1,158 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp + +import mmcv +import numpy as np + +from ..builder import PIPELINES + + +@PIPELINES.register_module() +class LoadImageFromFile(object): + """Load an image from file. + + Required keys are "img_prefix" and "img_info" (a dict that must contain the + key "filename"). Added or updated keys are "filename", "img", "img_shape", + "ori_shape" (same as `img_shape`), "pad_shape" (same as `img_shape`), + "scale_factor" (1.0) and "img_norm_cfg" (means=0 and stds=1). + + Args: + to_float32 (bool): Whether to convert the loaded image to a float32 + numpy array. If set to False, the loaded image is an uint8 array. + Defaults to False. + color_type (str): The flag argument for :func:`mmcv.imfrombytes`. + Defaults to 'color'. + file_client_args (dict): Arguments to instantiate a FileClient. + See :class:`mmcv.fileio.FileClient` for details. + Defaults to ``dict(backend='disk')``. + imdecode_backend (str): Backend for :func:`mmcv.imdecode`. Default: + 'cv2' + """ + + def __init__(self, + to_float32=False, + color_type='color', + file_client_args=dict(backend='disk'), + imdecode_backend='cv2'): + self.to_float32 = to_float32 + self.color_type = color_type + self.file_client_args = file_client_args.copy() + self.file_client = None + self.imdecode_backend = imdecode_backend + + def __call__(self, results): + """Call functions to load image and get image meta information. + + Args: + results (dict): Result dict from :obj:`mmseg.CustomDataset`. + + Returns: + dict: The dict contains loaded image and meta information. + """ + + if self.file_client is None: + self.file_client = mmcv.FileClient(**self.file_client_args) + + if results.get('img_prefix') is not None: + filename = osp.join(results['img_prefix'], + results['img_info']['filename']) + else: + filename = results['img_info']['filename'] + img_bytes = self.file_client.get(filename) + img = mmcv.imfrombytes( + img_bytes, flag=self.color_type, backend=self.imdecode_backend) + if self.to_float32: + img = img.astype(np.float32) + + results['filename'] = filename + results['ori_filename'] = results['img_info']['filename'] + results['img'] = img + results['img_shape'] = img.shape + results['ori_shape'] = img.shape + # Set initial values for default meta_keys + results['pad_shape'] = img.shape + results['scale_factor'] = 1.0 + num_channels = 1 if len(img.shape) < 3 else img.shape[2] + results['img_norm_cfg'] = dict( + mean=np.zeros(num_channels, dtype=np.float32), + std=np.ones(num_channels, dtype=np.float32), + to_rgb=False) + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(to_float32={self.to_float32},' + repr_str += f"color_type='{self.color_type}'," + repr_str += f"imdecode_backend='{self.imdecode_backend}')" + return repr_str + + +@PIPELINES.register_module() +class LoadAnnotations(object): + """Load annotations for semantic segmentation. + + Args: + reduce_zero_label (bool): Whether reduce all label value by 1. + Usually used for datasets where 0 is background label. + Default: False. + file_client_args (dict): Arguments to instantiate a FileClient. + See :class:`mmcv.fileio.FileClient` for details. + Defaults to ``dict(backend='disk')``. + imdecode_backend (str): Backend for :func:`mmcv.imdecode`. Default: + 'pillow' + """ + + def __init__(self, + reduce_zero_label=False, + file_client_args=dict(backend='disk'), + imdecode_backend='pillow'): + self.reduce_zero_label = reduce_zero_label + self.file_client_args = file_client_args.copy() + self.file_client = None + self.imdecode_backend = imdecode_backend + + def __call__(self, results): + """Call function to load multiple types annotations. + + Args: + results (dict): Result dict from :obj:`mmseg.CustomDataset`. + + Returns: + dict: The dict contains loaded semantic segmentation annotations. + """ + + if self.file_client is None: + self.file_client = mmcv.FileClient(**self.file_client_args) + + if results.get('seg_prefix', None) is not None: + filename = osp.join(results['seg_prefix'], + results['ann_info']['seg_map']) + else: + filename = results['ann_info']['seg_map'] + img_bytes = self.file_client.get(filename) + gt_semantic_seg = mmcv.imfrombytes( + img_bytes, flag='unchanged', + backend=self.imdecode_backend).squeeze().astype(np.uint8) + # modify if custom classes + if results.get('label_map', None) is not None: + # Add deep copy to solve bug of repeatedly + # replace `gt_semantic_seg`, which is reported in + # https://github.com/open-mmlab/mmsegmentation/pull/1445/ + gt_semantic_seg_copy = gt_semantic_seg.copy() + for old_id, new_id in results['label_map'].items(): + gt_semantic_seg[gt_semantic_seg_copy == old_id] = new_id + # reduce zero_label + if self.reduce_zero_label: + # avoid using underflow conversion + gt_semantic_seg[gt_semantic_seg == 0] = 255 + gt_semantic_seg = gt_semantic_seg - 1 + gt_semantic_seg[gt_semantic_seg == 254] = 255 + results['gt_semantic_seg'] = gt_semantic_seg + results['seg_fields'].append('gt_semantic_seg') + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(reduce_zero_label={self.reduce_zero_label},' + repr_str += f"imdecode_backend='{self.imdecode_backend}')" + return repr_str diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/test_time_aug.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/test_time_aug.py new file mode 100644 index 0000000000000000000000000000000000000000..5c17cbbba11fe99976f4d929215343ca403da688 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/test_time_aug.py @@ -0,0 +1,134 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import mmcv + +from ..builder import PIPELINES +from .compose import Compose + + +@PIPELINES.register_module() +class MultiScaleFlipAug(object): + """Test-time augmentation with multiple scales and flipping. + + An example configuration is as followed: + + .. code-block:: + + img_scale=(2048, 1024), + img_ratios=[0.5, 1.0], + flip=True, + transforms=[ + dict(type='Resize', keep_ratio=True), + dict(type='RandomFlip'), + dict(type='Normalize', **img_norm_cfg), + dict(type='Pad', size_divisor=32), + dict(type='ImageToTensor', keys=['img']), + dict(type='Collect', keys=['img']), + ] + + After MultiScaleFLipAug with above configuration, the results are wrapped + into lists of the same length as followed: + + .. code-block:: + + dict( + img=[...], + img_shape=[...], + scale=[(1024, 512), (1024, 512), (2048, 1024), (2048, 1024)] + flip=[False, True, False, True] + ... + ) + + Args: + transforms (list[dict]): Transforms to apply in each augmentation. + img_scale (None | tuple | list[tuple]): Images scales for resizing. + img_ratios (float | list[float]): Image ratios for resizing + flip (bool): Whether apply flip augmentation. Default: False. + flip_direction (str | list[str]): Flip augmentation directions, + options are "horizontal" and "vertical". If flip_direction is list, + multiple flip augmentations will be applied. + It has no effect when flip == False. Default: "horizontal". + """ + + def __init__(self, + transforms, + img_scale, + img_ratios=None, + flip=False, + flip_direction='horizontal'): + self.transforms = Compose(transforms) + if img_ratios is not None: + img_ratios = img_ratios if isinstance(img_ratios, + list) else [img_ratios] + assert mmcv.is_list_of(img_ratios, float) + if img_scale is None: + # mode 1: given img_scale=None and a range of image ratio + self.img_scale = None + assert mmcv.is_list_of(img_ratios, float) + elif isinstance(img_scale, tuple) and mmcv.is_list_of( + img_ratios, float): + assert len(img_scale) == 2 + # mode 2: given a scale and a range of image ratio + self.img_scale = [(int(img_scale[0] * ratio), + int(img_scale[1] * ratio)) + for ratio in img_ratios] + else: + # mode 3: given multiple scales + self.img_scale = img_scale if isinstance(img_scale, + list) else [img_scale] + assert mmcv.is_list_of(self.img_scale, tuple) or self.img_scale is None + self.flip = flip + self.img_ratios = img_ratios + self.flip_direction = flip_direction if isinstance( + flip_direction, list) else [flip_direction] + assert mmcv.is_list_of(self.flip_direction, str) + if not self.flip and self.flip_direction != ['horizontal']: + warnings.warn( + 'flip_direction has no effect when flip is set to False') + if (self.flip + and not any([t['type'] == 'RandomFlip' for t in transforms])): + warnings.warn( + 'flip has no effect when RandomFlip is not in transforms') + + def __call__(self, results): + """Call function to apply test time augment transforms on results. + + Args: + results (dict): Result dict contains the data to transform. + + Returns: + dict[str: list]: The augmented data, where each value is wrapped + into a list. + """ + + aug_data = [] + if self.img_scale is None and mmcv.is_list_of(self.img_ratios, float): + h, w = results['img'].shape[:2] + img_scale = [(int(w * ratio), int(h * ratio)) + for ratio in self.img_ratios] + else: + img_scale = self.img_scale + flip_aug = [False, True] if self.flip else [False] + for scale in img_scale: + for flip in flip_aug: + for direction in self.flip_direction: + _results = results.copy() + _results['scale'] = scale + _results['flip'] = flip + _results['flip_direction'] = direction + data = self.transforms(_results) + aug_data.append(data) + # list of dict to dict of list + aug_data_dict = {key: [] for key in aug_data[0]} + for data in aug_data: + for key, val in data.items(): + aug_data_dict[key].append(val) + return aug_data_dict + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(transforms={self.transforms}, ' + repr_str += f'img_scale={self.img_scale}, flip={self.flip})' + repr_str += f'flip_direction={self.flip_direction}' + return repr_str diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/transforms.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/transforms.py new file mode 100644 index 0000000000000000000000000000000000000000..5673b646fa654bcba39ea897d37a4e7371b5c77f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/pipelines/transforms.py @@ -0,0 +1,1335 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import copy + +import mmcv +import numpy as np +from mmcv.utils import deprecated_api_warning, is_tuple_of +from numpy import random + +from ..builder import PIPELINES + + +@PIPELINES.register_module() +class ResizeToMultiple(object): + """Resize images & seg to multiple of divisor. + + Args: + size_divisor (int): images and gt seg maps need to resize to multiple + of size_divisor. Default: 32. + interpolation (str, optional): The interpolation mode of image resize. + Default: None + """ + + def __init__(self, size_divisor=32, interpolation=None): + self.size_divisor = size_divisor + self.interpolation = interpolation + + def __call__(self, results): + """Call function to resize images, semantic segmentation map to + multiple of size divisor. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Resized results, 'img_shape', 'pad_shape' keys are updated. + """ + # Align image to multiple of size divisor. + img = results['img'] + img = mmcv.imresize_to_multiple( + img, + self.size_divisor, + scale_factor=1, + interpolation=self.interpolation + if self.interpolation else 'bilinear') + + results['img'] = img + results['img_shape'] = img.shape + results['pad_shape'] = img.shape + + # Align segmentation map to multiple of size divisor. + for key in results.get('seg_fields', []): + gt_seg = results[key] + gt_seg = mmcv.imresize_to_multiple( + gt_seg, + self.size_divisor, + scale_factor=1, + interpolation='nearest') + results[key] = gt_seg + + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += (f'(size_divisor={self.size_divisor}, ' + f'interpolation={self.interpolation})') + return repr_str + + +@PIPELINES.register_module() +class Resize(object): + """Resize images & seg. + + This transform resizes the input image to some scale. If the input dict + contains the key "scale", then the scale in the input dict is used, + otherwise the specified scale in the init method is used. + + ``img_scale`` can be None, a tuple (single-scale) or a list of tuple + (multi-scale). There are 4 multiscale modes: + + - ``ratio_range is not None``: + 1. When img_scale is None, img_scale is the shape of image in results + (img_scale = results['img'].shape[:2]) and the image is resized based + on the original size. (mode 1) + 2. When img_scale is a tuple (single-scale), randomly sample a ratio from + the ratio range and multiply it with the image scale. (mode 2) + + - ``ratio_range is None and multiscale_mode == "range"``: randomly sample a + scale from the a range. (mode 3) + + - ``ratio_range is None and multiscale_mode == "value"``: randomly sample a + scale from multiple scales. (mode 4) + + Args: + img_scale (tuple or list[tuple]): Images scales for resizing. + Default:None. + multiscale_mode (str): Either "range" or "value". + Default: 'range' + ratio_range (tuple[float]): (min_ratio, max_ratio). + Default: None + keep_ratio (bool): Whether to keep the aspect ratio when resizing the + image. Default: True + min_size (int, optional): The minimum size for input and the shape + of the image and seg map will not be less than ``min_size``. + As the shape of model input is fixed like 'SETR' and 'BEiT'. + Following the setting in these models, resized images must be + bigger than the crop size in ``slide_inference``. Default: None + """ + + def __init__(self, + img_scale=None, + multiscale_mode='range', + ratio_range=None, + keep_ratio=True, + min_size=None): + if img_scale is None: + self.img_scale = None + else: + if isinstance(img_scale, list): + self.img_scale = img_scale + else: + self.img_scale = [img_scale] + assert mmcv.is_list_of(self.img_scale, tuple) + + if ratio_range is not None: + # mode 1: given img_scale=None and a range of image ratio + # mode 2: given a scale and a range of image ratio + assert self.img_scale is None or len(self.img_scale) == 1 + else: + # mode 3 and 4: given multiple scales or a range of scales + assert multiscale_mode in ['value', 'range'] + + self.multiscale_mode = multiscale_mode + self.ratio_range = ratio_range + self.keep_ratio = keep_ratio + self.min_size = min_size + + @staticmethod + def random_select(img_scales): + """Randomly select an img_scale from given candidates. + + Args: + img_scales (list[tuple]): Images scales for selection. + + Returns: + (tuple, int): Returns a tuple ``(img_scale, scale_dix)``, + where ``img_scale`` is the selected image scale and + ``scale_idx`` is the selected index in the given candidates. + """ + + assert mmcv.is_list_of(img_scales, tuple) + scale_idx = np.random.randint(len(img_scales)) + img_scale = img_scales[scale_idx] + return img_scale, scale_idx + + @staticmethod + def random_sample(img_scales): + """Randomly sample an img_scale when ``multiscale_mode=='range'``. + + Args: + img_scales (list[tuple]): Images scale range for sampling. + There must be two tuples in img_scales, which specify the lower + and upper bound of image scales. + + Returns: + (tuple, None): Returns a tuple ``(img_scale, None)``, where + ``img_scale`` is sampled scale and None is just a placeholder + to be consistent with :func:`random_select`. + """ + + assert mmcv.is_list_of(img_scales, tuple) and len(img_scales) == 2 + img_scale_long = [max(s) for s in img_scales] + img_scale_short = [min(s) for s in img_scales] + long_edge = np.random.randint( + min(img_scale_long), + max(img_scale_long) + 1) + short_edge = np.random.randint( + min(img_scale_short), + max(img_scale_short) + 1) + img_scale = (long_edge, short_edge) + return img_scale, None + + @staticmethod + def random_sample_ratio(img_scale, ratio_range): + """Randomly sample an img_scale when ``ratio_range`` is specified. + + A ratio will be randomly sampled from the range specified by + ``ratio_range``. Then it would be multiplied with ``img_scale`` to + generate sampled scale. + + Args: + img_scale (tuple): Images scale base to multiply with ratio. + ratio_range (tuple[float]): The minimum and maximum ratio to scale + the ``img_scale``. + + Returns: + (tuple, None): Returns a tuple ``(scale, None)``, where + ``scale`` is sampled ratio multiplied with ``img_scale`` and + None is just a placeholder to be consistent with + :func:`random_select`. + """ + + assert isinstance(img_scale, tuple) and len(img_scale) == 2 + min_ratio, max_ratio = ratio_range + assert min_ratio <= max_ratio + ratio = np.random.random_sample() * (max_ratio - min_ratio) + min_ratio + scale = int(img_scale[0] * ratio), int(img_scale[1] * ratio) + return scale, None + + def _random_scale(self, results): + """Randomly sample an img_scale according to ``ratio_range`` and + ``multiscale_mode``. + + If ``ratio_range`` is specified, a ratio will be sampled and be + multiplied with ``img_scale``. + If multiple scales are specified by ``img_scale``, a scale will be + sampled according to ``multiscale_mode``. + Otherwise, single scale will be used. + + Args: + results (dict): Result dict from :obj:`dataset`. + + Returns: + dict: Two new keys 'scale` and 'scale_idx` are added into + ``results``, which would be used by subsequent pipelines. + """ + + if self.ratio_range is not None: + if self.img_scale is None: + h, w = results['img'].shape[:2] + scale, scale_idx = self.random_sample_ratio((w, h), + self.ratio_range) + else: + scale, scale_idx = self.random_sample_ratio( + self.img_scale[0], self.ratio_range) + elif len(self.img_scale) == 1: + scale, scale_idx = self.img_scale[0], 0 + elif self.multiscale_mode == 'range': + scale, scale_idx = self.random_sample(self.img_scale) + elif self.multiscale_mode == 'value': + scale, scale_idx = self.random_select(self.img_scale) + else: + raise NotImplementedError + + results['scale'] = scale + results['scale_idx'] = scale_idx + + def _resize_img(self, results): + """Resize images with ``results['scale']``.""" + if self.keep_ratio: + if self.min_size is not None: + # TODO: Now 'min_size' is an 'int' which means the minimum + # shape of images is (min_size, min_size, 3). 'min_size' + # with tuple type will be supported, i.e. the width and + # height are not equal. + if min(results['scale']) < self.min_size: + new_short = self.min_size + else: + new_short = min(results['scale']) + + h, w = results['img'].shape[:2] + if h > w: + new_h, new_w = new_short * h / w, new_short + else: + new_h, new_w = new_short, new_short * w / h + results['scale'] = (new_h, new_w) + + img, scale_factor = mmcv.imrescale( + results['img'], results['scale'], return_scale=True) + # the w_scale and h_scale has minor difference + # a real fix should be done in the mmcv.imrescale in the future + new_h, new_w = img.shape[:2] + h, w = results['img'].shape[:2] + w_scale = new_w / w + h_scale = new_h / h + else: + img, w_scale, h_scale = mmcv.imresize( + results['img'], results['scale'], return_scale=True) + scale_factor = np.array([w_scale, h_scale, w_scale, h_scale], + dtype=np.float32) + results['img'] = img + results['img_shape'] = img.shape + results['pad_shape'] = img.shape # in case that there is no padding + results['scale_factor'] = scale_factor + results['keep_ratio'] = self.keep_ratio + + def _resize_seg(self, results): + """Resize semantic segmentation map with ``results['scale']``.""" + for key in results.get('seg_fields', []): + if self.keep_ratio: + gt_seg = mmcv.imrescale( + results[key], results['scale'], interpolation='nearest') + else: + gt_seg = mmcv.imresize( + results[key], results['scale'], interpolation='nearest') + results[key] = gt_seg + + def __call__(self, results): + """Call function to resize images, bounding boxes, masks, semantic + segmentation map. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Resized results, 'img_shape', 'pad_shape', 'scale_factor', + 'keep_ratio' keys are added into result dict. + """ + + if 'scale' not in results: + self._random_scale(results) + self._resize_img(results) + self._resize_seg(results) + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += (f'(img_scale={self.img_scale}, ' + f'multiscale_mode={self.multiscale_mode}, ' + f'ratio_range={self.ratio_range}, ' + f'keep_ratio={self.keep_ratio})') + return repr_str + + +@PIPELINES.register_module() +class RandomFlip(object): + """Flip the image & seg. + + If the input dict contains the key "flip", then the flag will be used, + otherwise it will be randomly decided by a ratio specified in the init + method. + + Args: + prob (float, optional): The flipping probability. Default: None. + direction(str, optional): The flipping direction. Options are + 'horizontal' and 'vertical'. Default: 'horizontal'. + """ + + @deprecated_api_warning({'flip_ratio': 'prob'}, cls_name='RandomFlip') + def __init__(self, prob=None, direction='horizontal'): + self.prob = prob + self.direction = direction + if prob is not None: + assert prob >= 0 and prob <= 1 + assert direction in ['horizontal', 'vertical'] + + def __call__(self, results): + """Call function to flip bounding boxes, masks, semantic segmentation + maps. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Flipped results, 'flip', 'flip_direction' keys are added into + result dict. + """ + + if 'flip' not in results: + flip = True if np.random.rand() < self.prob else False + results['flip'] = flip + if 'flip_direction' not in results: + results['flip_direction'] = self.direction + if results['flip']: + # flip image + results['img'] = mmcv.imflip( + results['img'], direction=results['flip_direction']) + + # flip segs + for key in results.get('seg_fields', []): + # use copy() to make numpy stride positive + results[key] = mmcv.imflip( + results[key], direction=results['flip_direction']).copy() + return results + + def __repr__(self): + return self.__class__.__name__ + f'(prob={self.prob})' + + +@PIPELINES.register_module() +class Pad(object): + """Pad the image & mask. + + There are two padding modes: (1) pad to a fixed size and (2) pad to the + minimum size that is divisible by some number. + Added keys are "pad_shape", "pad_fixed_size", "pad_size_divisor", + + Args: + size (tuple, optional): Fixed padding size. + size_divisor (int, optional): The divisor of padded size. + pad_val (float, optional): Padding value. Default: 0. + seg_pad_val (float, optional): Padding value of segmentation map. + Default: 255. + """ + + def __init__(self, + size=None, + size_divisor=None, + pad_val=0, + seg_pad_val=255): + self.size = size + self.size_divisor = size_divisor + self.pad_val = pad_val + self.seg_pad_val = seg_pad_val + # only one of size and size_divisor should be valid + assert size is not None or size_divisor is not None + assert size is None or size_divisor is None + + def _pad_img(self, results): + """Pad images according to ``self.size``.""" + if self.size is not None: + padded_img = mmcv.impad( + results['img'], shape=self.size, pad_val=self.pad_val) + elif self.size_divisor is not None: + padded_img = mmcv.impad_to_multiple( + results['img'], self.size_divisor, pad_val=self.pad_val) + results['img'] = padded_img + results['pad_shape'] = padded_img.shape + results['pad_fixed_size'] = self.size + results['pad_size_divisor'] = self.size_divisor + + def _pad_seg(self, results): + """Pad masks according to ``results['pad_shape']``.""" + for key in results.get('seg_fields', []): + results[key] = mmcv.impad( + results[key], + shape=results['pad_shape'][:2], + pad_val=self.seg_pad_val) + + def __call__(self, results): + """Call function to pad images, masks, semantic segmentation maps. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Updated result dict. + """ + + self._pad_img(results) + self._pad_seg(results) + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(size={self.size}, size_divisor={self.size_divisor}, ' \ + f'pad_val={self.pad_val})' + return repr_str + + +@PIPELINES.register_module() +class Normalize(object): + """Normalize the image. + + Added key is "img_norm_cfg". + + Args: + mean (sequence): Mean values of 3 channels. + std (sequence): Std values of 3 channels. + to_rgb (bool): Whether to convert the image from BGR to RGB, + default is true. + """ + + def __init__(self, mean, std, to_rgb=True): + self.mean = np.array(mean, dtype=np.float32) + self.std = np.array(std, dtype=np.float32) + self.to_rgb = to_rgb + + def __call__(self, results): + """Call function to normalize images. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Normalized results, 'img_norm_cfg' key is added into + result dict. + """ + + results['img'] = mmcv.imnormalize(results['img'], self.mean, self.std, + self.to_rgb) + results['img_norm_cfg'] = dict( + mean=self.mean, std=self.std, to_rgb=self.to_rgb) + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(mean={self.mean}, std={self.std}, to_rgb=' \ + f'{self.to_rgb})' + return repr_str + + +@PIPELINES.register_module() +class Rerange(object): + """Rerange the image pixel value. + + Args: + min_value (float or int): Minimum value of the reranged image. + Default: 0. + max_value (float or int): Maximum value of the reranged image. + Default: 255. + """ + + def __init__(self, min_value=0, max_value=255): + assert isinstance(min_value, float) or isinstance(min_value, int) + assert isinstance(max_value, float) or isinstance(max_value, int) + assert min_value < max_value + self.min_value = min_value + self.max_value = max_value + + def __call__(self, results): + """Call function to rerange images. + + Args: + results (dict): Result dict from loading pipeline. + Returns: + dict: Reranged results. + """ + + img = results['img'] + img_min_value = np.min(img) + img_max_value = np.max(img) + + assert img_min_value < img_max_value + # rerange to [0, 1] + img = (img - img_min_value) / (img_max_value - img_min_value) + # rerange to [min_value, max_value] + img = img * (self.max_value - self.min_value) + self.min_value + results['img'] = img + + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(min_value={self.min_value}, max_value={self.max_value})' + return repr_str + + +@PIPELINES.register_module() +class CLAHE(object): + """Use CLAHE method to process the image. + + See `ZUIDERVELD,K. Contrast Limited Adaptive Histogram Equalization[J]. + Graphics Gems, 1994:474-485.` for more information. + + Args: + clip_limit (float): Threshold for contrast limiting. Default: 40.0. + tile_grid_size (tuple[int]): Size of grid for histogram equalization. + Input image will be divided into equally sized rectangular tiles. + It defines the number of tiles in row and column. Default: (8, 8). + """ + + def __init__(self, clip_limit=40.0, tile_grid_size=(8, 8)): + assert isinstance(clip_limit, (float, int)) + self.clip_limit = clip_limit + assert is_tuple_of(tile_grid_size, int) + assert len(tile_grid_size) == 2 + self.tile_grid_size = tile_grid_size + + def __call__(self, results): + """Call function to Use CLAHE method process images. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Processed results. + """ + + for i in range(results['img'].shape[2]): + results['img'][:, :, i] = mmcv.clahe( + np.array(results['img'][:, :, i], dtype=np.uint8), + self.clip_limit, self.tile_grid_size) + + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(clip_limit={self.clip_limit}, '\ + f'tile_grid_size={self.tile_grid_size})' + return repr_str + + +@PIPELINES.register_module() +class RandomCrop(object): + """Random crop the image & seg. + + Args: + crop_size (tuple): Expected size after cropping, (h, w). + cat_max_ratio (float): The maximum ratio that single category could + occupy. + """ + + def __init__(self, crop_size, cat_max_ratio=1., ignore_index=255): + assert crop_size[0] > 0 and crop_size[1] > 0 + self.crop_size = crop_size + self.cat_max_ratio = cat_max_ratio + self.ignore_index = ignore_index + + def get_crop_bbox(self, img): + """Randomly get a crop bounding box.""" + margin_h = max(img.shape[0] - self.crop_size[0], 0) + margin_w = max(img.shape[1] - self.crop_size[1], 0) + offset_h = np.random.randint(0, margin_h + 1) + offset_w = np.random.randint(0, margin_w + 1) + crop_y1, crop_y2 = offset_h, offset_h + self.crop_size[0] + crop_x1, crop_x2 = offset_w, offset_w + self.crop_size[1] + + return crop_y1, crop_y2, crop_x1, crop_x2 + + def crop(self, img, crop_bbox): + """Crop from ``img``""" + crop_y1, crop_y2, crop_x1, crop_x2 = crop_bbox + img = img[crop_y1:crop_y2, crop_x1:crop_x2, ...] + return img + + def __call__(self, results): + """Call function to randomly crop images, semantic segmentation maps. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Randomly cropped results, 'img_shape' key in result dict is + updated according to crop size. + """ + + img = results['img'] + crop_bbox = self.get_crop_bbox(img) + if self.cat_max_ratio < 1.: + # Repeat 10 times + for _ in range(10): + seg_temp = self.crop(results['gt_semantic_seg'], crop_bbox) + labels, cnt = np.unique(seg_temp, return_counts=True) + cnt = cnt[labels != self.ignore_index] + if len(cnt) > 1 and np.max(cnt) / np.sum( + cnt) < self.cat_max_ratio: + break + crop_bbox = self.get_crop_bbox(img) + + # crop the image + img = self.crop(img, crop_bbox) + img_shape = img.shape + results['img'] = img + results['img_shape'] = img_shape + + # crop semantic seg + for key in results.get('seg_fields', []): + results[key] = self.crop(results[key], crop_bbox) + + return results + + def __repr__(self): + return self.__class__.__name__ + f'(crop_size={self.crop_size})' + + +@PIPELINES.register_module() +class RandomRotate(object): + """Rotate the image & seg. + + Args: + prob (float): The rotation probability. + degree (float, tuple[float]): Range of degrees to select from. If + degree is a number instead of tuple like (min, max), + the range of degree will be (``-degree``, ``+degree``) + pad_val (float, optional): Padding value of image. Default: 0. + seg_pad_val (float, optional): Padding value of segmentation map. + Default: 255. + center (tuple[float], optional): Center point (w, h) of the rotation in + the source image. If not specified, the center of the image will be + used. Default: None. + auto_bound (bool): Whether to adjust the image size to cover the whole + rotated image. Default: False + """ + + def __init__(self, + prob, + degree, + pad_val=0, + seg_pad_val=255, + center=None, + auto_bound=False): + self.prob = prob + assert prob >= 0 and prob <= 1 + if isinstance(degree, (float, int)): + assert degree > 0, f'degree {degree} should be positive' + self.degree = (-degree, degree) + else: + self.degree = degree + assert len(self.degree) == 2, f'degree {self.degree} should be a ' \ + f'tuple of (min, max)' + self.pal_val = pad_val + self.seg_pad_val = seg_pad_val + self.center = center + self.auto_bound = auto_bound + + def __call__(self, results): + """Call function to rotate image, semantic segmentation maps. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Rotated results. + """ + + rotate = True if np.random.rand() < self.prob else False + degree = np.random.uniform(min(*self.degree), max(*self.degree)) + if rotate: + # rotate image + results['img'] = mmcv.imrotate( + results['img'], + angle=degree, + border_value=self.pal_val, + center=self.center, + auto_bound=self.auto_bound) + + # rotate segs + for key in results.get('seg_fields', []): + results[key] = mmcv.imrotate( + results[key], + angle=degree, + border_value=self.seg_pad_val, + center=self.center, + auto_bound=self.auto_bound, + interpolation='nearest') + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(prob={self.prob}, ' \ + f'degree={self.degree}, ' \ + f'pad_val={self.pal_val}, ' \ + f'seg_pad_val={self.seg_pad_val}, ' \ + f'center={self.center}, ' \ + f'auto_bound={self.auto_bound})' + return repr_str + + +@PIPELINES.register_module() +class RGB2Gray(object): + """Convert RGB image to grayscale image. + + This transform calculate the weighted mean of input image channels with + ``weights`` and then expand the channels to ``out_channels``. When + ``out_channels`` is None, the number of output channels is the same as + input channels. + + Args: + out_channels (int): Expected number of output channels after + transforming. Default: None. + weights (tuple[float]): The weights to calculate the weighted mean. + Default: (0.299, 0.587, 0.114). + """ + + def __init__(self, out_channels=None, weights=(0.299, 0.587, 0.114)): + assert out_channels is None or out_channels > 0 + self.out_channels = out_channels + assert isinstance(weights, tuple) + for item in weights: + assert isinstance(item, (float, int)) + self.weights = weights + + def __call__(self, results): + """Call function to convert RGB image to grayscale image. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Result dict with grayscale image. + """ + img = results['img'] + assert len(img.shape) == 3 + assert img.shape[2] == len(self.weights) + weights = np.array(self.weights).reshape((1, 1, -1)) + img = (img * weights).sum(2, keepdims=True) + if self.out_channels is None: + img = img.repeat(weights.shape[2], axis=2) + else: + img = img.repeat(self.out_channels, axis=2) + + results['img'] = img + results['img_shape'] = img.shape + + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(out_channels={self.out_channels}, ' \ + f'weights={self.weights})' + return repr_str + + +@PIPELINES.register_module() +class AdjustGamma(object): + """Using gamma correction to process the image. + + Args: + gamma (float or int): Gamma value used in gamma correction. + Default: 1.0. + """ + + def __init__(self, gamma=1.0): + assert isinstance(gamma, float) or isinstance(gamma, int) + assert gamma > 0 + self.gamma = gamma + inv_gamma = 1.0 / gamma + self.table = np.array([(i / 255.0)**inv_gamma * 255 + for i in np.arange(256)]).astype('uint8') + + def __call__(self, results): + """Call function to process the image with gamma correction. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Processed results. + """ + + results['img'] = mmcv.lut_transform( + np.array(results['img'], dtype=np.uint8), self.table) + + return results + + def __repr__(self): + return self.__class__.__name__ + f'(gamma={self.gamma})' + + +@PIPELINES.register_module() +class SegRescale(object): + """Rescale semantic segmentation maps. + + Args: + scale_factor (float): The scale factor of the final output. + """ + + def __init__(self, scale_factor=1): + self.scale_factor = scale_factor + + def __call__(self, results): + """Call function to scale the semantic segmentation map. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Result dict with semantic segmentation map scaled. + """ + for key in results.get('seg_fields', []): + if self.scale_factor != 1: + results[key] = mmcv.imrescale( + results[key], self.scale_factor, interpolation='nearest') + return results + + def __repr__(self): + return self.__class__.__name__ + f'(scale_factor={self.scale_factor})' + + +@PIPELINES.register_module() +class PhotoMetricDistortion(object): + """Apply photometric distortion to image sequentially, every transformation + is applied with a probability of 0.5. The position of random contrast is in + second or second to last. + + 1. random brightness + 2. random contrast (mode 0) + 3. convert color from BGR to HSV + 4. random saturation + 5. random hue + 6. convert color from HSV to BGR + 7. random contrast (mode 1) + + Args: + brightness_delta (int): delta of brightness. + contrast_range (tuple): range of contrast. + saturation_range (tuple): range of saturation. + hue_delta (int): delta of hue. + """ + + def __init__(self, + brightness_delta=32, + contrast_range=(0.5, 1.5), + saturation_range=(0.5, 1.5), + hue_delta=18): + self.brightness_delta = brightness_delta + self.contrast_lower, self.contrast_upper = contrast_range + self.saturation_lower, self.saturation_upper = saturation_range + self.hue_delta = hue_delta + + def convert(self, img, alpha=1, beta=0): + """Multiple with alpha and add beat with clip.""" + img = img.astype(np.float32) * alpha + beta + img = np.clip(img, 0, 255) + return img.astype(np.uint8) + + def brightness(self, img): + """Brightness distortion.""" + if random.randint(2): + return self.convert( + img, + beta=random.uniform(-self.brightness_delta, + self.brightness_delta)) + return img + + def contrast(self, img): + """Contrast distortion.""" + if random.randint(2): + return self.convert( + img, + alpha=random.uniform(self.contrast_lower, self.contrast_upper)) + return img + + def saturation(self, img): + """Saturation distortion.""" + if random.randint(2): + img = mmcv.bgr2hsv(img) + img[:, :, 1] = self.convert( + img[:, :, 1], + alpha=random.uniform(self.saturation_lower, + self.saturation_upper)) + img = mmcv.hsv2bgr(img) + return img + + def hue(self, img): + """Hue distortion.""" + if random.randint(2): + img = mmcv.bgr2hsv(img) + img[:, :, + 0] = (img[:, :, 0].astype(int) + + random.randint(-self.hue_delta, self.hue_delta)) % 180 + img = mmcv.hsv2bgr(img) + return img + + def __call__(self, results): + """Call function to perform photometric distortion on images. + + Args: + results (dict): Result dict from loading pipeline. + + Returns: + dict: Result dict with images distorted. + """ + + img = results['img'] + # random brightness + img = self.brightness(img) + + # mode == 0 --> do random contrast first + # mode == 1 --> do random contrast last + mode = random.randint(2) + if mode == 1: + img = self.contrast(img) + + # random saturation + img = self.saturation(img) + + # random hue + img = self.hue(img) + + # random contrast + if mode == 0: + img = self.contrast(img) + + results['img'] = img + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += (f'(brightness_delta={self.brightness_delta}, ' + f'contrast_range=({self.contrast_lower}, ' + f'{self.contrast_upper}), ' + f'saturation_range=({self.saturation_lower}, ' + f'{self.saturation_upper}), ' + f'hue_delta={self.hue_delta})') + return repr_str + + +@PIPELINES.register_module() +class RandomCutOut(object): + """CutOut operation. + + Randomly drop some regions of image used in + `Cutout `_. + Args: + prob (float): cutout probability. + n_holes (int | tuple[int, int]): Number of regions to be dropped. + If it is given as a list, number of holes will be randomly + selected from the closed interval [`n_holes[0]`, `n_holes[1]`]. + cutout_shape (tuple[int, int] | list[tuple[int, int]]): The candidate + shape of dropped regions. It can be `tuple[int, int]` to use a + fixed cutout shape, or `list[tuple[int, int]]` to randomly choose + shape from the list. + cutout_ratio (tuple[float, float] | list[tuple[float, float]]): The + candidate ratio of dropped regions. It can be `tuple[float, float]` + to use a fixed ratio or `list[tuple[float, float]]` to randomly + choose ratio from the list. Please note that `cutout_shape` + and `cutout_ratio` cannot be both given at the same time. + fill_in (tuple[float, float, float] | tuple[int, int, int]): The value + of pixel to fill in the dropped regions. Default: (0, 0, 0). + seg_fill_in (int): The labels of pixel to fill in the dropped regions. + If seg_fill_in is None, skip. Default: None. + """ + + def __init__(self, + prob, + n_holes, + cutout_shape=None, + cutout_ratio=None, + fill_in=(0, 0, 0), + seg_fill_in=None): + + assert 0 <= prob and prob <= 1 + assert (cutout_shape is None) ^ (cutout_ratio is None), \ + 'Either cutout_shape or cutout_ratio should be specified.' + assert (isinstance(cutout_shape, (list, tuple)) + or isinstance(cutout_ratio, (list, tuple))) + if isinstance(n_holes, tuple): + assert len(n_holes) == 2 and 0 <= n_holes[0] < n_holes[1] + else: + n_holes = (n_holes, n_holes) + if seg_fill_in is not None: + assert (isinstance(seg_fill_in, int) and 0 <= seg_fill_in + and seg_fill_in <= 255) + self.prob = prob + self.n_holes = n_holes + self.fill_in = fill_in + self.seg_fill_in = seg_fill_in + self.with_ratio = cutout_ratio is not None + self.candidates = cutout_ratio if self.with_ratio else cutout_shape + if not isinstance(self.candidates, list): + self.candidates = [self.candidates] + + def __call__(self, results): + """Call function to drop some regions of image.""" + cutout = True if np.random.rand() < self.prob else False + if cutout: + h, w, c = results['img'].shape + n_holes = np.random.randint(self.n_holes[0], self.n_holes[1] + 1) + for _ in range(n_holes): + x1 = np.random.randint(0, w) + y1 = np.random.randint(0, h) + index = np.random.randint(0, len(self.candidates)) + if not self.with_ratio: + cutout_w, cutout_h = self.candidates[index] + else: + cutout_w = int(self.candidates[index][0] * w) + cutout_h = int(self.candidates[index][1] * h) + + x2 = np.clip(x1 + cutout_w, 0, w) + y2 = np.clip(y1 + cutout_h, 0, h) + results['img'][y1:y2, x1:x2, :] = self.fill_in + + if self.seg_fill_in is not None: + for key in results.get('seg_fields', []): + results[key][y1:y2, x1:x2] = self.seg_fill_in + + return results + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(prob={self.prob}, ' + repr_str += f'n_holes={self.n_holes}, ' + repr_str += (f'cutout_ratio={self.candidates}, ' if self.with_ratio + else f'cutout_shape={self.candidates}, ') + repr_str += f'fill_in={self.fill_in}, ' + repr_str += f'seg_fill_in={self.seg_fill_in})' + return repr_str + + +@PIPELINES.register_module() +class RandomMosaic(object): + """Mosaic augmentation. Given 4 images, mosaic transform combines them into + one output image. The output image is composed of the parts from each sub- + image. + + .. code:: text + + mosaic transform + center_x + +------------------------------+ + | pad | pad | + | +-----------+ | + | | | | + | | image1 |--------+ | + | | | | | + | | | image2 | | + center_y |----+-------------+-----------| + | | cropped | | + |pad | image3 | image4 | + | | | | + +----|-------------+-----------+ + | | + +-------------+ + + The mosaic transform steps are as follows: + 1. Choose the mosaic center as the intersections of 4 images + 2. Get the left top image according to the index, and randomly + sample another 3 images from the custom dataset. + 3. Sub image will be cropped if image is larger than mosaic patch + + Args: + prob (float): mosaic probability. + img_scale (Sequence[int]): Image size after mosaic pipeline of + a single image. The size of the output image is four times + that of a single image. The output image comprises 4 single images. + Default: (640, 640). + center_ratio_range (Sequence[float]): Center ratio range of mosaic + output. Default: (0.5, 1.5). + pad_val (int): Pad value. Default: 0. + seg_pad_val (int): Pad value of segmentation map. Default: 255. + """ + + def __init__(self, + prob, + img_scale=(640, 640), + center_ratio_range=(0.5, 1.5), + pad_val=0, + seg_pad_val=255): + assert 0 <= prob and prob <= 1 + assert isinstance(img_scale, tuple) + self.prob = prob + self.img_scale = img_scale + self.center_ratio_range = center_ratio_range + self.pad_val = pad_val + self.seg_pad_val = seg_pad_val + + def __call__(self, results): + """Call function to make a mosaic of image. + + Args: + results (dict): Result dict. + + Returns: + dict: Result dict with mosaic transformed. + """ + mosaic = True if np.random.rand() < self.prob else False + if mosaic: + results = self._mosaic_transform_img(results) + results = self._mosaic_transform_seg(results) + return results + + def get_indexes(self, dataset): + """Call function to collect indexes. + + Args: + dataset (:obj:`MultiImageMixDataset`): The dataset. + + Returns: + list: indexes. + """ + + indexes = [random.randint(0, len(dataset)) for _ in range(3)] + return indexes + + def _mosaic_transform_img(self, results): + """Mosaic transform function. + + Args: + results (dict): Result dict. + + Returns: + dict: Updated result dict. + """ + + assert 'mix_results' in results + if len(results['img'].shape) == 3: + mosaic_img = np.full( + (int(self.img_scale[0] * 2), int(self.img_scale[1] * 2), 3), + self.pad_val, + dtype=results['img'].dtype) + else: + mosaic_img = np.full( + (int(self.img_scale[0] * 2), int(self.img_scale[1] * 2)), + self.pad_val, + dtype=results['img'].dtype) + + # mosaic center x, y + self.center_x = int( + random.uniform(*self.center_ratio_range) * self.img_scale[1]) + self.center_y = int( + random.uniform(*self.center_ratio_range) * self.img_scale[0]) + center_position = (self.center_x, self.center_y) + + loc_strs = ('top_left', 'top_right', 'bottom_left', 'bottom_right') + for i, loc in enumerate(loc_strs): + if loc == 'top_left': + result_patch = copy.deepcopy(results) + else: + result_patch = copy.deepcopy(results['mix_results'][i - 1]) + + img_i = result_patch['img'] + h_i, w_i = img_i.shape[:2] + # keep_ratio resize + scale_ratio_i = min(self.img_scale[0] / h_i, + self.img_scale[1] / w_i) + img_i = mmcv.imresize( + img_i, (int(w_i * scale_ratio_i), int(h_i * scale_ratio_i))) + + # compute the combine parameters + paste_coord, crop_coord = self._mosaic_combine( + loc, center_position, img_i.shape[:2][::-1]) + x1_p, y1_p, x2_p, y2_p = paste_coord + x1_c, y1_c, x2_c, y2_c = crop_coord + + # crop and paste image + mosaic_img[y1_p:y2_p, x1_p:x2_p] = img_i[y1_c:y2_c, x1_c:x2_c] + + results['img'] = mosaic_img + results['img_shape'] = mosaic_img.shape + results['ori_shape'] = mosaic_img.shape + + return results + + def _mosaic_transform_seg(self, results): + """Mosaic transform function for label annotations. + + Args: + results (dict): Result dict. + + Returns: + dict: Updated result dict. + """ + + assert 'mix_results' in results + for key in results.get('seg_fields', []): + mosaic_seg = np.full( + (int(self.img_scale[0] * 2), int(self.img_scale[1] * 2)), + self.seg_pad_val, + dtype=results[key].dtype) + + # mosaic center x, y + center_position = (self.center_x, self.center_y) + + loc_strs = ('top_left', 'top_right', 'bottom_left', 'bottom_right') + for i, loc in enumerate(loc_strs): + if loc == 'top_left': + result_patch = copy.deepcopy(results) + else: + result_patch = copy.deepcopy(results['mix_results'][i - 1]) + + gt_seg_i = result_patch[key] + h_i, w_i = gt_seg_i.shape[:2] + # keep_ratio resize + scale_ratio_i = min(self.img_scale[0] / h_i, + self.img_scale[1] / w_i) + gt_seg_i = mmcv.imresize( + gt_seg_i, + (int(w_i * scale_ratio_i), int(h_i * scale_ratio_i)), + interpolation='nearest') + + # compute the combine parameters + paste_coord, crop_coord = self._mosaic_combine( + loc, center_position, gt_seg_i.shape[:2][::-1]) + x1_p, y1_p, x2_p, y2_p = paste_coord + x1_c, y1_c, x2_c, y2_c = crop_coord + + # crop and paste image + mosaic_seg[y1_p:y2_p, x1_p:x2_p] = gt_seg_i[y1_c:y2_c, + x1_c:x2_c] + + results[key] = mosaic_seg + + return results + + def _mosaic_combine(self, loc, center_position_xy, img_shape_wh): + """Calculate global coordinate of mosaic image and local coordinate of + cropped sub-image. + + Args: + loc (str): Index for the sub-image, loc in ('top_left', + 'top_right', 'bottom_left', 'bottom_right'). + center_position_xy (Sequence[float]): Mixing center for 4 images, + (x, y). + img_shape_wh (Sequence[int]): Width and height of sub-image + + Returns: + tuple[tuple[float]]: Corresponding coordinate of pasting and + cropping + - paste_coord (tuple): paste corner coordinate in mosaic image. + - crop_coord (tuple): crop corner coordinate in mosaic image. + """ + + assert loc in ('top_left', 'top_right', 'bottom_left', 'bottom_right') + if loc == 'top_left': + # index0 to top left part of image + x1, y1, x2, y2 = max(center_position_xy[0] - img_shape_wh[0], 0), \ + max(center_position_xy[1] - img_shape_wh[1], 0), \ + center_position_xy[0], \ + center_position_xy[1] + crop_coord = img_shape_wh[0] - (x2 - x1), img_shape_wh[1] - ( + y2 - y1), img_shape_wh[0], img_shape_wh[1] + + elif loc == 'top_right': + # index1 to top right part of image + x1, y1, x2, y2 = center_position_xy[0], \ + max(center_position_xy[1] - img_shape_wh[1], 0), \ + min(center_position_xy[0] + img_shape_wh[0], + self.img_scale[1] * 2), \ + center_position_xy[1] + crop_coord = 0, img_shape_wh[1] - (y2 - y1), min( + img_shape_wh[0], x2 - x1), img_shape_wh[1] + + elif loc == 'bottom_left': + # index2 to bottom left part of image + x1, y1, x2, y2 = max(center_position_xy[0] - img_shape_wh[0], 0), \ + center_position_xy[1], \ + center_position_xy[0], \ + min(self.img_scale[0] * 2, center_position_xy[1] + + img_shape_wh[1]) + crop_coord = img_shape_wh[0] - (x2 - x1), 0, img_shape_wh[0], min( + y2 - y1, img_shape_wh[1]) + + else: + # index3 to bottom right part of image + x1, y1, x2, y2 = center_position_xy[0], \ + center_position_xy[1], \ + min(center_position_xy[0] + img_shape_wh[0], + self.img_scale[1] * 2), \ + min(self.img_scale[0] * 2, center_position_xy[1] + + img_shape_wh[1]) + crop_coord = 0, 0, min(img_shape_wh[0], + x2 - x1), min(y2 - y1, img_shape_wh[1]) + + paste_coord = x1, y1, x2, y2 + return paste_coord, crop_coord + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(prob={self.prob}, ' + repr_str += f'img_scale={self.img_scale}, ' + repr_str += f'center_ratio_range={self.center_ratio_range}, ' + repr_str += f'pad_val={self.pad_val}, ' + repr_str += f'seg_pad_val={self.pad_val})' + return repr_str diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/samplers/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/samplers/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..da09effaf20fefe1a102277672b98db7d884f002 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/samplers/__init__.py @@ -0,0 +1,4 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .distributed_sampler import DistributedSampler + +__all__ = ['DistributedSampler'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/samplers/distributed_sampler.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/samplers/distributed_sampler.py new file mode 100644 index 0000000000000000000000000000000000000000..d1a13c7162b060ac3047a8a4b5b215ab63ef6454 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/samplers/distributed_sampler.py @@ -0,0 +1,71 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from __future__ import division +from typing import Iterator, Optional + +import torch +from torch.utils.data import Dataset +from torch.utils.data import DistributedSampler as _DistributedSampler + +from mmseg.core.utils import sync_random_seed + + +class DistributedSampler(_DistributedSampler): + """DistributedSampler inheriting from + `torch.utils.data.DistributedSampler`. + + Args: + datasets (Dataset): the dataset will be loaded. + num_replicas (int, optional): Number of processes participating in + distributed training. By default, world_size is retrieved from the + current distributed group. + rank (int, optional): Rank of the current process within num_replicas. + By default, rank is retrieved from the current distributed group. + shuffle (bool): If True (default), sampler will shuffle the indices. + seed (int): random seed used to shuffle the sampler if + :attr:`shuffle=True`. This number should be identical across all + processes in the distributed group. Default: ``0``. + """ + + def __init__(self, + dataset: Dataset, + num_replicas: Optional[int] = None, + rank: Optional[int] = None, + shuffle: bool = True, + seed=0) -> None: + super().__init__( + dataset, num_replicas=num_replicas, rank=rank, shuffle=shuffle) + + # In distributed sampling, different ranks should sample + # non-overlapped data in the dataset. Therefore, this function + # is used to make sure that each rank shuffles the data indices + # in the same order based on the same seed. Then different ranks + # could use different indices to select non-overlapped data from the + # same data list. + self.seed = sync_random_seed(seed) + + def __iter__(self) -> Iterator: + """ + Yields: + Iterator: iterator of indices for rank. + """ + # deterministically shuffle based on epoch + if self.shuffle: + g = torch.Generator() + # When :attr:`shuffle=True`, this ensures all replicas + # use a different random ordering for each epoch. + # Otherwise, the next iteration of this sampler will + # yield the same ordering. + g.manual_seed(self.epoch + self.seed) + indices = torch.randperm(len(self.dataset), generator=g).tolist() + else: + indices = torch.arange(len(self.dataset)).tolist() + + # add extra samples to make it evenly divisible + indices += indices[:(self.total_size - len(indices))] + assert len(indices) == self.total_size + + # subsample + indices = indices[self.rank:self.total_size:self.num_replicas] + assert len(indices) == self.num_samples + + return iter(indices) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/voc.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/voc.py new file mode 100644 index 0000000000000000000000000000000000000000..9eecc344f2bd19df6b9108e37e4360d1a72ee986 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/datasets/voc.py @@ -0,0 +1,39 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os.path as osp + +from .builder import DATASETS +from .custom import CustomDataset + + +@DATASETS.register_module() +class PascalVOCDataset(CustomDataset): + """Pascal VOC dataset. + + Args: + split (str): Split txt file for Pascal VOC. + """ + + CLASSES = ('background', 'aeroplane', 'bicycle', 'bird', 'boat', 'bottle', + 'bus', 'car', 'cat', 'chair', 'cow', 'diningtable', 'dog', + 'horse', 'motorbike', 'person', 'pottedplant', 'sheep', 'sofa', + 'train', 'tvmonitor') + + PALETTE = [[0, 0, 0], [128, 0, 0], [0, 128, 0], [128, 128, 0], [0, 0, 128], + [128, 0, 128], [0, 128, 128], [128, 128, 128], [64, 0, 0], + [192, 0, 0], [64, 128, 0], [192, 128, 0], [64, 0, 128], + [192, 0, 128], [64, 128, 128], [192, 128, 128], [0, 64, 0], + [128, 64, 0], [0, 192, 0], [128, 192, 0], [0, 64, 128]] + + def __init__(self, split, **kwargs): + + if "img_dir" in kwargs: + image_dir = kwargs["img_dir"] + if not osp.join(kwargs['data_root'], image_dir): + image_dir = "images" + if osp.join(kwargs['data_root'], image_dir): + kwargs["img_dir"] = image_dir + + super(PascalVOCDataset, self).__init__( + img_suffix='.jpg', seg_map_suffix='.png', split=split, **kwargs) + + assert osp.exists(self.img_dir) and self.split is not None diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..87d8108e3f1977bf4830fa83ad7498081d2a9a51 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/__init__.py @@ -0,0 +1,13 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .backbones import * # noqa: F401,F403 +from .builder import (BACKBONES, HEADS, LOSSES, SEGMENTORS, build_backbone, + build_head, build_loss, build_segmentor) +from .decode_heads import * # noqa: F401,F403 +from .losses import * # noqa: F401,F403 +from .necks import * # noqa: F401,F403 +from .segmentors import * # noqa: F401,F403 + +__all__ = [ + 'BACKBONES', 'HEADS', 'LOSSES', 'SEGMENTORS', 'build_backbone', + 'build_head', 'build_loss', 'build_segmentor' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/backbones/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/backbones/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..464bfa9f47643a01276c60ff8664e2d017363b88 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/backbones/__init__.py @@ -0,0 +1,3 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .resnet import ResNet, ResNetV1c, ResNetV1d + diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/backbones/resnet.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/backbones/resnet.py new file mode 100644 index 0000000000000000000000000000000000000000..e8b961d5faed3a25eb09576b1339e5ce18ca9627 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/backbones/resnet.py @@ -0,0 +1,714 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch.nn as nn +import torch.utils.checkpoint as cp +from mmcv.cnn import build_conv_layer, build_norm_layer, build_plugin_layer +from mmcv.runner import BaseModule +from mmcv.utils.parrots_wrapper import _BatchNorm + +from ..builder import BACKBONES +from ..utils import ResLayer + + +class BasicBlock(BaseModule): + """Basic block for ResNet.""" + + expansion = 1 + + def __init__(self, + inplanes, + planes, + stride=1, + dilation=1, + downsample=None, + style='pytorch', + with_cp=False, + conv_cfg=None, + norm_cfg=dict(type='BN'), + dcn=None, + plugins=None, + init_cfg=None): + super(BasicBlock, self).__init__(init_cfg) + assert dcn is None, 'Not implemented yet.' + assert plugins is None, 'Not implemented yet.' + + self.norm1_name, norm1 = build_norm_layer(norm_cfg, planes, postfix=1) + self.norm2_name, norm2 = build_norm_layer(norm_cfg, planes, postfix=2) + + self.conv1 = build_conv_layer( + conv_cfg, + inplanes, + planes, + 3, + stride=stride, + padding=dilation, + dilation=dilation, + bias=False) + self.add_module(self.norm1_name, norm1) + self.conv2 = build_conv_layer( + conv_cfg, planes, planes, 3, padding=1, bias=False) + self.add_module(self.norm2_name, norm2) + + self.relu = nn.ReLU(inplace=True) + self.downsample = downsample + self.stride = stride + self.dilation = dilation + self.with_cp = with_cp + + @property + def norm1(self): + """nn.Module: normalization layer after the first convolution layer""" + return getattr(self, self.norm1_name) + + @property + def norm2(self): + """nn.Module: normalization layer after the second convolution layer""" + return getattr(self, self.norm2_name) + + def forward(self, x): + """Forward function.""" + + def _inner_forward(x): + identity = x + + out = self.conv1(x) + out = self.norm1(out) + out = self.relu(out) + + out = self.conv2(out) + out = self.norm2(out) + + if self.downsample is not None: + identity = self.downsample(x) + + out += identity + + return out + + if self.with_cp and x.requires_grad: + out = cp.checkpoint(_inner_forward, x) + else: + out = _inner_forward(x) + + out = self.relu(out) + + return out + + +class Bottleneck(BaseModule): + """Bottleneck block for ResNet. + + If style is "pytorch", the stride-two layer is the 3x3 conv layer, if it is + "caffe", the stride-two layer is the first 1x1 conv layer. + """ + + expansion = 4 + + def __init__(self, + inplanes, + planes, + stride=1, + dilation=1, + downsample=None, + style='pytorch', + with_cp=False, + conv_cfg=None, + norm_cfg=dict(type='BN'), + dcn=None, + plugins=None, + init_cfg=None): + super(Bottleneck, self).__init__(init_cfg) + assert style in ['pytorch', 'caffe'] + assert dcn is None or isinstance(dcn, dict) + assert plugins is None or isinstance(plugins, list) + if plugins is not None: + allowed_position = ['after_conv1', 'after_conv2', 'after_conv3'] + assert all(p['position'] in allowed_position for p in plugins) + + self.inplanes = inplanes + self.planes = planes + self.stride = stride + self.dilation = dilation + self.style = style + self.with_cp = with_cp + self.conv_cfg = conv_cfg + self.norm_cfg = norm_cfg + self.dcn = dcn + self.with_dcn = dcn is not None + self.plugins = plugins + self.with_plugins = plugins is not None + + if self.with_plugins: + # collect plugins for conv1/conv2/conv3 + self.after_conv1_plugins = [ + plugin['cfg'] for plugin in plugins + if plugin['position'] == 'after_conv1' + ] + self.after_conv2_plugins = [ + plugin['cfg'] for plugin in plugins + if plugin['position'] == 'after_conv2' + ] + self.after_conv3_plugins = [ + plugin['cfg'] for plugin in plugins + if plugin['position'] == 'after_conv3' + ] + + if self.style == 'pytorch': + self.conv1_stride = 1 + self.conv2_stride = stride + else: + self.conv1_stride = stride + self.conv2_stride = 1 + + self.norm1_name, norm1 = build_norm_layer(norm_cfg, planes, postfix=1) + self.norm2_name, norm2 = build_norm_layer(norm_cfg, planes, postfix=2) + self.norm3_name, norm3 = build_norm_layer( + norm_cfg, planes * self.expansion, postfix=3) + + self.conv1 = build_conv_layer( + conv_cfg, + inplanes, + planes, + kernel_size=1, + stride=self.conv1_stride, + bias=False) + self.add_module(self.norm1_name, norm1) + fallback_on_stride = False + if self.with_dcn: + fallback_on_stride = dcn.pop('fallback_on_stride', False) + if not self.with_dcn or fallback_on_stride: + self.conv2 = build_conv_layer( + conv_cfg, + planes, + planes, + kernel_size=3, + stride=self.conv2_stride, + padding=dilation, + dilation=dilation, + bias=False) + else: + assert self.conv_cfg is None, 'conv_cfg must be None for DCN' + self.conv2 = build_conv_layer( + dcn, + planes, + planes, + kernel_size=3, + stride=self.conv2_stride, + padding=dilation, + dilation=dilation, + bias=False) + + self.add_module(self.norm2_name, norm2) + self.conv3 = build_conv_layer( + conv_cfg, + planes, + planes * self.expansion, + kernel_size=1, + bias=False) + self.add_module(self.norm3_name, norm3) + + self.relu = nn.ReLU(inplace=True) + self.downsample = downsample + + if self.with_plugins: + self.after_conv1_plugin_names = self.make_block_plugins( + planes, self.after_conv1_plugins) + self.after_conv2_plugin_names = self.make_block_plugins( + planes, self.after_conv2_plugins) + self.after_conv3_plugin_names = self.make_block_plugins( + planes * self.expansion, self.after_conv3_plugins) + + def make_block_plugins(self, in_channels, plugins): + """make plugins for block. + + Args: + in_channels (int): Input channels of plugin. + plugins (list[dict]): List of plugins cfg to build. + + Returns: + list[str]: List of the names of plugin. + """ + assert isinstance(plugins, list) + plugin_names = [] + for plugin in plugins: + plugin = plugin.copy() + name, layer = build_plugin_layer( + plugin, + in_channels=in_channels, + postfix=plugin.pop('postfix', '')) + assert not hasattr(self, name), f'duplicate plugin {name}' + self.add_module(name, layer) + plugin_names.append(name) + return plugin_names + + def forward_plugin(self, x, plugin_names): + """Forward function for plugins.""" + out = x + for name in plugin_names: + out = getattr(self, name)(x) + return out + + @property + def norm1(self): + """nn.Module: normalization layer after the first convolution layer""" + return getattr(self, self.norm1_name) + + @property + def norm2(self): + """nn.Module: normalization layer after the second convolution layer""" + return getattr(self, self.norm2_name) + + @property + def norm3(self): + """nn.Module: normalization layer after the third convolution layer""" + return getattr(self, self.norm3_name) + + def forward(self, x): + """Forward function.""" + + def _inner_forward(x): + identity = x + + out = self.conv1(x) + out = self.norm1(out) + out = self.relu(out) + + if self.with_plugins: + out = self.forward_plugin(out, self.after_conv1_plugin_names) + + out = self.conv2(out) + out = self.norm2(out) + out = self.relu(out) + + if self.with_plugins: + out = self.forward_plugin(out, self.after_conv2_plugin_names) + + out = self.conv3(out) + out = self.norm3(out) + + if self.with_plugins: + out = self.forward_plugin(out, self.after_conv3_plugin_names) + + if self.downsample is not None: + identity = self.downsample(x) + + out += identity + + return out + + if self.with_cp and x.requires_grad: + out = cp.checkpoint(_inner_forward, x) + else: + out = _inner_forward(x) + + out = self.relu(out) + + return out + + +@BACKBONES.register_module() +class ResNet(BaseModule): + """ResNet backbone. + + This backbone is the improved implementation of `Deep Residual Learning + for Image Recognition `_. + + Args: + depth (int): Depth of resnet, from {18, 34, 50, 101, 152}. + in_channels (int): Number of input image channels. Default: 3. + stem_channels (int): Number of stem channels. Default: 64. + base_channels (int): Number of base channels of res layer. Default: 64. + num_stages (int): Resnet stages, normally 4. Default: 4. + strides (Sequence[int]): Strides of the first block of each stage. + Default: (1, 2, 2, 2). + dilations (Sequence[int]): Dilation of each stage. + Default: (1, 1, 1, 1). + out_indices (Sequence[int]): Output from which stages. + Default: (0, 1, 2, 3). + style (str): `pytorch` or `caffe`. If set to "pytorch", the stride-two + layer is the 3x3 conv layer, otherwise the stride-two layer is + the first 1x1 conv layer. Default: 'pytorch'. + deep_stem (bool): Replace 7x7 conv in input stem with 3 3x3 conv. + Default: False. + avg_down (bool): Use AvgPool instead of stride conv when + downsampling in the bottleneck. Default: False. + frozen_stages (int): Stages to be frozen (stop grad and set eval mode). + -1 means not freezing any parameters. Default: -1. + conv_cfg (dict | None): Dictionary to construct and config conv layer. + When conv_cfg is None, cfg will be set to dict(type='Conv2d'). + Default: None. + norm_cfg (dict): Dictionary to construct and config norm layer. + Default: dict(type='BN', requires_grad=True). + norm_eval (bool): Whether to set norm layers to eval mode, namely, + freeze running stats (mean and var). Note: Effect on Batch Norm + and its variants only. Default: False. + dcn (dict | None): Dictionary to construct and config DCN conv layer. + When dcn is not None, conv_cfg must be None. Default: None. + stage_with_dcn (Sequence[bool]): Whether to set DCN conv for each + stage. The length of stage_with_dcn is equal to num_stages. + Default: (False, False, False, False). + plugins (list[dict]): List of plugins for stages, each dict contains: + + - cfg (dict, required): Cfg dict to build plugin. + + - position (str, required): Position inside block to insert plugin, + options: 'after_conv1', 'after_conv2', 'after_conv3'. + + - stages (tuple[bool], optional): Stages to apply plugin, length + should be same as 'num_stages'. + Default: None. + multi_grid (Sequence[int]|None): Multi grid dilation rates of last + stage. Default: None. + contract_dilation (bool): Whether contract first dilation of each layer + Default: False. + with_cp (bool): Use checkpoint or not. Using checkpoint will save some + memory while slowing down the training speed. Default: False. + zero_init_residual (bool): Whether to use zero init for last norm layer + in resblocks to let them behave as identity. Default: True. + pretrained (str, optional): model pretrained path. Default: None. + init_cfg (dict or list[dict], optional): Initialization config dict. + Default: None. + + Example: + >>> from mmseg.models import ResNet + >>> import torch + >>> self = ResNet(depth=18) + >>> self.eval() + >>> inputs = torch.rand(1, 3, 32, 32) + >>> level_outputs = self.forward(inputs) + >>> for level_out in level_outputs: + ... print(tuple(level_out.shape)) + (1, 64, 8, 8) + (1, 128, 4, 4) + (1, 256, 2, 2) + (1, 512, 1, 1) + """ + + arch_settings = { + 18: (BasicBlock, (2, 2, 2, 2)), + 34: (BasicBlock, (3, 4, 6, 3)), + 50: (Bottleneck, (3, 4, 6, 3)), + 101: (Bottleneck, (3, 4, 23, 3)), + 152: (Bottleneck, (3, 8, 36, 3)) + } + + def __init__(self, + depth, + in_channels=3, + stem_channels=64, + base_channels=64, + num_stages=4, + strides=(1, 2, 2, 2), + dilations=(1, 1, 1, 1), + out_indices=(0, 1, 2, 3), + style='pytorch', + deep_stem=False, + avg_down=False, + frozen_stages=-1, + conv_cfg=None, + norm_cfg=dict(type='BN', requires_grad=True), + norm_eval=False, + dcn=None, + stage_with_dcn=(False, False, False, False), + plugins=None, + multi_grid=None, + contract_dilation=False, + with_cp=False, + zero_init_residual=True, + pretrained=None, + init_cfg=None): + super(ResNet, self).__init__(init_cfg) + if depth not in self.arch_settings: + raise KeyError(f'invalid depth {depth} for resnet') + + self.pretrained = pretrained + self.zero_init_residual = zero_init_residual + block_init_cfg = None + assert not (init_cfg and pretrained), \ + 'init_cfg and pretrained cannot be setting at the same time' + if isinstance(pretrained, str): + warnings.warn('DeprecationWarning: pretrained is a deprecated, ' + 'please use "init_cfg" instead') + self.init_cfg = dict(type='Pretrained', checkpoint=pretrained) + elif pretrained is None: + if init_cfg is None: + self.init_cfg = [ + dict(type='Kaiming', layer='Conv2d'), + dict( + type='Constant', + val=1, + layer=['_BatchNorm', 'GroupNorm']) + ] + block = self.arch_settings[depth][0] + if self.zero_init_residual: + if block is BasicBlock: + block_init_cfg = dict( + type='Constant', + val=0, + override=dict(name='norm2')) + elif block is Bottleneck: + block_init_cfg = dict( + type='Constant', + val=0, + override=dict(name='norm3')) + else: + raise TypeError('pretrained must be a str or None') + + self.depth = depth + self.stem_channels = stem_channels + self.base_channels = base_channels + self.num_stages = num_stages + assert num_stages >= 1 and num_stages <= 4 + self.strides = strides + self.dilations = dilations + assert len(strides) == len(dilations) == num_stages + self.out_indices = out_indices + assert max(out_indices) < num_stages + self.style = style + self.deep_stem = deep_stem + self.avg_down = avg_down + self.frozen_stages = frozen_stages + self.conv_cfg = conv_cfg + self.norm_cfg = norm_cfg + self.with_cp = with_cp + self.norm_eval = norm_eval + self.dcn = dcn + self.stage_with_dcn = stage_with_dcn + if dcn is not None: + assert len(stage_with_dcn) == num_stages + self.plugins = plugins + self.multi_grid = multi_grid + self.contract_dilation = contract_dilation + self.block, stage_blocks = self.arch_settings[depth] + self.stage_blocks = stage_blocks[:num_stages] + self.inplanes = stem_channels + + self._make_stem_layer(in_channels, stem_channels) + + self.res_layers = [] + for i, num_blocks in enumerate(self.stage_blocks): + stride = strides[i] + dilation = dilations[i] + dcn = self.dcn if self.stage_with_dcn[i] else None + if plugins is not None: + stage_plugins = self.make_stage_plugins(plugins, i) + else: + stage_plugins = None + # multi grid is applied to last layer only + stage_multi_grid = multi_grid if i == len( + self.stage_blocks) - 1 else None + planes = base_channels * 2**i + res_layer = self.make_res_layer( + block=self.block, + inplanes=self.inplanes, + planes=planes, + num_blocks=num_blocks, + stride=stride, + dilation=dilation, + style=self.style, + avg_down=self.avg_down, + with_cp=with_cp, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + dcn=dcn, + plugins=stage_plugins, + multi_grid=stage_multi_grid, + contract_dilation=contract_dilation, + init_cfg=block_init_cfg) + self.inplanes = planes * self.block.expansion + layer_name = f'layer{i+1}' + self.add_module(layer_name, res_layer) + self.res_layers.append(layer_name) + + self._freeze_stages() + + self.feat_dim = self.block.expansion * base_channels * 2**( + len(self.stage_blocks) - 1) + + def make_stage_plugins(self, plugins, stage_idx): + """make plugins for ResNet 'stage_idx'th stage . + + Currently we support to insert 'context_block', + 'empirical_attention_block', 'nonlocal_block' into the backbone like + ResNet/ResNeXt. They could be inserted after conv1/conv2/conv3 of + Bottleneck. + + An example of plugins format could be : + >>> plugins=[ + ... dict(cfg=dict(type='xxx', arg1='xxx'), + ... stages=(False, True, True, True), + ... position='after_conv2'), + ... dict(cfg=dict(type='yyy'), + ... stages=(True, True, True, True), + ... position='after_conv3'), + ... dict(cfg=dict(type='zzz', postfix='1'), + ... stages=(True, True, True, True), + ... position='after_conv3'), + ... dict(cfg=dict(type='zzz', postfix='2'), + ... stages=(True, True, True, True), + ... position='after_conv3') + ... ] + >>> self = ResNet(depth=18) + >>> stage_plugins = self.make_stage_plugins(plugins, 0) + >>> assert len(stage_plugins) == 3 + + Suppose 'stage_idx=0', the structure of blocks in the stage would be: + conv1-> conv2->conv3->yyy->zzz1->zzz2 + Suppose 'stage_idx=1', the structure of blocks in the stage would be: + conv1-> conv2->xxx->conv3->yyy->zzz1->zzz2 + + If stages is missing, the plugin would be applied to all stages. + + Args: + plugins (list[dict]): List of plugins cfg to build. The postfix is + required if multiple same type plugins are inserted. + stage_idx (int): Index of stage to build + + Returns: + list[dict]: Plugins for current stage + """ + stage_plugins = [] + for plugin in plugins: + plugin = plugin.copy() + stages = plugin.pop('stages', None) + assert stages is None or len(stages) == self.num_stages + # whether to insert plugin into current stage + if stages is None or stages[stage_idx]: + stage_plugins.append(plugin) + + return stage_plugins + + def make_res_layer(self, **kwargs): + """Pack all blocks in a stage into a ``ResLayer``.""" + return ResLayer(**kwargs) + + @property + def norm1(self): + """nn.Module: the normalization layer named "norm1" """ + return getattr(self, self.norm1_name) + + def _make_stem_layer(self, in_channels, stem_channels): + """Make stem layer for ResNet.""" + if self.deep_stem: + self.stem = nn.Sequential( + build_conv_layer( + self.conv_cfg, + in_channels, + stem_channels // 2, + kernel_size=3, + stride=2, + padding=1, + bias=False), + build_norm_layer(self.norm_cfg, stem_channels // 2)[1], + nn.ReLU(inplace=True), + build_conv_layer( + self.conv_cfg, + stem_channels // 2, + stem_channels // 2, + kernel_size=3, + stride=1, + padding=1, + bias=False), + build_norm_layer(self.norm_cfg, stem_channels // 2)[1], + nn.ReLU(inplace=True), + build_conv_layer( + self.conv_cfg, + stem_channels // 2, + stem_channels, + kernel_size=3, + stride=1, + padding=1, + bias=False), + build_norm_layer(self.norm_cfg, stem_channels)[1], + nn.ReLU(inplace=True)) + else: + self.conv1 = build_conv_layer( + self.conv_cfg, + in_channels, + stem_channels, + kernel_size=7, + stride=2, + padding=3, + bias=False) + self.norm1_name, norm1 = build_norm_layer( + self.norm_cfg, stem_channels, postfix=1) + self.add_module(self.norm1_name, norm1) + self.relu = nn.ReLU(inplace=True) + self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1) + + def _freeze_stages(self): + """Freeze stages param and norm stats.""" + if self.frozen_stages >= 0: + if self.deep_stem: + self.stem.eval() + for param in self.stem.parameters(): + param.requires_grad = False + else: + self.norm1.eval() + for m in [self.conv1, self.norm1]: + for param in m.parameters(): + param.requires_grad = False + + for i in range(1, self.frozen_stages + 1): + m = getattr(self, f'layer{i}') + m.eval() + for param in m.parameters(): + param.requires_grad = False + + def forward(self, x): + """Forward function.""" + if self.deep_stem: + x = self.stem(x) + else: + x = self.conv1(x) + x = self.norm1(x) + x = self.relu(x) + x = self.maxpool(x) + outs = [] + for i, layer_name in enumerate(self.res_layers): + res_layer = getattr(self, layer_name) + x = res_layer(x) + if i in self.out_indices: + outs.append(x) + return tuple(outs) + + def train(self, mode=True): + """Convert the model into training mode while keep normalization layer + freezed.""" + super(ResNet, self).train(mode) + self._freeze_stages() + if mode and self.norm_eval: + for m in self.modules(): + # trick: eval have effect on BatchNorm only + if isinstance(m, _BatchNorm): + m.eval() + + +@BACKBONES.register_module() +class ResNetV1c(ResNet): + """ResNetV1c variant described in [1]_. + + Compared with default ResNet(ResNetV1b), ResNetV1c replaces the 7x7 conv in + the input stem with three 3x3 convs. For more details please refer to `Bag + of Tricks for Image Classification with Convolutional Neural Networks + `_. + """ + + def __init__(self, **kwargs): + super(ResNetV1c, self).__init__( + deep_stem=True, avg_down=False, **kwargs) + + +@BACKBONES.register_module() +class ResNetV1d(ResNet): + """ResNetV1d variant described in [1]_. + + Compared with default ResNet(ResNetV1b), ResNetV1d replaces the 7x7 conv in + the input stem with three 3x3 convs. And in the downsampling block, a 2x2 + avg_pool with stride 2 is added before conv, whose stride is changed to 1. + """ + + def __init__(self, **kwargs): + super(ResNetV1d, self).__init__( + deep_stem=True, avg_down=True, **kwargs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/builder.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/builder.py new file mode 100644 index 0000000000000000000000000000000000000000..5e18e4e6430a80434d7cd2a3eed55ea343fccab6 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/builder.py @@ -0,0 +1,49 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +from mmcv.cnn import MODELS as MMCV_MODELS +from mmcv.cnn.bricks.registry import ATTENTION as MMCV_ATTENTION +from mmcv.utils import Registry + +MODELS = Registry('models', parent=MMCV_MODELS) +ATTENTION = Registry('attention', parent=MMCV_ATTENTION) + +BACKBONES = MODELS +NECKS = MODELS +HEADS = MODELS +LOSSES = MODELS +SEGMENTORS = MODELS + + +def build_backbone(cfg): + """Build backbone.""" + return BACKBONES.build(cfg) + + +def build_neck(cfg): + """Build neck.""" + return NECKS.build(cfg) + + +def build_head(cfg): + """Build head.""" + return HEADS.build(cfg) + + +def build_loss(cfg): + """Build loss.""" + return LOSSES.build(cfg) + + +def build_segmentor(cfg, train_cfg=None, test_cfg=None): + """Build segmentor.""" + if train_cfg is not None or test_cfg is not None: + warnings.warn( + 'train_cfg and test_cfg is deprecated, ' + 'please specify them in model', UserWarning) + assert cfg.get('train_cfg') is None or train_cfg is None, \ + 'train_cfg specified in both outer field and model field ' + assert cfg.get('test_cfg') is None or test_cfg is None, \ + 'test_cfg specified in both outer field and model field ' + return SEGMENTORS.build( + cfg, default_args=dict(train_cfg=train_cfg, test_cfg=test_cfg)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..a7b0b2d685ec4c494e714fc64d4e17c9310a2b7a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/__init__.py @@ -0,0 +1,4 @@ +# Copyright (c) OpenMMLab. All rights reserved. + +from .decode_head import BaseDecodeHead +from .attunet_head import ATTUNetHead diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/attunet_head.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/attunet_head.py new file mode 100644 index 0000000000000000000000000000000000000000..86ed0f56b4f2790f8e2383cd15332dddf730cba5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/attunet_head.py @@ -0,0 +1,124 @@ +import torch +import torch.nn as nn +import torch.nn.functional as F + +from ..builder import HEADS +from .decode_head import BaseDecodeHead + +def conv_bn_relu(in_channels, out_channels, kernel_size=3, padding=1, stride=1): + return nn.Sequential(nn.Conv2d(in_channels, + out_channels, + kernel_size=kernel_size, + padding=padding, + stride=stride), + nn.BatchNorm2d(out_channels), + nn.ReLU(inplace=True)) + +class AttBlock(nn.Module): + def __init__(self, F_g, F_l, F_int): + super(AttBlock, self).__init__() + self.W_g = nn.Sequential( + nn.Conv2d(in_channels=F_g, + out_channels=F_int, + kernel_size=1, + stride=1, + padding=0, + bias=True), + nn.BatchNorm2d(F_int) + ) + + self.W_x = nn.Sequential( + nn.Conv2d(in_channels=F_l, + out_channels=F_int, + kernel_size=1, + stride=1, + padding=0, + bias=True), + nn.BatchNorm2d(F_int) + ) + + self.psi = nn.Sequential( + nn.Conv2d(in_channels=F_int, + out_channels=1, + kernel_size=1, + stride=1, + padding=0, + bias=True), + nn.BatchNorm2d(1), + nn.Sigmoid() + ) + + self.relu = nn.ReLU(inplace=True) + + def forward(self, g, x): + g1 = self.W_g(g) + x1 = self.W_x(x) + psi = self.relu(g1 + x1) + psi = self.psi(psi) + + return x * psi + + +class DecBlock(nn.Module): + def __init__( + self, + in_channels, + skip_channels, + out_channels + ): + super().__init__() + self.conv1 = conv_bn_relu(in_channels=in_channels + skip_channels, + out_channels=out_channels) + + self.conv2 = conv_bn_relu(in_channels=out_channels, + out_channels=out_channels) + + self.up = nn.Upsample(scale_factor=2, + mode='bilinear', + align_corners=True) + + self.att = AttBlock(F_g=in_channels, F_l=skip_channels, F_int=in_channels) + + def forward(self, x, skip=None): + x = self.up(x) + if skip is not None: + if hasattr(self, "att"): + skip = self.att(g=x, x=skip) + x = torch.cat([x, skip], dim=1) + x = self.conv1(x) + x = self.conv2(x) + return x + + +@HEADS.register_module() +class ATTUNetHead(BaseDecodeHead): + def __init__(self, **kwargs): + super(ATTUNetHead, self).__init__(**kwargs) + self.decoders = nn.ModuleList() + in_channels = self.in_channels[::-1] + skip_channels = in_channels[1:] + for in_c, skip_c in zip(in_channels, skip_channels): + self.decoders.append(DecBlock(in_c, skip_c, skip_c)) + + def forward(self, features): + features = features[::-1] + x = features[0] + skips = features[1:] + + for i, layer in enumerate(self.decoders): + if i < len(skips): + x = layer(x, skips[i]) + else: + x = layer(x) + + output = self.cls_seg(x) + + return output + + def init_weights(self): + for m in self.modules(): + if isinstance(m, nn.Conv2d): + nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu') + elif isinstance(m, nn.BatchNorm2d): + nn.init.constant_(m.weight, 1) + nn.init.constant_(m.bias, 0) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/decode_head.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/decode_head.py new file mode 100644 index 0000000000000000000000000000000000000000..098e1b4e49f18eb613da6f0389445ddcdf2da3f5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/decode_heads/decode_head.py @@ -0,0 +1,266 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from abc import ABCMeta, abstractmethod + +import torch +import torch.nn as nn +from mmcv.runner import BaseModule, auto_fp16, force_fp32 + +from mmseg.core import build_pixel_sampler +from mmseg.ops import resize +from ..builder import build_loss +from ..losses import accuracy + + +class BaseDecodeHead(BaseModule, metaclass=ABCMeta): + """Base class for BaseDecodeHead. + + Args: + in_channels (int|Sequence[int]): Input channels. + channels (int): Channels after modules, before conv_seg. + num_classes (int): Number of classes. + dropout_ratio (float): Ratio of dropout layer. Default: 0.1. + conv_cfg (dict|None): Config of conv layers. Default: None. + norm_cfg (dict|None): Config of norm layers. Default: None. + act_cfg (dict): Config of activation layers. + Default: dict(type='ReLU') + in_index (int|Sequence[int]): Input feature index. Default: -1 + input_transform (str|None): Transformation type of input features. + Options: 'resize_concat', 'multiple_select', None. + 'resize_concat': Multiple feature maps will be resize to the + same size as first one and than concat together. + Usually used in FCN head of HRNet. + 'multiple_select': Multiple feature maps will be bundle into + a list and passed into decode head. + None: Only one select feature map is allowed. + Default: None. + loss_decode (dict | Sequence[dict]): Config of decode loss. + The `loss_name` is property of corresponding loss function which + could be shown in training log. If you want this loss + item to be included into the backward graph, `loss_` must be the + prefix of the name. Defaults to 'loss_ce'. + e.g. dict(type='CrossEntropyLoss'), + [dict(type='CrossEntropyLoss', loss_name='loss_ce'), + dict(type='DiceLoss', loss_name='loss_dice')] + Default: dict(type='CrossEntropyLoss'). + ignore_index (int | None): The label index to be ignored. When using + masked BCE loss, ignore_index should be set to None. Default: 255. + sampler (dict|None): The config of segmentation map sampler. + Default: None. + align_corners (bool): align_corners argument of F.interpolate. + Default: False. + init_cfg (dict or list[dict], optional): Initialization config dict. + """ + + def __init__(self, + in_channels, + channels, + *, + num_classes, + dropout_ratio=0.1, + conv_cfg=None, + norm_cfg=None, + act_cfg=dict(type='ReLU'), + in_index=-1, + input_transform=None, + loss_decode=dict( + type='CrossEntropyLoss', + use_sigmoid=False, + loss_weight=1.0), + ignore_index=255, + sampler=None, + align_corners=False, + init_cfg=dict( + type='Normal', std=0.01, override=dict(name='conv_seg'))): + super(BaseDecodeHead, self).__init__(init_cfg) + self._init_inputs(in_channels, in_index, input_transform) + self.channels = channels + self.num_classes = num_classes + self.dropout_ratio = dropout_ratio + self.conv_cfg = conv_cfg + self.norm_cfg = norm_cfg + self.act_cfg = act_cfg + self.in_index = in_index + + self.ignore_index = ignore_index + self.align_corners = align_corners + + if isinstance(loss_decode, dict): + self.loss_decode = build_loss(loss_decode) + elif isinstance(loss_decode, (list, tuple)): + self.loss_decode = nn.ModuleList() + for loss in loss_decode: + self.loss_decode.append(build_loss(loss)) + else: + raise TypeError(f'loss_decode must be a dict or sequence of dict,\ + but got {type(loss_decode)}') + + if sampler is not None: + self.sampler = build_pixel_sampler(sampler, context=self) + else: + self.sampler = None + + self.conv_seg = nn.Conv2d(channels, num_classes, kernel_size=1) + if dropout_ratio > 0: + self.dropout = nn.Dropout2d(dropout_ratio) + else: + self.dropout = None + self.fp16_enabled = False + + def extra_repr(self): + """Extra repr.""" + s = f'input_transform={self.input_transform}, ' \ + f'ignore_index={self.ignore_index}, ' \ + f'align_corners={self.align_corners}' + return s + + def _init_inputs(self, in_channels, in_index, input_transform): + """Check and initialize input transforms. + + The in_channels, in_index and input_transform must match. + Specifically, when input_transform is None, only single feature map + will be selected. So in_channels and in_index must be of type int. + When input_transform + + Args: + in_channels (int|Sequence[int]): Input channels. + in_index (int|Sequence[int]): Input feature index. + input_transform (str|None): Transformation type of input features. + Options: 'resize_concat', 'multiple_select', None. + 'resize_concat': Multiple feature maps will be resize to the + same size as first one and than concat together. + Usually used in FCN head of HRNet. + 'multiple_select': Multiple feature maps will be bundle into + a list and passed into decode head. + None: Only one select feature map is allowed. + """ + + if input_transform is not None: + assert input_transform in ['resize_concat', 'multiple_select'] + self.input_transform = input_transform + self.in_index = in_index + if input_transform is not None: + assert isinstance(in_channels, (list, tuple)) + assert isinstance(in_index, (list, tuple)) + assert len(in_channels) == len(in_index) + if input_transform == 'resize_concat': + self.in_channels = sum(in_channels) + else: + self.in_channels = in_channels + else: + assert isinstance(in_channels, (int, list, tuple)) + assert isinstance(in_index, int) + self.in_channels = in_channels + + def _transform_inputs(self, inputs): + """Transform inputs for decoder. + + Args: + inputs (list[Tensor]): List of multi-level img features. + + Returns: + Tensor: The transformed inputs + """ + + if self.input_transform == 'resize_concat': + inputs = [inputs[i] for i in self.in_index] + upsampled_inputs = [ + resize( + input=x, + size=inputs[0].shape[2:], + mode='bilinear', + align_corners=self.align_corners) for x in inputs + ] + inputs = torch.cat(upsampled_inputs, dim=1) + elif self.input_transform == 'multiple_select': + inputs = [inputs[i] for i in self.in_index] + else: + inputs = inputs[self.in_index] + + return inputs + + @auto_fp16() + @abstractmethod + def forward(self, inputs): + """Placeholder of forward function.""" + pass + + def forward_train(self, inputs, img_metas, gt_semantic_seg, train_cfg): + """Forward function for training. + Args: + inputs (list[Tensor]): List of multi-level img features. + img_metas (list[dict]): List of image info dict where each dict + has: 'img_shape', 'scale_factor', 'flip', and may also contain + 'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'. + For details on the values of these keys see + `mmseg/datasets/pipelines/formatting.py:Collect`. + gt_semantic_seg (Tensor): Semantic segmentation masks + used if the architecture supports semantic segmentation task. + train_cfg (dict): The training config. + + Returns: + dict[str, Tensor]: a dictionary of loss components + """ + seg_logits = self.forward(inputs) + losses = self.losses(seg_logits, gt_semantic_seg) + return losses + + def forward_test(self, inputs, img_metas, test_cfg): + """Forward function for testing. + + Args: + inputs (list[Tensor]): List of multi-level img features. + img_metas (list[dict]): List of image info dict where each dict + has: 'img_shape', 'scale_factor', 'flip', and may also contain + 'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'. + For details on the values of these keys see + `mmseg/datasets/pipelines/formatting.py:Collect`. + test_cfg (dict): The testing config. + + Returns: + Tensor: Output segmentation map. + """ + return self.forward(inputs) + + def cls_seg(self, feat): + """Classify each pixel.""" + if self.dropout is not None: + feat = self.dropout(feat) + output = self.conv_seg(feat) + return output + + @force_fp32(apply_to=('seg_logit', )) + def losses(self, seg_logit, seg_label): + """Compute segmentation loss.""" + loss = dict() + seg_logit = resize( + input=seg_logit, + size=seg_label.shape[2:], + mode='bilinear', + align_corners=self.align_corners) + if self.sampler is not None: + seg_weight = self.sampler.sample(seg_logit, seg_label) + else: + seg_weight = None + seg_label = seg_label.squeeze(1) + + if not isinstance(self.loss_decode, nn.ModuleList): + losses_decode = [self.loss_decode] + else: + losses_decode = self.loss_decode + for loss_decode in losses_decode: + if loss_decode.loss_name not in loss: + loss[loss_decode.loss_name] = loss_decode( + seg_logit, + seg_label, + weight=seg_weight, + ignore_index=self.ignore_index) + else: + loss[loss_decode.loss_name] += loss_decode( + seg_logit, + seg_label, + weight=seg_weight, + ignore_index=self.ignore_index) + + loss['acc_seg'] = accuracy( + seg_logit, seg_label, ignore_index=self.ignore_index) + return loss diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..fbc5b2d1b9d4425f3a1938cbf8aa3b50a1a41d31 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/__init__.py @@ -0,0 +1,15 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .accuracy import Accuracy, accuracy +from .cross_entropy_loss import (CrossEntropyLoss, binary_cross_entropy, + cross_entropy, mask_cross_entropy) +from .dice_loss import DiceLoss +from .focal_loss import FocalLoss +from .lovasz_loss import LovaszLoss +from .utils import reduce_loss, weight_reduce_loss, weighted_loss + +__all__ = [ + 'accuracy', 'Accuracy', 'cross_entropy', 'binary_cross_entropy', + 'mask_cross_entropy', 'CrossEntropyLoss', 'reduce_loss', + 'weight_reduce_loss', 'weighted_loss', 'LovaszLoss', 'DiceLoss', + 'FocalLoss' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/accuracy.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/accuracy.py new file mode 100644 index 0000000000000000000000000000000000000000..1d9e2d7701088adadd5b6bb71c718c986b87a066 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/accuracy.py @@ -0,0 +1,92 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn + + +def accuracy(pred, target, topk=1, thresh=None, ignore_index=None): + """Calculate accuracy according to the prediction and target. + + Args: + pred (torch.Tensor): The model prediction, shape (N, num_class, ...) + target (torch.Tensor): The target of each prediction, shape (N, , ...) + ignore_index (int | None): The label index to be ignored. Default: None + topk (int | tuple[int], optional): If the predictions in ``topk`` + matches the target, the predictions will be regarded as + correct ones. Defaults to 1. + thresh (float, optional): If not None, predictions with scores under + this threshold are considered incorrect. Default to None. + + Returns: + float | tuple[float]: If the input ``topk`` is a single integer, + the function will return a single float as accuracy. If + ``topk`` is a tuple containing multiple integers, the + function will return a tuple containing accuracies of + each ``topk`` number. + """ + assert isinstance(topk, (int, tuple)) + if isinstance(topk, int): + topk = (topk, ) + return_single = True + else: + return_single = False + + maxk = max(topk) + if pred.size(0) == 0: + accu = [pred.new_tensor(0.) for i in range(len(topk))] + return accu[0] if return_single else accu + assert pred.ndim == target.ndim + 1 + assert pred.size(0) == target.size(0) + assert maxk <= pred.size(1), \ + f'maxk {maxk} exceeds pred dimension {pred.size(1)}' + pred_value, pred_label = pred.topk(maxk, dim=1) + # transpose to shape (maxk, N, ...) + pred_label = pred_label.transpose(0, 1) + correct = pred_label.eq(target.unsqueeze(0).expand_as(pred_label)) + if thresh is not None: + # Only prediction values larger than thresh are counted as correct + correct = correct & (pred_value > thresh).t() + if ignore_index is not None: + correct = correct[:, target != ignore_index] + res = [] + eps = torch.finfo(torch.float32).eps + for k in topk: + # Avoid causing ZeroDivisionError when all pixels + # of an image are ignored + correct_k = correct[:k].reshape(-1).float().sum(0, keepdim=True) + eps + if ignore_index is not None: + total_num = target[target != ignore_index].numel() + eps + else: + total_num = target.numel() + eps + res.append(correct_k.mul_(100.0 / total_num)) + return res[0] if return_single else res + + +class Accuracy(nn.Module): + """Accuracy calculation module.""" + + def __init__(self, topk=(1, ), thresh=None, ignore_index=None): + """Module to calculate the accuracy. + + Args: + topk (tuple, optional): The criterion used to calculate the + accuracy. Defaults to (1,). + thresh (float, optional): If not None, predictions with scores + under this threshold are considered incorrect. Default to None. + """ + super().__init__() + self.topk = topk + self.thresh = thresh + self.ignore_index = ignore_index + + def forward(self, pred, target): + """Forward function to calculate accuracy. + + Args: + pred (torch.Tensor): Prediction of models. + target (torch.Tensor): Target for each prediction. + + Returns: + tuple[float]: The accuracies under different topk criterions. + """ + return accuracy(pred, target, self.topk, self.thresh, + self.ignore_index) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/cross_entropy_loss.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/cross_entropy_loss.py new file mode 100644 index 0000000000000000000000000000000000000000..623fd58dbc7d909962f00d85517720ec732c6ff2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/cross_entropy_loss.py @@ -0,0 +1,296 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch +import torch.nn as nn +import torch.nn.functional as F + +from ..builder import LOSSES +from .utils import get_class_weight, weight_reduce_loss + + +def cross_entropy(pred, + label, + weight=None, + class_weight=None, + reduction='mean', + avg_factor=None, + ignore_index=-100, + avg_non_ignore=False): + """cross_entropy. The wrapper function for :func:`F.cross_entropy` + + Args: + pred (torch.Tensor): The prediction with shape (N, 1). + label (torch.Tensor): The learning label of the prediction. + weight (torch.Tensor, optional): Sample-wise loss weight. + Default: None. + class_weight (list[float], optional): The weight for each class. + Default: None. + reduction (str, optional): The method used to reduce the loss. + Options are 'none', 'mean' and 'sum'. Default: 'mean'. + avg_factor (int, optional): Average factor that is used to average + the loss. Default: None. + ignore_index (int): Specifies a target value that is ignored and + does not contribute to the input gradients. When + ``avg_non_ignore `` is ``True``, and the ``reduction`` is + ``''mean''``, the loss is averaged over non-ignored targets. + Defaults: -100. + avg_non_ignore (bool): The flag decides to whether the loss is + only averaged over non-ignored targets. Default: False. + `New in version 0.23.0.` + """ + + # class_weight is a manual rescaling weight given to each class. + # If given, has to be a Tensor of size C element-wise losses + loss = F.cross_entropy( + pred, + label, + weight=class_weight, + reduction='none', + ignore_index=ignore_index) + + # apply weights and do the reduction + # average loss over non-ignored elements + # pytorch's official cross_entropy average loss over non-ignored elements + # refer to https://github.com/pytorch/pytorch/blob/56b43f4fec1f76953f15a627694d4bba34588969/torch/nn/functional.py#L2660 # noqa + if (avg_factor is None) and avg_non_ignore and reduction == 'mean': + avg_factor = label.numel() - (label == ignore_index).sum().item() + if weight is not None: + weight = weight.float() + loss = weight_reduce_loss( + loss, weight=weight, reduction=reduction, avg_factor=avg_factor) + + return loss + + +def _expand_onehot_labels(labels, label_weights, target_shape, ignore_index): + """Expand onehot labels to match the size of prediction.""" + bin_labels = labels.new_zeros(target_shape) + valid_mask = (labels >= 0) & (labels != ignore_index) + inds = torch.nonzero(valid_mask, as_tuple=True) + + if inds[0].numel() > 0: + if labels.dim() == 3: + bin_labels[inds[0], labels[valid_mask], inds[1], inds[2]] = 1 + else: + bin_labels[inds[0], labels[valid_mask]] = 1 + + valid_mask = valid_mask.unsqueeze(1).expand(target_shape).float() + + if label_weights is None: + bin_label_weights = valid_mask + else: + bin_label_weights = label_weights.unsqueeze(1).expand(target_shape) + bin_label_weights = bin_label_weights * valid_mask + + return bin_labels, bin_label_weights, valid_mask + + +def binary_cross_entropy(pred, + label, + weight=None, + reduction='mean', + avg_factor=None, + class_weight=None, + ignore_index=-100, + avg_non_ignore=False, + **kwargs): + """Calculate the binary CrossEntropy loss. + + Args: + pred (torch.Tensor): The prediction with shape (N, 1). + label (torch.Tensor): The learning label of the prediction. + Note: In bce loss, label < 0 is invalid. + weight (torch.Tensor, optional): Sample-wise loss weight. + reduction (str, optional): The method used to reduce the loss. + Options are "none", "mean" and "sum". + avg_factor (int, optional): Average factor that is used to average + the loss. Defaults to None. + class_weight (list[float], optional): The weight for each class. + ignore_index (int): The label index to be ignored. Default: -100. + avg_non_ignore (bool): The flag decides to whether the loss is + only averaged over non-ignored targets. Default: False. + `New in version 0.23.0.` + + Returns: + torch.Tensor: The calculated loss + """ + if pred.size(1) == 1: + # For binary class segmentation, the shape of pred is + # [N, 1, H, W] and that of label is [N, H, W]. + # As the ignore_index often set as 255, so the + # binary class label check should mask out + # ignore_index + assert label[label != ignore_index].max() <= 1, \ + 'For pred with shape [N, 1, H, W], its label must have at ' \ + 'most 2 classes' + pred = pred.squeeze() + if pred.dim() != label.dim(): + assert (pred.dim() == 2 and label.dim() == 1) or ( + pred.dim() == 4 and label.dim() == 3), \ + 'Only pred shape [N, C], label shape [N] or pred shape [N, C, ' \ + 'H, W], label shape [N, H, W] are supported' + # `weight` returned from `_expand_onehot_labels` + # has been treated for valid (non-ignore) pixels + label, weight, valid_mask = _expand_onehot_labels( + label, weight, pred.shape, ignore_index) + else: + # should mask out the ignored elements + valid_mask = ((label >= 0) & (label != ignore_index)).float() + if weight is not None: + weight = weight * valid_mask + else: + weight = valid_mask + # average loss over non-ignored and valid elements + if reduction == 'mean' and avg_factor is None and avg_non_ignore: + avg_factor = valid_mask.sum().item() + + loss = F.binary_cross_entropy_with_logits( + pred, label.float(), pos_weight=class_weight, reduction='none') + # do the reduction for the weighted loss + loss = weight_reduce_loss( + loss, weight, reduction=reduction, avg_factor=avg_factor) + + return loss + + +def mask_cross_entropy(pred, + target, + label, + reduction='mean', + avg_factor=None, + class_weight=None, + ignore_index=None, + **kwargs): + """Calculate the CrossEntropy loss for masks. + + Args: + pred (torch.Tensor): The prediction with shape (N, C), C is the number + of classes. + target (torch.Tensor): The learning label of the prediction. + label (torch.Tensor): ``label`` indicates the class label of the mask' + corresponding object. This will be used to select the mask in the + of the class which the object belongs to when the mask prediction + if not class-agnostic. + reduction (str, optional): The method used to reduce the loss. + Options are "none", "mean" and "sum". + avg_factor (int, optional): Average factor that is used to average + the loss. Defaults to None. + class_weight (list[float], optional): The weight for each class. + ignore_index (None): Placeholder, to be consistent with other loss. + Default: None. + + Returns: + torch.Tensor: The calculated loss + """ + assert ignore_index is None, 'BCE loss does not support ignore_index' + # TODO: handle these two reserved arguments + assert reduction == 'mean' and avg_factor is None + num_rois = pred.size()[0] + inds = torch.arange(0, num_rois, dtype=torch.long, device=pred.device) + pred_slice = pred[inds, label].squeeze(1) + return F.binary_cross_entropy_with_logits( + pred_slice, target, weight=class_weight, reduction='mean')[None] + + +@LOSSES.register_module() +class CrossEntropyLoss(nn.Module): + """CrossEntropyLoss. + + Args: + use_sigmoid (bool, optional): Whether the prediction uses sigmoid + of softmax. Defaults to False. + use_mask (bool, optional): Whether to use mask cross entropy loss. + Defaults to False. + reduction (str, optional): . Defaults to 'mean'. + Options are "none", "mean" and "sum". + class_weight (list[float] | str, optional): Weight of each class. If in + str format, read them from a file. Defaults to None. + loss_weight (float, optional): Weight of the loss. Defaults to 1.0. + loss_name (str, optional): Name of the loss item. If you want this loss + item to be included into the backward graph, `loss_` must be the + prefix of the name. Defaults to 'loss_ce'. + avg_non_ignore (bool): The flag decides to whether the loss is + only averaged over non-ignored targets. Default: False. + `New in version 0.23.0.` + """ + + def __init__(self, + use_sigmoid=False, + use_mask=False, + reduction='mean', + class_weight=None, + loss_weight=1.0, + loss_name='loss_ce', + avg_non_ignore=False): + super(CrossEntropyLoss, self).__init__() + assert (use_sigmoid is False) or (use_mask is False) + self.use_sigmoid = use_sigmoid + self.use_mask = use_mask + self.reduction = reduction + self.loss_weight = loss_weight + self.class_weight = get_class_weight(class_weight) + self.avg_non_ignore = avg_non_ignore + if not self.avg_non_ignore and self.reduction == 'mean': + warnings.warn( + 'Default ``avg_non_ignore`` is False, if you would like to ' + 'ignore the certain label and average loss over non-ignore ' + 'labels, which is the same with PyTorch official ' + 'cross_entropy, set ``avg_non_ignore=True``.') + + if self.use_sigmoid: + self.cls_criterion = binary_cross_entropy + elif self.use_mask: + self.cls_criterion = mask_cross_entropy + else: + self.cls_criterion = cross_entropy + self._loss_name = loss_name + + def extra_repr(self): + """Extra repr.""" + s = f'avg_non_ignore={self.avg_non_ignore}' + return s + + def forward(self, + cls_score, + label, + weight=None, + avg_factor=None, + reduction_override=None, + ignore_index=-100, + **kwargs): + """Forward function.""" + assert reduction_override in (None, 'none', 'mean', 'sum') + reduction = ( + reduction_override if reduction_override else self.reduction) + if self.class_weight is not None: + class_weight = cls_score.new_tensor(self.class_weight) + else: + class_weight = None + # Note: for BCE loss, label < 0 is invalid. + loss_cls = self.loss_weight * self.cls_criterion( + cls_score, + label, + weight, + class_weight=class_weight, + reduction=reduction, + avg_factor=avg_factor, + avg_non_ignore=self.avg_non_ignore, + ignore_index=ignore_index, + **kwargs) + return loss_cls + + @property + def loss_name(self): + """Loss Name. + + This function must be implemented and will return the name of this + loss function. This name will be used to combine different loss items + by simple sum operation. In addition, if you want this loss item to be + included into the backward graph, `loss_` must be the prefix of the + name. + + Returns: + str: The name of this loss item. + """ + return self._loss_name diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/dice_loss.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/dice_loss.py new file mode 100644 index 0000000000000000000000000000000000000000..79a3abfc2f633042b065987975a94edc64be06b1 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/dice_loss.py @@ -0,0 +1,137 @@ +# Copyright (c) OpenMMLab. All rights reserved. +"""Modified from https://github.com/LikeLy-Journey/SegmenTron/blob/master/ +segmentron/solver/loss.py (Apache-2.0 License)""" +import torch +import torch.nn as nn +import torch.nn.functional as F + +from ..builder import LOSSES +from .utils import get_class_weight, weighted_loss + + +@weighted_loss +def dice_loss(pred, + target, + valid_mask, + smooth=1, + exponent=2, + class_weight=None, + ignore_index=255): + assert pred.shape[0] == target.shape[0] + total_loss = 0 + num_classes = pred.shape[1] + for i in range(num_classes): + if i != ignore_index: + dice_loss = binary_dice_loss( + pred[:, i], + target[..., i], + valid_mask=valid_mask, + smooth=smooth, + exponent=exponent) + if class_weight is not None: + dice_loss *= class_weight[i] + total_loss += dice_loss + return total_loss / num_classes + + +@weighted_loss +def binary_dice_loss(pred, target, valid_mask, smooth=1, exponent=2, **kwards): + assert pred.shape[0] == target.shape[0] + pred = pred.reshape(pred.shape[0], -1) + target = target.reshape(target.shape[0], -1) + valid_mask = valid_mask.reshape(valid_mask.shape[0], -1) + + num = torch.sum(torch.mul(pred, target) * valid_mask, dim=1) * 2 + smooth + den = torch.sum(pred.pow(exponent) + target.pow(exponent), dim=1) + smooth + + return 1 - num / den + + +@LOSSES.register_module() +class DiceLoss(nn.Module): + """DiceLoss. + + This loss is proposed in `V-Net: Fully Convolutional Neural Networks for + Volumetric Medical Image Segmentation `_. + + Args: + smooth (float): A float number to smooth loss, and avoid NaN error. + Default: 1 + exponent (float): An float number to calculate denominator + value: \\sum{x^exponent} + \\sum{y^exponent}. Default: 2. + reduction (str, optional): The method used to reduce the loss. Options + are "none", "mean" and "sum". This parameter only works when + per_image is True. Default: 'mean'. + class_weight (list[float] | str, optional): Weight of each class. If in + str format, read them from a file. Defaults to None. + loss_weight (float, optional): Weight of the loss. Default to 1.0. + ignore_index (int | None): The label index to be ignored. Default: 255. + loss_name (str, optional): Name of the loss item. If you want this loss + item to be included into the backward graph, `loss_` must be the + prefix of the name. Defaults to 'loss_dice'. + """ + + def __init__(self, + smooth=1, + exponent=2, + reduction='mean', + class_weight=None, + loss_weight=1.0, + ignore_index=255, + loss_name='loss_dice', + **kwards): + super(DiceLoss, self).__init__() + self.smooth = smooth + self.exponent = exponent + self.reduction = reduction + self.class_weight = get_class_weight(class_weight) + self.loss_weight = loss_weight + self.ignore_index = ignore_index + self._loss_name = loss_name + + def forward(self, + pred, + target, + avg_factor=None, + reduction_override=None, + **kwards): + assert reduction_override in (None, 'none', 'mean', 'sum') + reduction = ( + reduction_override if reduction_override else self.reduction) + if self.class_weight is not None: + class_weight = pred.new_tensor(self.class_weight) + else: + class_weight = None + + pred = F.softmax(pred, dim=1) + num_classes = pred.shape[1] + one_hot_target = F.one_hot( + torch.clamp(target.long(), 0, num_classes - 1), + num_classes=num_classes) + valid_mask = (target != self.ignore_index).long() + + loss = self.loss_weight * dice_loss( + pred, + one_hot_target, + valid_mask=valid_mask, + reduction=reduction, + avg_factor=avg_factor, + smooth=self.smooth, + exponent=self.exponent, + class_weight=class_weight, + ignore_index=self.ignore_index) + return loss + + @property + def loss_name(self): + """Loss Name. + + This function must be implemented and will return the name of this + loss function. This name will be used to combine different loss items + by simple sum operation. In addition, if you want this loss item to be + included into the backward graph, `loss_` must be the prefix of the + name. + Returns: + str: The name of this loss item. + """ + return self._loss_name diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/focal_loss.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/focal_loss.py new file mode 100644 index 0000000000000000000000000000000000000000..af1c711dfbce65907e77d37e7a8f0aa9bd2e5e3a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/focal_loss.py @@ -0,0 +1,327 @@ +# Copyright (c) OpenMMLab. All rights reserved. +# Modified from https://github.com/open-mmlab/mmdetection +import torch +import torch.nn as nn +import torch.nn.functional as F +from mmcv.ops import sigmoid_focal_loss as _sigmoid_focal_loss + +from ..builder import LOSSES +from .utils import weight_reduce_loss + + +# This method is used when cuda is not available +def py_sigmoid_focal_loss(pred, + target, + one_hot_target=None, + weight=None, + gamma=2.0, + alpha=0.5, + class_weight=None, + valid_mask=None, + reduction='mean', + avg_factor=None): + """PyTorch version of `Focal Loss `_. + + Args: + pred (torch.Tensor): The prediction with shape (N, C), C is the + number of classes + target (torch.Tensor): The learning label of the prediction with + shape (N, C) + one_hot_target (None): Placeholder. It should be None. + weight (torch.Tensor, optional): Sample-wise loss weight. + gamma (float, optional): The gamma for calculating the modulating + factor. Defaults to 2.0. + alpha (float | list[float], optional): A balanced form for Focal Loss. + Defaults to 0.5. + class_weight (list[float], optional): Weight of each class. + Defaults to None. + valid_mask (torch.Tensor, optional): A mask uses 1 to mark the valid + samples and uses 0 to mark the ignored samples. Default: None. + reduction (str, optional): The method used to reduce the loss into + a scalar. Defaults to 'mean'. + avg_factor (int, optional): Average factor that is used to average + the loss. Defaults to None. + """ + if isinstance(alpha, list): + alpha = pred.new_tensor(alpha) + pred_sigmoid = pred.sigmoid() + target = target.type_as(pred) + one_minus_pt = (1 - pred_sigmoid) * target + pred_sigmoid * (1 - target) + focal_weight = (alpha * target + (1 - alpha) * + (1 - target)) * one_minus_pt.pow(gamma) + + loss = F.binary_cross_entropy_with_logits( + pred, target, reduction='none') * focal_weight + final_weight = torch.ones(1, pred.size(1)).type_as(loss) + if weight is not None: + if weight.shape != loss.shape and weight.size(0) == loss.size(0): + # For most cases, weight is of shape (N, ), + # which means it does not have the second axis num_class + weight = weight.view(-1, 1) + assert weight.dim() == loss.dim() + final_weight = final_weight * weight + if class_weight is not None: + final_weight = final_weight * pred.new_tensor(class_weight) + if valid_mask is not None: + final_weight = final_weight * valid_mask + loss = weight_reduce_loss(loss, final_weight, reduction, avg_factor) + return loss + + +def sigmoid_focal_loss(pred, + target, + one_hot_target, + weight=None, + gamma=2.0, + alpha=0.5, + class_weight=None, + valid_mask=None, + reduction='mean', + avg_factor=None): + r"""A warpper of cuda version `Focal Loss + `_. + Args: + pred (torch.Tensor): The prediction with shape (N, C), C is the number + of classes. + target (torch.Tensor): The learning label of the prediction. It's shape + should be (N, ) + one_hot_target (torch.Tensor): The learning label with shape (N, C) + weight (torch.Tensor, optional): Sample-wise loss weight. + gamma (float, optional): The gamma for calculating the modulating + factor. Defaults to 2.0. + alpha (float | list[float], optional): A balanced form for Focal Loss. + Defaults to 0.5. + class_weight (list[float], optional): Weight of each class. + Defaults to None. + valid_mask (torch.Tensor, optional): A mask uses 1 to mark the valid + samples and uses 0 to mark the ignored samples. Default: None. + reduction (str, optional): The method used to reduce the loss into + a scalar. Defaults to 'mean'. Options are "none", "mean" and "sum". + avg_factor (int, optional): Average factor that is used to average + the loss. Defaults to None. + """ + # Function.apply does not accept keyword arguments, so the decorator + # "weighted_loss" is not applicable + final_weight = torch.ones(1, pred.size(1)).type_as(pred) + if isinstance(alpha, list): + # _sigmoid_focal_loss doesn't accept alpha of list type. Therefore, if + # a list is given, we set the input alpha as 0.5. This means setting + # equal weight for foreground class and background class. By + # multiplying the loss by 2, the effect of setting alpha as 0.5 is + # undone. The alpha of type list is used to regulate the loss in the + # post-processing process. + loss = _sigmoid_focal_loss(pred.contiguous(), target.contiguous(), + gamma, 0.5, None, 'none') * 2 + alpha = pred.new_tensor(alpha) + final_weight = final_weight * ( + alpha * one_hot_target + (1 - alpha) * (1 - one_hot_target)) + else: + loss = _sigmoid_focal_loss(pred.contiguous(), target.contiguous(), + gamma, alpha, None, 'none') + if weight is not None: + if weight.shape != loss.shape and weight.size(0) == loss.size(0): + # For most cases, weight is of shape (N, ), + # which means it does not have the second axis num_class + weight = weight.view(-1, 1) + assert weight.dim() == loss.dim() + final_weight = final_weight * weight + if class_weight is not None: + final_weight = final_weight * pred.new_tensor(class_weight) + if valid_mask is not None: + final_weight = final_weight * valid_mask + loss = weight_reduce_loss(loss, final_weight, reduction, avg_factor) + return loss + + +@LOSSES.register_module() +class FocalLoss(nn.Module): + + def __init__(self, + use_sigmoid=True, + gamma=2.0, + alpha=0.5, + reduction='mean', + class_weight=None, + loss_weight=1.0, + loss_name='loss_focal'): + """`Focal Loss `_ + Args: + use_sigmoid (bool, optional): Whether to the prediction is + used for sigmoid or softmax. Defaults to True. + gamma (float, optional): The gamma for calculating the modulating + factor. Defaults to 2.0. + alpha (float | list[float], optional): A balanced form for Focal + Loss. Defaults to 0.5. When a list is provided, the length + of the list should be equal to the number of classes. + Please be careful that this parameter is not the + class-wise weight but the weight of a binary classification + problem. This binary classification problem regards the + pixels which belong to one class as the foreground + and the other pixels as the background, each element in + the list is the weight of the corresponding foreground class. + The value of alpha or each element of alpha should be a float + in the interval [0, 1]. If you want to specify the class-wise + weight, please use `class_weight` parameter. + reduction (str, optional): The method used to reduce the loss into + a scalar. Defaults to 'mean'. Options are "none", "mean" and + "sum". + class_weight (list[float], optional): Weight of each class. + Defaults to None. + loss_weight (float, optional): Weight of loss. Defaults to 1.0. + loss_name (str, optional): Name of the loss item. If you want this + loss item to be included into the backward graph, `loss_` must + be the prefix of the name. Defaults to 'loss_focal'. + """ + super(FocalLoss, self).__init__() + assert use_sigmoid is True, \ + 'AssertionError: Only sigmoid focal loss supported now.' + assert reduction in ('none', 'mean', 'sum'), \ + "AssertionError: reduction should be 'none', 'mean' or " \ + "'sum'" + assert isinstance(alpha, (float, list)), \ + 'AssertionError: alpha should be of type float' + assert isinstance(gamma, float), \ + 'AssertionError: gamma should be of type float' + assert isinstance(loss_weight, float), \ + 'AssertionError: loss_weight should be of type float' + assert isinstance(loss_name, str), \ + 'AssertionError: loss_name should be of type str' + assert isinstance(class_weight, list) or class_weight is None, \ + 'AssertionError: class_weight must be None or of type list' + self.use_sigmoid = use_sigmoid + self.gamma = gamma + self.alpha = alpha + self.reduction = reduction + self.class_weight = class_weight + self.loss_weight = loss_weight + self._loss_name = loss_name + + def forward(self, + pred, + target, + weight=None, + avg_factor=None, + reduction_override=None, + ignore_index=255, + **kwargs): + """Forward function. + + Args: + pred (torch.Tensor): The prediction with shape + (N, C) where C = number of classes, or + (N, C, d_1, d_2, ..., d_K) with K≥1 in the + case of K-dimensional loss. + target (torch.Tensor): The ground truth. If containing class + indices, shape (N) where each value is 0≤targets[i]≤C−1, + or (N, d_1, d_2, ..., d_K) with K≥1 in the case of + K-dimensional loss. If containing class probabilities, + same shape as the input. + weight (torch.Tensor, optional): The weight of loss for each + prediction. Defaults to None. + avg_factor (int, optional): Average factor that is used to + average the loss. Defaults to None. + reduction_override (str, optional): The reduction method used + to override the original reduction method of the loss. + Options are "none", "mean" and "sum". + ignore_index (int, optional): The label index to be ignored. + Default: 255 + Returns: + torch.Tensor: The calculated loss + """ + assert isinstance(ignore_index, int), \ + 'ignore_index must be of type int' + assert reduction_override in (None, 'none', 'mean', 'sum'), \ + "AssertionError: reduction should be 'none', 'mean' or " \ + "'sum'" + assert pred.shape == target.shape or \ + (pred.size(0) == target.size(0) and + pred.shape[2:] == target.shape[1:]), \ + "The shape of pred doesn't match the shape of target" + + original_shape = pred.shape + + # [B, C, d_1, d_2, ..., d_k] -> [C, B, d_1, d_2, ..., d_k] + pred = pred.transpose(0, 1) + # [C, B, d_1, d_2, ..., d_k] -> [C, N] + pred = pred.reshape(pred.size(0), -1) + # [C, N] -> [N, C] + pred = pred.transpose(0, 1).contiguous() + + if original_shape == target.shape: + # target with shape [B, C, d_1, d_2, ...] + # transform it's shape into [N, C] + # [B, C, d_1, d_2, ...] -> [C, B, d_1, d_2, ..., d_k] + target = target.transpose(0, 1) + # [C, B, d_1, d_2, ..., d_k] -> [C, N] + target = target.reshape(target.size(0), -1) + # [C, N] -> [N, C] + target = target.transpose(0, 1).contiguous() + else: + # target with shape [B, d_1, d_2, ...] + # transform it's shape into [N, ] + target = target.view(-1).contiguous() + valid_mask = (target != ignore_index).view(-1, 1) + # avoid raising error when using F.one_hot() + target = torch.where(target == ignore_index, target.new_tensor(0), + target) + + reduction = ( + reduction_override if reduction_override else self.reduction) + if self.use_sigmoid: + num_classes = pred.size(1) + if torch.cuda.is_available() and pred.is_cuda: + if target.dim() == 1: + one_hot_target = F.one_hot(target, num_classes=num_classes) + else: + one_hot_target = target + target = target.argmax(dim=1) + valid_mask = (target != ignore_index).view(-1, 1) + calculate_loss_func = sigmoid_focal_loss + else: + one_hot_target = None + if target.dim() == 1: + target = F.one_hot(target, num_classes=num_classes) + else: + valid_mask = (target.argmax(dim=1) != ignore_index).view( + -1, 1) + calculate_loss_func = py_sigmoid_focal_loss + + loss_cls = self.loss_weight * calculate_loss_func( + pred, + target, + one_hot_target, + weight, + gamma=self.gamma, + alpha=self.alpha, + class_weight=self.class_weight, + valid_mask=valid_mask, + reduction=reduction, + avg_factor=avg_factor) + + if reduction == 'none': + # [N, C] -> [C, N] + loss_cls = loss_cls.transpose(0, 1) + # [C, N] -> [C, B, d1, d2, ...] + # original_shape: [B, C, d1, d2, ...] + loss_cls = loss_cls.reshape(original_shape[1], + original_shape[0], + *original_shape[2:]) + # [C, B, d1, d2, ...] -> [B, C, d1, d2, ...] + loss_cls = loss_cls.transpose(0, 1).contiguous() + else: + raise NotImplementedError + return loss_cls + + @property + def loss_name(self): + """Loss Name. + + This function must be implemented and will return the name of this + loss function. This name will be used to combine different loss items + by simple sum operation. In addition, if you want this loss item to be + included into the backward graph, `loss_` must be the prefix of the + name. + Returns: + str: The name of this loss item. + """ + return self._loss_name diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/lovasz_loss.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/lovasz_loss.py new file mode 100644 index 0000000000000000000000000000000000000000..2bb0fad3931ea7d6140beca8b32a09379fbd7670 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/lovasz_loss.py @@ -0,0 +1,323 @@ +# Copyright (c) OpenMMLab. All rights reserved. +"""Modified from https://github.com/bermanmaxim/LovaszSoftmax/blob/master/pytor +ch/lovasz_losses.py Lovasz-Softmax and Jaccard hinge loss in PyTorch Maxim +Berman 2018 ESAT-PSI KU Leuven (MIT License)""" + +import mmcv +import torch +import torch.nn as nn +import torch.nn.functional as F + +from ..builder import LOSSES +from .utils import get_class_weight, weight_reduce_loss + + +def lovasz_grad(gt_sorted): + """Computes gradient of the Lovasz extension w.r.t sorted errors. + + See Alg. 1 in paper. + """ + p = len(gt_sorted) + gts = gt_sorted.sum() + intersection = gts - gt_sorted.float().cumsum(0) + union = gts + (1 - gt_sorted).float().cumsum(0) + jaccard = 1. - intersection / union + if p > 1: # cover 1-pixel case + jaccard[1:p] = jaccard[1:p] - jaccard[0:-1] + return jaccard + + +def flatten_binary_logits(logits, labels, ignore_index=None): + """Flattens predictions in the batch (binary case) Remove labels equal to + 'ignore_index'.""" + logits = logits.view(-1) + labels = labels.view(-1) + if ignore_index is None: + return logits, labels + valid = (labels != ignore_index) + vlogits = logits[valid] + vlabels = labels[valid] + return vlogits, vlabels + + +def flatten_probs(probs, labels, ignore_index=None): + """Flattens predictions in the batch.""" + if probs.dim() == 3: + # assumes output of a sigmoid layer + B, H, W = probs.size() + probs = probs.view(B, 1, H, W) + B, C, H, W = probs.size() + probs = probs.permute(0, 2, 3, 1).contiguous().view(-1, C) # B*H*W, C=P,C + labels = labels.view(-1) + if ignore_index is None: + return probs, labels + valid = (labels != ignore_index) + vprobs = probs[valid.nonzero().squeeze()] + vlabels = labels[valid] + return vprobs, vlabels + + +def lovasz_hinge_flat(logits, labels): + """Binary Lovasz hinge loss. + + Args: + logits (torch.Tensor): [P], logits at each prediction + (between -infty and +infty). + labels (torch.Tensor): [P], binary ground truth labels (0 or 1). + + Returns: + torch.Tensor: The calculated loss. + """ + if len(labels) == 0: + # only void pixels, the gradients should be 0 + return logits.sum() * 0. + signs = 2. * labels.float() - 1. + errors = (1. - logits * signs) + errors_sorted, perm = torch.sort(errors, dim=0, descending=True) + perm = perm.data + gt_sorted = labels[perm] + grad = lovasz_grad(gt_sorted) + loss = torch.dot(F.relu(errors_sorted), grad) + return loss + + +def lovasz_hinge(logits, + labels, + classes='present', + per_image=False, + class_weight=None, + reduction='mean', + avg_factor=None, + ignore_index=255): + """Binary Lovasz hinge loss. + + Args: + logits (torch.Tensor): [B, H, W], logits at each pixel + (between -infty and +infty). + labels (torch.Tensor): [B, H, W], binary ground truth masks (0 or 1). + classes (str | list[int], optional): Placeholder, to be consistent with + other loss. Default: None. + per_image (bool, optional): If per_image is True, compute the loss per + image instead of per batch. Default: False. + class_weight (list[float], optional): Placeholder, to be consistent + with other loss. Default: None. + reduction (str, optional): The method used to reduce the loss. Options + are "none", "mean" and "sum". This parameter only works when + per_image is True. Default: 'mean'. + avg_factor (int, optional): Average factor that is used to average + the loss. This parameter only works when per_image is True. + Default: None. + ignore_index (int | None): The label index to be ignored. Default: 255. + + Returns: + torch.Tensor: The calculated loss. + """ + if per_image: + loss = [ + lovasz_hinge_flat(*flatten_binary_logits( + logit.unsqueeze(0), label.unsqueeze(0), ignore_index)) + for logit, label in zip(logits, labels) + ] + loss = weight_reduce_loss( + torch.stack(loss), None, reduction, avg_factor) + else: + loss = lovasz_hinge_flat( + *flatten_binary_logits(logits, labels, ignore_index)) + return loss + + +def lovasz_softmax_flat(probs, labels, classes='present', class_weight=None): + """Multi-class Lovasz-Softmax loss. + + Args: + probs (torch.Tensor): [P, C], class probabilities at each prediction + (between 0 and 1). + labels (torch.Tensor): [P], ground truth labels (between 0 and C - 1). + classes (str | list[int], optional): Classes chosen to calculate loss. + 'all' for all classes, 'present' for classes present in labels, or + a list of classes to average. Default: 'present'. + class_weight (list[float], optional): The weight for each class. + Default: None. + + Returns: + torch.Tensor: The calculated loss. + """ + if probs.numel() == 0: + # only void pixels, the gradients should be 0 + return probs * 0. + C = probs.size(1) + losses = [] + class_to_sum = list(range(C)) if classes in ['all', 'present'] else classes + for c in class_to_sum: + fg = (labels == c).float() # foreground for class c + if (classes == 'present' and fg.sum() == 0): + continue + if C == 1: + if len(classes) > 1: + raise ValueError('Sigmoid output possible only with 1 class') + class_pred = probs[:, 0] + else: + class_pred = probs[:, c] + errors = (fg - class_pred).abs() + errors_sorted, perm = torch.sort(errors, 0, descending=True) + perm = perm.data + fg_sorted = fg[perm] + loss = torch.dot(errors_sorted, lovasz_grad(fg_sorted)) + if class_weight is not None: + loss *= class_weight[c] + losses.append(loss) + return torch.stack(losses).mean() + + +def lovasz_softmax(probs, + labels, + classes='present', + per_image=False, + class_weight=None, + reduction='mean', + avg_factor=None, + ignore_index=255): + """Multi-class Lovasz-Softmax loss. + + Args: + probs (torch.Tensor): [B, C, H, W], class probabilities at each + prediction (between 0 and 1). + labels (torch.Tensor): [B, H, W], ground truth labels (between 0 and + C - 1). + classes (str | list[int], optional): Classes chosen to calculate loss. + 'all' for all classes, 'present' for classes present in labels, or + a list of classes to average. Default: 'present'. + per_image (bool, optional): If per_image is True, compute the loss per + image instead of per batch. Default: False. + class_weight (list[float], optional): The weight for each class. + Default: None. + reduction (str, optional): The method used to reduce the loss. Options + are "none", "mean" and "sum". This parameter only works when + per_image is True. Default: 'mean'. + avg_factor (int, optional): Average factor that is used to average + the loss. This parameter only works when per_image is True. + Default: None. + ignore_index (int | None): The label index to be ignored. Default: 255. + + Returns: + torch.Tensor: The calculated loss. + """ + + if per_image: + loss = [ + lovasz_softmax_flat( + *flatten_probs( + prob.unsqueeze(0), label.unsqueeze(0), ignore_index), + classes=classes, + class_weight=class_weight) + for prob, label in zip(probs, labels) + ] + loss = weight_reduce_loss( + torch.stack(loss), None, reduction, avg_factor) + else: + loss = lovasz_softmax_flat( + *flatten_probs(probs, labels, ignore_index), + classes=classes, + class_weight=class_weight) + return loss + + +@LOSSES.register_module() +class LovaszLoss(nn.Module): + """LovaszLoss. + + This loss is proposed in `The Lovasz-Softmax loss: A tractable surrogate + for the optimization of the intersection-over-union measure in neural + networks `_. + + Args: + loss_type (str, optional): Binary or multi-class loss. + Default: 'multi_class'. Options are "binary" and "multi_class". + classes (str | list[int], optional): Classes chosen to calculate loss. + 'all' for all classes, 'present' for classes present in labels, or + a list of classes to average. Default: 'present'. + per_image (bool, optional): If per_image is True, compute the loss per + image instead of per batch. Default: False. + reduction (str, optional): The method used to reduce the loss. Options + are "none", "mean" and "sum". This parameter only works when + per_image is True. Default: 'mean'. + class_weight (list[float] | str, optional): Weight of each class. If in + str format, read them from a file. Defaults to None. + loss_weight (float, optional): Weight of the loss. Defaults to 1.0. + loss_name (str, optional): Name of the loss item. If you want this loss + item to be included into the backward graph, `loss_` must be the + prefix of the name. Defaults to 'loss_lovasz'. + """ + + def __init__(self, + loss_type='multi_class', + classes='present', + per_image=False, + reduction='mean', + class_weight=None, + loss_weight=1.0, + loss_name='loss_lovasz'): + super(LovaszLoss, self).__init__() + assert loss_type in ('binary', 'multi_class'), "loss_type should be \ + 'binary' or 'multi_class'." + + if loss_type == 'binary': + self.cls_criterion = lovasz_hinge + else: + self.cls_criterion = lovasz_softmax + assert classes in ('all', 'present') or mmcv.is_list_of(classes, int) + if not per_image: + assert reduction == 'none', "reduction should be 'none' when \ + per_image is False." + + self.classes = classes + self.per_image = per_image + self.reduction = reduction + self.loss_weight = loss_weight + self.class_weight = get_class_weight(class_weight) + self._loss_name = loss_name + + def forward(self, + cls_score, + label, + weight=None, + avg_factor=None, + reduction_override=None, + **kwargs): + """Forward function.""" + assert reduction_override in (None, 'none', 'mean', 'sum') + reduction = ( + reduction_override if reduction_override else self.reduction) + if self.class_weight is not None: + class_weight = cls_score.new_tensor(self.class_weight) + else: + class_weight = None + + # if multi-class loss, transform logits to probs + if self.cls_criterion == lovasz_softmax: + cls_score = F.softmax(cls_score, dim=1) + + loss_cls = self.loss_weight * self.cls_criterion( + cls_score, + label, + self.classes, + self.per_image, + class_weight=class_weight, + reduction=reduction, + avg_factor=avg_factor, + **kwargs) + return loss_cls + + @property + def loss_name(self): + """Loss Name. + + This function must be implemented and will return the name of this + loss function. This name will be used to combine different loss items + by simple sum operation. In addition, if you want this loss item to be + included into the backward graph, `loss_` must be the prefix of the + name. + Returns: + str: The name of this loss item. + """ + return self._loss_name diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/utils.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/utils.py new file mode 100644 index 0000000000000000000000000000000000000000..621f57c746c4de0abc3ad7a3c2ad35ef65c2fe32 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/losses/utils.py @@ -0,0 +1,126 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import functools + +import mmcv +import numpy as np +import torch +import torch.nn.functional as F + + +def get_class_weight(class_weight): + """Get class weight for loss function. + + Args: + class_weight (list[float] | str | None): If class_weight is a str, + take it as a file name and read from it. + """ + if isinstance(class_weight, str): + # take it as a file path + if class_weight.endswith('.npy'): + class_weight = np.load(class_weight) + else: + # pkl, json or yaml + class_weight = mmcv.load(class_weight) + + return class_weight + + +def reduce_loss(loss, reduction): + """Reduce loss as specified. + + Args: + loss (Tensor): Elementwise loss tensor. + reduction (str): Options are "none", "mean" and "sum". + + Return: + Tensor: Reduced loss tensor. + """ + reduction_enum = F._Reduction.get_enum(reduction) + # none: 0, elementwise_mean:1, sum: 2 + if reduction_enum == 0: + return loss + elif reduction_enum == 1: + return loss.mean() + elif reduction_enum == 2: + return loss.sum() + + +def weight_reduce_loss(loss, weight=None, reduction='mean', avg_factor=None): + """Apply element-wise weight and reduce loss. + + Args: + loss (Tensor): Element-wise loss. + weight (Tensor): Element-wise weights. + reduction (str): Same as built-in losses of PyTorch. + avg_factor (float): Average factor when computing the mean of losses. + + Returns: + Tensor: Processed loss values. + """ + # if weight is specified, apply element-wise weight + if weight is not None: + assert weight.dim() == loss.dim() + if weight.dim() > 1: + assert weight.size(1) == 1 or weight.size(1) == loss.size(1) + loss = loss * weight + + # if avg_factor is not specified, just reduce the loss + if avg_factor is None: + loss = reduce_loss(loss, reduction) + else: + # if reduction is mean, then average the loss by avg_factor + if reduction == 'mean': + # Avoid causing ZeroDivisionError when avg_factor is 0.0, + # i.e., all labels of an image belong to ignore index. + eps = torch.finfo(torch.float32).eps + loss = loss.sum() / (avg_factor + eps) + # if reduction is 'none', then do nothing, otherwise raise an error + elif reduction != 'none': + raise ValueError('avg_factor can not be used with reduction="sum"') + return loss + + +def weighted_loss(loss_func): + """Create a weighted version of a given loss function. + + To use this decorator, the loss function must have the signature like + `loss_func(pred, target, **kwargs)`. The function only needs to compute + element-wise loss without any reduction. This decorator will add weight + and reduction arguments to the function. The decorated function will have + the signature like `loss_func(pred, target, weight=None, reduction='mean', + avg_factor=None, **kwargs)`. + + :Example: + + >>> import torch + >>> @weighted_loss + >>> def l1_loss(pred, target): + >>> return (pred - target).abs() + + >>> pred = torch.Tensor([0, 2, 3]) + >>> target = torch.Tensor([1, 1, 1]) + >>> weight = torch.Tensor([1, 0, 1]) + + >>> l1_loss(pred, target) + tensor(1.3333) + >>> l1_loss(pred, target, weight) + tensor(1.) + >>> l1_loss(pred, target, reduction='none') + tensor([1., 1., 2.]) + >>> l1_loss(pred, target, weight, avg_factor=2) + tensor(1.5000) + """ + + @functools.wraps(loss_func) + def wrapper(pred, + target, + weight=None, + reduction='mean', + avg_factor=None, + **kwargs): + # get element-wise loss + loss = loss_func(pred, target, **kwargs) + loss = weight_reduce_loss(loss, weight, reduction, avg_factor) + return loss + + return wrapper diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..ff03186a92b78f942e79cff9eec9f5e2784c359a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/__init__.py @@ -0,0 +1,11 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .featurepyramid import Feature2Pyramid +from .fpn import FPN +from .ic_neck import ICNeck +from .jpu import JPU +from .mla_neck import MLANeck +from .multilevel_neck import MultiLevelNeck + +__all__ = [ + 'FPN', 'MultiLevelNeck', 'MLANeck', 'ICNeck', 'JPU', 'Feature2Pyramid' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/featurepyramid.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/featurepyramid.py new file mode 100644 index 0000000000000000000000000000000000000000..82a00ceb1c4fa792538143f7af86b957822cce4d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/featurepyramid.py @@ -0,0 +1,67 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn +from mmcv.cnn import build_norm_layer + +from ..builder import NECKS + + +@NECKS.register_module() +class Feature2Pyramid(nn.Module): + """Feature2Pyramid. + + A neck structure connect ViT backbone and decoder_heads. + + Args: + embed_dims (int): Embedding dimension. + rescales (list[float]): Different sampling multiples were + used to obtain pyramid features. Default: [4, 2, 1, 0.5]. + norm_cfg (dict): Config dict for normalization layer. + Default: dict(type='SyncBN', requires_grad=True). + """ + + def __init__(self, + embed_dim, + rescales=[4, 2, 1, 0.5], + norm_cfg=dict(type='SyncBN', requires_grad=True)): + super(Feature2Pyramid, self).__init__() + self.rescales = rescales + self.upsample_4x = None + for k in self.rescales: + if k == 4: + self.upsample_4x = nn.Sequential( + nn.ConvTranspose2d( + embed_dim, embed_dim, kernel_size=2, stride=2), + build_norm_layer(norm_cfg, embed_dim)[1], + nn.GELU(), + nn.ConvTranspose2d( + embed_dim, embed_dim, kernel_size=2, stride=2), + ) + elif k == 2: + self.upsample_2x = nn.Sequential( + nn.ConvTranspose2d( + embed_dim, embed_dim, kernel_size=2, stride=2)) + elif k == 1: + self.identity = nn.Identity() + elif k == 0.5: + self.downsample_2x = nn.MaxPool2d(kernel_size=2, stride=2) + elif k == 0.25: + self.downsample_4x = nn.MaxPool2d(kernel_size=4, stride=4) + else: + raise KeyError(f'invalid {k} for feature2pyramid') + + def forward(self, inputs): + assert len(inputs) == len(self.rescales) + outputs = [] + if self.upsample_4x is not None: + ops = [ + self.upsample_4x, self.upsample_2x, self.identity, + self.downsample_2x + ] + else: + ops = [ + self.upsample_2x, self.identity, self.downsample_2x, + self.downsample_4x + ] + for i in range(len(inputs)): + outputs.append(ops[i](inputs[i])) + return tuple(outputs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/fpn.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/fpn.py new file mode 100644 index 0000000000000000000000000000000000000000..6997de9d4428a0625241b55c6d509afd8e2b515a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/fpn.py @@ -0,0 +1,213 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn +import torch.nn.functional as F +from mmcv.cnn import ConvModule +from mmcv.runner import BaseModule, auto_fp16 + +from mmseg.ops import resize +from ..builder import NECKS + + +@NECKS.register_module() +class FPN(BaseModule): + """Feature Pyramid Network. + + This neck is the implementation of `Feature Pyramid Networks for Object + Detection `_. + + Args: + in_channels (list[int]): Number of input channels per scale. + out_channels (int): Number of output channels (used at each scale). + num_outs (int): Number of output scales. + start_level (int): Index of the start input backbone level used to + build the feature pyramid. Default: 0. + end_level (int): Index of the end input backbone level (exclusive) to + build the feature pyramid. Default: -1, which means the last level. + add_extra_convs (bool | str): If bool, it decides whether to add conv + layers on top of the original feature maps. Default to False. + If True, its actual mode is specified by `extra_convs_on_inputs`. + If str, it specifies the source feature map of the extra convs. + Only the following options are allowed + + - 'on_input': Last feat map of neck inputs (i.e. backbone feature). + - 'on_lateral': Last feature map after lateral convs. + - 'on_output': The last output feature map after fpn convs. + extra_convs_on_inputs (bool, deprecated): Whether to apply extra convs + on the original feature from the backbone. If True, + it is equivalent to `add_extra_convs='on_input'`. If False, it is + equivalent to set `add_extra_convs='on_output'`. Default to True. + relu_before_extra_convs (bool): Whether to apply relu before the extra + conv. Default: False. + no_norm_on_lateral (bool): Whether to apply norm on lateral. + Default: False. + conv_cfg (dict): Config dict for convolution layer. Default: None. + norm_cfg (dict): Config dict for normalization layer. Default: None. + act_cfg (dict): Config dict for activation layer in ConvModule. + Default: None. + upsample_cfg (dict): Config dict for interpolate layer. + Default: dict(mode='nearest'). + init_cfg (dict or list[dict], optional): Initialization config dict. + + Example: + >>> import torch + >>> in_channels = [2, 3, 5, 7] + >>> scales = [340, 170, 84, 43] + >>> inputs = [torch.rand(1, c, s, s) + ... for c, s in zip(in_channels, scales)] + >>> self = FPN(in_channels, 11, len(in_channels)).eval() + >>> outputs = self.forward(inputs) + >>> for i in range(len(outputs)): + ... print(f'outputs[{i}].shape = {outputs[i].shape}') + outputs[0].shape = torch.Size([1, 11, 340, 340]) + outputs[1].shape = torch.Size([1, 11, 170, 170]) + outputs[2].shape = torch.Size([1, 11, 84, 84]) + outputs[3].shape = torch.Size([1, 11, 43, 43]) + """ + + def __init__(self, + in_channels, + out_channels, + num_outs, + start_level=0, + end_level=-1, + add_extra_convs=False, + extra_convs_on_inputs=False, + relu_before_extra_convs=False, + no_norm_on_lateral=False, + conv_cfg=None, + norm_cfg=None, + act_cfg=None, + upsample_cfg=dict(mode='nearest'), + init_cfg=dict( + type='Xavier', layer='Conv2d', distribution='uniform')): + super(FPN, self).__init__(init_cfg) + assert isinstance(in_channels, list) + self.in_channels = in_channels + self.out_channels = out_channels + self.num_ins = len(in_channels) + self.num_outs = num_outs + self.relu_before_extra_convs = relu_before_extra_convs + self.no_norm_on_lateral = no_norm_on_lateral + self.fp16_enabled = False + self.upsample_cfg = upsample_cfg.copy() + + if end_level == -1: + self.backbone_end_level = self.num_ins + assert num_outs >= self.num_ins - start_level + else: + # if end_level < inputs, no extra level is allowed + self.backbone_end_level = end_level + assert end_level <= len(in_channels) + assert num_outs == end_level - start_level + self.start_level = start_level + self.end_level = end_level + self.add_extra_convs = add_extra_convs + assert isinstance(add_extra_convs, (str, bool)) + if isinstance(add_extra_convs, str): + # Extra_convs_source choices: 'on_input', 'on_lateral', 'on_output' + assert add_extra_convs in ('on_input', 'on_lateral', 'on_output') + elif add_extra_convs: # True + if extra_convs_on_inputs: + # For compatibility with previous release + # TODO: deprecate `extra_convs_on_inputs` + self.add_extra_convs = 'on_input' + else: + self.add_extra_convs = 'on_output' + + self.lateral_convs = nn.ModuleList() + self.fpn_convs = nn.ModuleList() + + for i in range(self.start_level, self.backbone_end_level): + l_conv = ConvModule( + in_channels[i], + out_channels, + 1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg if not self.no_norm_on_lateral else None, + act_cfg=act_cfg, + inplace=False) + fpn_conv = ConvModule( + out_channels, + out_channels, + 3, + padding=1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg, + inplace=False) + + self.lateral_convs.append(l_conv) + self.fpn_convs.append(fpn_conv) + + # add extra conv layers (e.g., RetinaNet) + extra_levels = num_outs - self.backbone_end_level + self.start_level + if self.add_extra_convs and extra_levels >= 1: + for i in range(extra_levels): + if i == 0 and self.add_extra_convs == 'on_input': + in_channels = self.in_channels[self.backbone_end_level - 1] + else: + in_channels = out_channels + extra_fpn_conv = ConvModule( + in_channels, + out_channels, + 3, + stride=2, + padding=1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg, + inplace=False) + self.fpn_convs.append(extra_fpn_conv) + + @auto_fp16() + def forward(self, inputs): + assert len(inputs) == len(self.in_channels) + + # build laterals + laterals = [ + lateral_conv(inputs[i + self.start_level]) + for i, lateral_conv in enumerate(self.lateral_convs) + ] + + # build top-down path + used_backbone_levels = len(laterals) + for i in range(used_backbone_levels - 1, 0, -1): + # In some cases, fixing `scale factor` (e.g. 2) is preferred, but + # it cannot co-exist with `size` in `F.interpolate`. + if 'scale_factor' in self.upsample_cfg: + laterals[i - 1] = laterals[i - 1] + resize( + laterals[i], **self.upsample_cfg) + else: + prev_shape = laterals[i - 1].shape[2:] + laterals[i - 1] = laterals[i - 1] + resize( + laterals[i], size=prev_shape, **self.upsample_cfg) + + # build outputs + # part 1: from original levels + outs = [ + self.fpn_convs[i](laterals[i]) for i in range(used_backbone_levels) + ] + # part 2: add extra levels + if self.num_outs > len(outs): + # use max pool to get more levels on top of outputs + # (e.g., Faster R-CNN, Mask R-CNN) + if not self.add_extra_convs: + for i in range(self.num_outs - used_backbone_levels): + outs.append(F.max_pool2d(outs[-1], 1, stride=2)) + # add conv layers on top of original feature maps (RetinaNet) + else: + if self.add_extra_convs == 'on_input': + extra_source = inputs[self.backbone_end_level - 1] + elif self.add_extra_convs == 'on_lateral': + extra_source = laterals[-1] + elif self.add_extra_convs == 'on_output': + extra_source = outs[-1] + else: + raise NotImplementedError + outs.append(self.fpn_convs[used_backbone_levels](extra_source)) + for i in range(used_backbone_levels + 1, self.num_outs): + if self.relu_before_extra_convs: + outs.append(self.fpn_convs[i](F.relu(outs[-1]))) + else: + outs.append(self.fpn_convs[i](outs[-1])) + return tuple(outs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/ic_neck.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/ic_neck.py new file mode 100644 index 0000000000000000000000000000000000000000..a5d81cef8e16bd82f11c5ed35c04564a702b667d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/ic_neck.py @@ -0,0 +1,148 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn.functional as F +from mmcv.cnn import ConvModule +from mmcv.runner import BaseModule + +from mmseg.ops import resize +from ..builder import NECKS + + +class CascadeFeatureFusion(BaseModule): + """Cascade Feature Fusion Unit in ICNet. + + Args: + low_channels (int): The number of input channels for + low resolution feature map. + high_channels (int): The number of input channels for + high resolution feature map. + out_channels (int): The number of output channels. + conv_cfg (dict): Dictionary to construct and config conv layer. + Default: None. + norm_cfg (dict): Dictionary to construct and config norm layer. + Default: dict(type='BN'). + act_cfg (dict): Dictionary to construct and config act layer. + Default: dict(type='ReLU'). + align_corners (bool): align_corners argument of F.interpolate. + Default: False. + init_cfg (dict or list[dict], optional): Initialization config dict. + Default: None. + + Returns: + x (Tensor): The output tensor of shape (N, out_channels, H, W). + x_low (Tensor): The output tensor of shape (N, out_channels, H, W) + for Cascade Label Guidance in auxiliary heads. + """ + + def __init__(self, + low_channels, + high_channels, + out_channels, + conv_cfg=None, + norm_cfg=dict(type='BN'), + act_cfg=dict(type='ReLU'), + align_corners=False, + init_cfg=None): + super(CascadeFeatureFusion, self).__init__(init_cfg=init_cfg) + self.align_corners = align_corners + self.conv_low = ConvModule( + low_channels, + out_channels, + 3, + padding=2, + dilation=2, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + self.conv_high = ConvModule( + high_channels, + out_channels, + 1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + + def forward(self, x_low, x_high): + x_low = resize( + x_low, + size=x_high.size()[2:], + mode='bilinear', + align_corners=self.align_corners) + # Note: Different from original paper, `x_low` is underwent + # `self.conv_low` rather than another 1x1 conv classifier + # before being used for auxiliary head. + x_low = self.conv_low(x_low) + x_high = self.conv_high(x_high) + x = x_low + x_high + x = F.relu(x, inplace=True) + return x, x_low + + +@NECKS.register_module() +class ICNeck(BaseModule): + """ICNet for Real-Time Semantic Segmentation on High-Resolution Images. + + This head is the implementation of `ICHead + `_. + + Args: + in_channels (int): The number of input image channels. Default: 3. + out_channels (int): The numbers of output feature channels. + Default: 128. + conv_cfg (dict): Dictionary to construct and config conv layer. + Default: None. + norm_cfg (dict): Dictionary to construct and config norm layer. + Default: dict(type='BN'). + act_cfg (dict): Dictionary to construct and config act layer. + Default: dict(type='ReLU'). + align_corners (bool): align_corners argument of F.interpolate. + Default: False. + init_cfg (dict or list[dict], optional): Initialization config dict. + Default: None. + """ + + def __init__(self, + in_channels=(64, 256, 256), + out_channels=128, + conv_cfg=None, + norm_cfg=dict(type='BN'), + act_cfg=dict(type='ReLU'), + align_corners=False, + init_cfg=None): + super(ICNeck, self).__init__(init_cfg=init_cfg) + assert len(in_channels) == 3, 'Length of input channels \ + must be 3!' + + self.in_channels = in_channels + self.out_channels = out_channels + self.conv_cfg = conv_cfg + self.norm_cfg = norm_cfg + self.act_cfg = act_cfg + self.align_corners = align_corners + self.cff_24 = CascadeFeatureFusion( + self.in_channels[2], + self.in_channels[1], + self.out_channels, + conv_cfg=self.conv_cfg, + norm_cfg=self.norm_cfg, + act_cfg=self.act_cfg, + align_corners=self.align_corners) + + self.cff_12 = CascadeFeatureFusion( + self.out_channels, + self.in_channels[0], + self.out_channels, + conv_cfg=self.conv_cfg, + norm_cfg=self.norm_cfg, + act_cfg=self.act_cfg, + align_corners=self.align_corners) + + def forward(self, inputs): + assert len(inputs) == 3, 'Length of input feature \ + maps must be 3!' + + x_sub1, x_sub2, x_sub4 = inputs + x_cff_24, x_24 = self.cff_24(x_sub4, x_sub2) + x_cff_12, x_12 = self.cff_12(x_cff_24, x_sub1) + # Note: `x_cff_12` is used for decode_head, + # `x_24` and `x_12` are used for auxiliary head. + return x_24, x_12, x_cff_12 diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/jpu.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/jpu.py new file mode 100644 index 0000000000000000000000000000000000000000..3cc6b9f428911a48a7fe3f1f2913812e45e4737e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/jpu.py @@ -0,0 +1,131 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +from mmcv.cnn import ConvModule, DepthwiseSeparableConvModule +from mmcv.runner import BaseModule + +from mmseg.ops import resize +from ..builder import NECKS + + +@NECKS.register_module() +class JPU(BaseModule): + """FastFCN: Rethinking Dilated Convolution in the Backbone + for Semantic Segmentation. + + This Joint Pyramid Upsampling (JPU) neck is the implementation of + `FastFCN `_. + + Args: + in_channels (Tuple[int], optional): The number of input channels + for each convolution operations before upsampling. + Default: (512, 1024, 2048). + mid_channels (int): The number of output channels of JPU. + Default: 512. + start_level (int): Index of the start input backbone level used to + build the feature pyramid. Default: 0. + end_level (int): Index of the end input backbone level (exclusive) to + build the feature pyramid. Default: -1, which means the last level. + dilations (tuple[int]): Dilation rate of each Depthwise + Separable ConvModule. Default: (1, 2, 4, 8). + align_corners (bool, optional): The align_corners argument of + resize operation. Default: False. + conv_cfg (dict | None): Config of conv layers. + Default: None. + norm_cfg (dict | None): Config of norm layers. + Default: dict(type='BN'). + act_cfg (dict): Config of activation layers. + Default: dict(type='ReLU'). + init_cfg (dict or list[dict], optional): Initialization config dict. + Default: None. + """ + + def __init__(self, + in_channels=(512, 1024, 2048), + mid_channels=512, + start_level=0, + end_level=-1, + dilations=(1, 2, 4, 8), + align_corners=False, + conv_cfg=None, + norm_cfg=dict(type='BN'), + act_cfg=dict(type='ReLU'), + init_cfg=None): + super(JPU, self).__init__(init_cfg=init_cfg) + assert isinstance(in_channels, tuple) + assert isinstance(dilations, tuple) + self.in_channels = in_channels + self.mid_channels = mid_channels + self.start_level = start_level + self.num_ins = len(in_channels) + if end_level == -1: + self.backbone_end_level = self.num_ins + else: + self.backbone_end_level = end_level + assert end_level <= len(in_channels) + + self.dilations = dilations + self.align_corners = align_corners + + self.conv_layers = nn.ModuleList() + self.dilation_layers = nn.ModuleList() + for i in range(self.start_level, self.backbone_end_level): + conv_layer = nn.Sequential( + ConvModule( + self.in_channels[i], + self.mid_channels, + kernel_size=3, + padding=1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg)) + self.conv_layers.append(conv_layer) + for i in range(len(dilations)): + dilation_layer = nn.Sequential( + DepthwiseSeparableConvModule( + in_channels=(self.backbone_end_level - self.start_level) * + self.mid_channels, + out_channels=self.mid_channels, + kernel_size=3, + stride=1, + padding=dilations[i], + dilation=dilations[i], + dw_norm_cfg=norm_cfg, + dw_act_cfg=None, + pw_norm_cfg=norm_cfg, + pw_act_cfg=act_cfg)) + self.dilation_layers.append(dilation_layer) + + def forward(self, inputs): + """Forward function.""" + assert len(inputs) == len(self.in_channels), 'Length of inputs must \ + be the same with self.in_channels!' + + feats = [ + self.conv_layers[i - self.start_level](inputs[i]) + for i in range(self.start_level, self.backbone_end_level) + ] + + h, w = feats[0].shape[2:] + for i in range(1, len(feats)): + feats[i] = resize( + feats[i], + size=(h, w), + mode='bilinear', + align_corners=self.align_corners) + + feat = torch.cat(feats, dim=1) + concat_feat = torch.cat([ + self.dilation_layers[i](feat) for i in range(len(self.dilations)) + ], + dim=1) + + outs = [] + + # Default: outs[2] is the output of JPU for decoder head, outs[1] is + # the feature map from backbone for auxiliary head. Additionally, + # outs[0] can also be used for auxiliary head. + for i in range(self.start_level, self.backbone_end_level - 1): + outs.append(inputs[i]) + outs.append(concat_feat) + return tuple(outs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/mla_neck.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/mla_neck.py new file mode 100644 index 0000000000000000000000000000000000000000..1513e296daaedf83bea23ab2c168fb63482bed23 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/mla_neck.py @@ -0,0 +1,118 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn +from mmcv.cnn import ConvModule, build_norm_layer + +from ..builder import NECKS + + +class MLAModule(nn.Module): + + def __init__(self, + in_channels=[1024, 1024, 1024, 1024], + out_channels=256, + norm_cfg=None, + act_cfg=None): + super(MLAModule, self).__init__() + self.channel_proj = nn.ModuleList() + for i in range(len(in_channels)): + self.channel_proj.append( + ConvModule( + in_channels=in_channels[i], + out_channels=out_channels, + kernel_size=1, + norm_cfg=norm_cfg, + act_cfg=act_cfg)) + self.feat_extract = nn.ModuleList() + for i in range(len(in_channels)): + self.feat_extract.append( + ConvModule( + in_channels=out_channels, + out_channels=out_channels, + kernel_size=3, + padding=1, + norm_cfg=norm_cfg, + act_cfg=act_cfg)) + + def forward(self, inputs): + + # feat_list -> [p2, p3, p4, p5] + feat_list = [] + for x, conv in zip(inputs, self.channel_proj): + feat_list.append(conv(x)) + + # feat_list -> [p5, p4, p3, p2] + # mid_list -> [m5, m4, m3, m2] + feat_list = feat_list[::-1] + mid_list = [] + for feat in feat_list: + if len(mid_list) == 0: + mid_list.append(feat) + else: + mid_list.append(mid_list[-1] + feat) + + # mid_list -> [m5, m4, m3, m2] + # out_list -> [o2, o3, o4, o5] + out_list = [] + for mid, conv in zip(mid_list, self.feat_extract): + out_list.append(conv(mid)) + + return tuple(out_list) + + +@NECKS.register_module() +class MLANeck(nn.Module): + """Multi-level Feature Aggregation. + + This neck is `The Multi-level Feature Aggregation construction of + SETR `_. + + + Args: + in_channels (List[int]): Number of input channels per scale. + out_channels (int): Number of output channels (used at each scale). + norm_layer (dict): Config dict for input normalization. + Default: norm_layer=dict(type='LN', eps=1e-6, requires_grad=True). + norm_cfg (dict): Config dict for normalization layer. Default: None. + act_cfg (dict): Config dict for activation layer in ConvModule. + Default: None. + """ + + def __init__(self, + in_channels, + out_channels, + norm_layer=dict(type='LN', eps=1e-6, requires_grad=True), + norm_cfg=None, + act_cfg=None): + super(MLANeck, self).__init__() + assert isinstance(in_channels, list) + self.in_channels = in_channels + self.out_channels = out_channels + + # In order to build general vision transformer backbone, we have to + # move MLA to neck. + self.norm = nn.ModuleList([ + build_norm_layer(norm_layer, in_channels[i])[1] + for i in range(len(in_channels)) + ]) + + self.mla = MLAModule( + in_channels=in_channels, + out_channels=out_channels, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + + def forward(self, inputs): + assert len(inputs) == len(self.in_channels) + + # Convert from nchw to nlc + outs = [] + for i in range(len(inputs)): + x = inputs[i] + n, c, h, w = x.shape + x = x.reshape(n, c, h * w).transpose(2, 1).contiguous() + x = self.norm[i](x) + x = x.transpose(1, 2).reshape(n, c, h, w).contiguous() + outs.append(x) + + outs = self.mla(outs) + return tuple(outs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/multilevel_neck.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/multilevel_neck.py new file mode 100644 index 0000000000000000000000000000000000000000..5151f8762de46ae3e41da8f9683ee8df6f70711e --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/necks/multilevel_neck.py @@ -0,0 +1,78 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch.nn as nn +from mmcv.cnn import ConvModule, xavier_init + +from mmseg.ops import resize +from ..builder import NECKS + + +@NECKS.register_module() +class MultiLevelNeck(nn.Module): + """MultiLevelNeck. + + A neck structure connect vit backbone and decoder_heads. + + Args: + in_channels (List[int]): Number of input channels per scale. + out_channels (int): Number of output channels (used at each scale). + scales (List[float]): Scale factors for each input feature map. + Default: [0.5, 1, 2, 4] + norm_cfg (dict): Config dict for normalization layer. Default: None. + act_cfg (dict): Config dict for activation layer in ConvModule. + Default: None. + """ + + def __init__(self, + in_channels, + out_channels, + scales=[0.5, 1, 2, 4], + norm_cfg=None, + act_cfg=None): + super(MultiLevelNeck, self).__init__() + assert isinstance(in_channels, list) + self.in_channels = in_channels + self.out_channels = out_channels + self.scales = scales + self.num_outs = len(scales) + self.lateral_convs = nn.ModuleList() + self.convs = nn.ModuleList() + for in_channel in in_channels: + self.lateral_convs.append( + ConvModule( + in_channel, + out_channels, + kernel_size=1, + norm_cfg=norm_cfg, + act_cfg=act_cfg)) + for _ in range(self.num_outs): + self.convs.append( + ConvModule( + out_channels, + out_channels, + kernel_size=3, + padding=1, + stride=1, + norm_cfg=norm_cfg, + act_cfg=act_cfg)) + + # default init_weights for conv(msra) and norm in ConvModule + def init_weights(self): + for m in self.modules(): + if isinstance(m, nn.Conv2d): + xavier_init(m, distribution='uniform') + + def forward(self, inputs): + assert len(inputs) == len(self.in_channels) + inputs = [ + lateral_conv(inputs[i]) + for i, lateral_conv in enumerate(self.lateral_convs) + ] + # for len(inputs) not equal to self.num_outs + if len(inputs) == 1: + inputs = [inputs[0] for _ in range(self.num_outs)] + outs = [] + for i in range(self.num_outs): + x_resize = resize( + inputs[i], scale_factor=self.scales[i], mode='bilinear') + outs.append(self.convs[i](x_resize)) + return tuple(outs) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..387c858bd7a4e1e222db0fe99d85f4728ff48f21 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/__init__.py @@ -0,0 +1,6 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .base import BaseSegmentor +from .cascade_encoder_decoder import CascadeEncoderDecoder +from .encoder_decoder import EncoderDecoder + +__all__ = ['BaseSegmentor', 'EncoderDecoder', 'CascadeEncoderDecoder'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/base.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/base.py new file mode 100644 index 0000000000000000000000000000000000000000..76dc8f075a848d3df9e0d8d4f123ca458fb93aa4 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/base.py @@ -0,0 +1,291 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings +from abc import ABCMeta, abstractmethod +from collections import OrderedDict + +import mmcv +import numpy as np +import torch +import torch.distributed as dist +from mmcv.runner import BaseModule, auto_fp16 + + +class BaseSegmentor(BaseModule, metaclass=ABCMeta): + """Base class for segmentors.""" + + def __init__(self, init_cfg=None): + super(BaseSegmentor, self).__init__(init_cfg) + self.fp16_enabled = False + + @property + def with_neck(self): + """bool: whether the segmentor has neck""" + return hasattr(self, 'neck') and self.neck is not None + + @property + def with_auxiliary_head(self): + """bool: whether the segmentor has auxiliary head""" + return hasattr(self, + 'auxiliary_head') and self.auxiliary_head is not None + + @property + def with_decode_head(self): + """bool: whether the segmentor has decode head""" + return hasattr(self, 'decode_head') and self.decode_head is not None + + @abstractmethod + def extract_feat(self, imgs): + """Placeholder for extract features from images.""" + pass + + @abstractmethod + def encode_decode(self, img, img_metas): + """Placeholder for encode images with backbone and decode into a + semantic segmentation map of the same size as input.""" + pass + + @abstractmethod + def forward_train(self, imgs, img_metas, **kwargs): + """Placeholder for Forward function for training.""" + pass + + @abstractmethod + def simple_test(self, img, img_meta, **kwargs): + """Placeholder for single image test.""" + pass + + @abstractmethod + def aug_test(self, imgs, img_metas, **kwargs): + """Placeholder for augmentation test.""" + pass + + def forward_test(self, imgs, img_metas, **kwargs): + """ + Args: + imgs (List[Tensor]): the outer list indicates test-time + augmentations and inner Tensor should have a shape NxCxHxW, + which contains all images in the batch. + img_metas (List[List[dict]]): the outer list indicates test-time + augs (multiscale, flip, etc.) and the inner list indicates + images in a batch. + """ + for var, name in [(imgs, 'imgs'), (img_metas, 'img_metas')]: + if not isinstance(var, list): + raise TypeError(f'{name} must be a list, but got ' + f'{type(var)}') + + num_augs = len(imgs) + if num_augs != len(img_metas): + raise ValueError(f'num of augmentations ({len(imgs)}) != ' + f'num of image meta ({len(img_metas)})') + # all images in the same aug batch all of the same ori_shape and pad + # shape + for img_meta in img_metas: + ori_shapes = [_['ori_shape'] for _ in img_meta] + assert all(shape == ori_shapes[0] for shape in ori_shapes) + img_shapes = [_['img_shape'] for _ in img_meta] + assert all(shape == img_shapes[0] for shape in img_shapes) + pad_shapes = [_['pad_shape'] for _ in img_meta] + assert all(shape == pad_shapes[0] for shape in pad_shapes) + + if num_augs == 1: + return self.simple_test(imgs[0], img_metas[0], **kwargs) + else: + return self.aug_test(imgs, img_metas, **kwargs) + + @auto_fp16(apply_to=('img', )) + def forward(self, img, img_metas, return_loss=True, **kwargs): + """Calls either :func:`forward_train` or :func:`forward_test` depending + on whether ``return_loss`` is ``True``. + + Note this setting will change the expected inputs. When + ``return_loss=True``, img and img_meta are single-nested (i.e. Tensor + and List[dict]), and when ``resturn_loss=False``, img and img_meta + should be double nested (i.e. List[Tensor], List[List[dict]]), with + the outer list indicating test time augmentations. + """ + if return_loss: + return self.forward_train(img, img_metas, **kwargs) + else: + return self.forward_test(img, img_metas, **kwargs) + + def train_step(self, data_batch, optimizer, **kwargs): + """The iteration step during training. + + This method defines an iteration step during training, except for the + back propagation and optimizer updating, which are done in an optimizer + hook. Note that in some complicated cases or models, the whole process + including back propagation and optimizer updating is also defined in + this method, such as GAN. + + Args: + data (dict): The output of dataloader. + optimizer (:obj:`torch.optim.Optimizer` | dict): The optimizer of + runner is passed to ``train_step()``. This argument is unused + and reserved. + + Returns: + dict: It should contain at least 3 keys: ``loss``, ``log_vars``, + ``num_samples``. + ``loss`` is a tensor for back propagation, which can be a + weighted sum of multiple losses. + ``log_vars`` contains all the variables to be sent to the + logger. + ``num_samples`` indicates the batch size (when the model is + DDP, it means the batch size on each GPU), which is used for + averaging the logs. + """ + losses = self(**data_batch) + loss, log_vars = self._parse_losses(losses) + + outputs = dict( + loss=loss, + log_vars=log_vars, + num_samples=len(data_batch['img_metas'])) + + return outputs + + def val_step(self, data_batch, optimizer=None, **kwargs): + """The iteration step during validation. + + This method shares the same signature as :func:`train_step`, but used + during val epochs. Note that the evaluation after training epochs is + not implemented with this method, but an evaluation hook. + """ + losses = self(**data_batch) + loss, log_vars = self._parse_losses(losses) + + log_vars_ = dict() + for loss_name, loss_value in log_vars.items(): + k = loss_name + '_val' + log_vars_[k] = loss_value + + outputs = dict( + loss=loss, + log_vars=log_vars_, + num_samples=len(data_batch['img_metas'])) + + return outputs + + @staticmethod + def _parse_losses(losses): + """Parse the raw outputs (losses) of the network. + + Args: + losses (dict): Raw output of the network, which usually contain + losses and other necessary information. + + Returns: + tuple[Tensor, dict]: (loss, log_vars), loss is the loss tensor + which may be a weighted sum of all losses, log_vars contains + all the variables to be sent to the logger. + """ + log_vars = OrderedDict() + for loss_name, loss_value in losses.items(): + if isinstance(loss_value, torch.Tensor): + log_vars[loss_name] = loss_value.mean() + elif isinstance(loss_value, list): + log_vars[loss_name] = sum(_loss.mean() for _loss in loss_value) + else: + raise TypeError( + f'{loss_name} is not a tensor or list of tensors') + + loss = sum(_value for _key, _value in log_vars.items() + if 'loss' in _key) + + # If the loss_vars has different length, raise assertion error + # to prevent GPUs from infinite waiting. + if dist.is_available() and dist.is_initialized(): + log_var_length = torch.tensor(len(log_vars), device=loss.device) + dist.all_reduce(log_var_length) + message = (f'rank {dist.get_rank()}' + + f' len(log_vars): {len(log_vars)}' + ' keys: ' + + ','.join(log_vars.keys()) + '\n') + assert log_var_length == len(log_vars) * dist.get_world_size(), \ + 'loss log variables are different across GPUs!\n' + message + + log_vars['loss'] = loss + for loss_name, loss_value in log_vars.items(): + # reduce loss when distributed training + if dist.is_available() and dist.is_initialized(): + loss_value = loss_value.data.clone() + dist.all_reduce(loss_value.div_(dist.get_world_size())) + log_vars[loss_name] = loss_value.item() + + return loss, log_vars + + def show_result(self, + img, + result, + palette=None, + win_name='', + show=False, + wait_time=0, + out_file=None, + opacity=0.5): + """Draw `result` over `img`. + + Args: + img (str or Tensor): The image to be displayed. + result (Tensor): The semantic segmentation results to draw over + `img`. + palette (list[list[int]]] | np.ndarray | None): The palette of + segmentation map. If None is given, random palette will be + generated. Default: None + win_name (str): The window name. + wait_time (int): Value of waitKey param. + Default: 0. + show (bool): Whether to show the image. + Default: False. + out_file (str or None): The filename to write the image. + Default: None. + opacity(float): Opacity of painted segmentation map. + Default 0.5. + Must be in (0, 1] range. + Returns: + img (Tensor): Only if not `show` or `out_file` + """ + img = mmcv.imread(img) + img = img.copy() + seg = result[0] + if palette is None: + if self.PALETTE is None: + # Get random state before set seed, + # and restore random state later. + # It will prevent loss of randomness, as the palette + # may be different in each iteration if not specified. + # See: https://github.com/open-mmlab/mmdetection/issues/5844 + state = np.random.get_state() + np.random.seed(42) + # random palette + palette = np.random.randint( + 0, 255, size=(len(self.CLASSES), 3)) + np.random.set_state(state) + else: + palette = self.PALETTE + palette = np.array(palette) + assert palette.shape[0] == len(self.CLASSES) + assert palette.shape[1] == 3 + assert len(palette.shape) == 2 + assert 0 < opacity <= 1.0 + color_seg = np.zeros((seg.shape[0], seg.shape[1], 3), dtype=np.uint8) + for label, color in enumerate(palette): + color_seg[seg == label, :] = color + # convert to BGR + color_seg = color_seg[..., ::-1] + + img = img * (1 - opacity) + color_seg * opacity + img = img.astype(np.uint8) + # if out_file specified, do not show image in window + if out_file is not None: + show = False + + if show: + mmcv.imshow(img, win_name, wait_time) + if out_file is not None: + mmcv.imwrite(img, out_file) + + if not (show or out_file): + warnings.warn('show==False and out_file is not specified, only ' + 'result image will be returned') + return img diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/cascade_encoder_decoder.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/cascade_encoder_decoder.py new file mode 100644 index 0000000000000000000000000000000000000000..1913a22e204bd77d0d98fc4f9e9ebaf9723d3ffe --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/cascade_encoder_decoder.py @@ -0,0 +1,88 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from torch import nn + +from mmseg.core import add_prefix +from mmseg.ops import resize +from .. import builder +from ..builder import SEGMENTORS +from .encoder_decoder import EncoderDecoder + + +@SEGMENTORS.register_module() +class CascadeEncoderDecoder(EncoderDecoder): + """Cascade Encoder Decoder segmentors. + + CascadeEncoderDecoder almost the same as EncoderDecoder, while decoders of + CascadeEncoderDecoder are cascaded. The output of previous decoder_head + will be the input of next decoder_head. + """ + + def __init__(self, + num_stages, + backbone, + decode_head, + neck=None, + auxiliary_head=None, + train_cfg=None, + test_cfg=None, + pretrained=None, + init_cfg=None): + self.num_stages = num_stages + super(CascadeEncoderDecoder, self).__init__( + backbone=backbone, + decode_head=decode_head, + neck=neck, + auxiliary_head=auxiliary_head, + train_cfg=train_cfg, + test_cfg=test_cfg, + pretrained=pretrained, + init_cfg=init_cfg) + + def _init_decode_head(self, decode_head): + """Initialize ``decode_head``""" + assert isinstance(decode_head, list) + assert len(decode_head) == self.num_stages + self.decode_head = nn.ModuleList() + for i in range(self.num_stages): + self.decode_head.append(builder.build_head(decode_head[i])) + self.align_corners = self.decode_head[-1].align_corners + self.num_classes = self.decode_head[-1].num_classes + + def encode_decode(self, img, img_metas): + """Encode images with backbone and decode into a semantic segmentation + map of the same size as input.""" + x = self.extract_feat(img) + out = self.decode_head[0].forward_test(x, img_metas, self.test_cfg) + for i in range(1, self.num_stages): + out = self.decode_head[i].forward_test(x, out, img_metas, + self.test_cfg) + out = resize( + input=out, + size=img.shape[2:], + mode='bilinear', + align_corners=self.align_corners) + return out + + def _decode_head_forward_train(self, x, img_metas, gt_semantic_seg): + """Run forward function and calculate loss for decode head in + training.""" + losses = dict() + + loss_decode = self.decode_head[0].forward_train( + x, img_metas, gt_semantic_seg, self.train_cfg) + + losses.update(add_prefix(loss_decode, 'decode_0')) + + for i in range(1, self.num_stages): + # forward test again, maybe unnecessary for most methods. + if i == 1: + prev_outputs = self.decode_head[0].forward_test( + x, img_metas, self.test_cfg) + else: + prev_outputs = self.decode_head[i - 1].forward_test( + x, prev_outputs, img_metas, self.test_cfg) + loss_decode = self.decode_head[i].forward_train( + x, prev_outputs, img_metas, gt_semantic_seg, self.train_cfg) + losses.update(add_prefix(loss_decode, f'decode_{i}')) + + return losses diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/encoder_decoder.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/encoder_decoder.py new file mode 100644 index 0000000000000000000000000000000000000000..72467b4690263e981e73e87a52b438a10d743746 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/segmentors/encoder_decoder.py @@ -0,0 +1,284 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +import torch.nn.functional as F + +from mmseg.core import add_prefix +from mmseg.ops import resize +from .. import builder +from ..builder import SEGMENTORS +from .base import BaseSegmentor + + +@SEGMENTORS.register_module() +class EncoderDecoder(BaseSegmentor): + """Encoder Decoder segmentors. + + EncoderDecoder typically consists of backbone, decode_head, auxiliary_head. + Note that auxiliary_head is only used for deep supervision during training, + which could be dumped during inference. + """ + + def __init__(self, + backbone, + decode_head, + neck=None, + auxiliary_head=None, + train_cfg=None, + test_cfg=None, + pretrained=None, + init_cfg=None): + super(EncoderDecoder, self).__init__(init_cfg) + if pretrained is not None: + assert backbone.get('pretrained') is None, \ + 'both backbone and segmentor set pretrained weight' + backbone.pretrained = pretrained + self.backbone = builder.build_backbone(backbone) + if neck is not None: + self.neck = builder.build_neck(neck) + self._init_decode_head(decode_head) + self._init_auxiliary_head(auxiliary_head) + + self.train_cfg = train_cfg + self.test_cfg = test_cfg + + assert self.with_decode_head + + def _init_decode_head(self, decode_head): + """Initialize ``decode_head``""" + self.decode_head = builder.build_head(decode_head) + self.align_corners = self.decode_head.align_corners + self.num_classes = self.decode_head.num_classes + + def _init_auxiliary_head(self, auxiliary_head): + """Initialize ``auxiliary_head``""" + if auxiliary_head is not None: + if isinstance(auxiliary_head, list): + self.auxiliary_head = nn.ModuleList() + for head_cfg in auxiliary_head: + self.auxiliary_head.append(builder.build_head(head_cfg)) + else: + self.auxiliary_head = builder.build_head(auxiliary_head) + + def extract_feat(self, img): + """Extract features from images.""" + x = self.backbone(img) + if self.with_neck: + x = self.neck(x) + return x + + def encode_decode(self, img, img_metas): + """Encode images with backbone and decode into a semantic segmentation + map of the same size as input.""" + x = self.extract_feat(img) + out = self._decode_head_forward_test(x, img_metas) + out = resize( + input=out, + size=img.shape[2:], + mode='bilinear', + align_corners=self.align_corners) + return out + + def _decode_head_forward_train(self, x, img_metas, gt_semantic_seg): + """Run forward function and calculate loss for decode head in + training.""" + losses = dict() + loss_decode = self.decode_head.forward_train(x, img_metas, + gt_semantic_seg, + self.train_cfg) + + losses.update(add_prefix(loss_decode, 'decode')) + return losses + + def _decode_head_forward_test(self, x, img_metas): + """Run forward function and calculate loss for decode head in + inference.""" + seg_logits = self.decode_head.forward_test(x, img_metas, self.test_cfg) + return seg_logits + + def _auxiliary_head_forward_train(self, x, img_metas, gt_semantic_seg): + """Run forward function and calculate loss for auxiliary head in + training.""" + losses = dict() + if isinstance(self.auxiliary_head, nn.ModuleList): + for idx, aux_head in enumerate(self.auxiliary_head): + loss_aux = aux_head.forward_train(x, img_metas, + gt_semantic_seg, + self.train_cfg) + losses.update(add_prefix(loss_aux, f'aux_{idx}')) + else: + loss_aux = self.auxiliary_head.forward_train( + x, img_metas, gt_semantic_seg, self.train_cfg) + losses.update(add_prefix(loss_aux, 'aux')) + + return losses + + def forward_dummy(self, img): + """Dummy forward function.""" + seg_logit = self.encode_decode(img, None) + + return seg_logit + + def forward_train(self, img, img_metas, gt_semantic_seg): + """Forward function for training. + + Args: + img (Tensor): Input images. + img_metas (list[dict]): List of image info dict where each dict + has: 'img_shape', 'scale_factor', 'flip', and may also contain + 'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'. + For details on the values of these keys see + `mmseg/datasets/pipelines/formatting.py:Collect`. + gt_semantic_seg (Tensor): Semantic segmentation masks + used if the architecture supports semantic segmentation task. + + Returns: + dict[str, Tensor]: a dictionary of loss components + """ + + x = self.extract_feat(img) + + losses = dict() + + loss_decode = self._decode_head_forward_train(x, img_metas, + gt_semantic_seg) + losses.update(loss_decode) + + if self.with_auxiliary_head: + loss_aux = self._auxiliary_head_forward_train( + x, img_metas, gt_semantic_seg) + losses.update(loss_aux) + + return losses + + # TODO refactor + def slide_inference(self, img, img_meta, rescale): + """Inference by sliding-window with overlap. + + If h_crop > h_img or w_crop > w_img, the small patch will be used to + decode without padding. + """ + + h_stride, w_stride = self.test_cfg.stride + h_crop, w_crop = self.test_cfg.crop_size + batch_size, _, h_img, w_img = img.size() + num_classes = self.num_classes + h_grids = max(h_img - h_crop + h_stride - 1, 0) // h_stride + 1 + w_grids = max(w_img - w_crop + w_stride - 1, 0) // w_stride + 1 + preds = img.new_zeros((batch_size, num_classes, h_img, w_img)) + count_mat = img.new_zeros((batch_size, 1, h_img, w_img)) + for h_idx in range(h_grids): + for w_idx in range(w_grids): + y1 = h_idx * h_stride + x1 = w_idx * w_stride + y2 = min(y1 + h_crop, h_img) + x2 = min(x1 + w_crop, w_img) + y1 = max(y2 - h_crop, 0) + x1 = max(x2 - w_crop, 0) + crop_img = img[:, :, y1:y2, x1:x2] + crop_seg_logit = self.encode_decode(crop_img, img_meta) + preds += F.pad(crop_seg_logit, + (int(x1), int(preds.shape[3] - x2), int(y1), + int(preds.shape[2] - y2))) + + count_mat[:, :, y1:y2, x1:x2] += 1 + assert (count_mat == 0).sum() == 0 + if torch.onnx.is_in_onnx_export(): + # cast count_mat to constant while exporting to ONNX + count_mat = torch.from_numpy( + count_mat.cpu().detach().numpy()).to(device=img.device) + preds = preds / count_mat + if rescale: + preds = resize( + preds, + size=img_meta[0]['ori_shape'][:2], + mode='bilinear', + align_corners=self.align_corners, + warning=False) + return preds + + def whole_inference(self, img, img_meta, rescale): + """Inference with full image.""" + + seg_logit = self.encode_decode(img, img_meta) + if rescale: + # support dynamic shape for onnx + if torch.onnx.is_in_onnx_export(): + size = img.shape[2:] + else: + size = img_meta[0]['ori_shape'][:2] + seg_logit = resize( + seg_logit, + size=size, + mode='bilinear', + align_corners=self.align_corners, + warning=False) + + return seg_logit + + def inference(self, img, img_meta, rescale): + """Inference with slide/whole style. + + Args: + img (Tensor): The input image of shape (N, 3, H, W). + img_meta (dict): Image info dict where each dict has: 'img_shape', + 'scale_factor', 'flip', and may also contain + 'filename', 'ori_shape', 'pad_shape', and 'img_norm_cfg'. + For details on the values of these keys see + `mmseg/datasets/pipelines/formatting.py:Collect`. + rescale (bool): Whether rescale back to original shape. + + Returns: + Tensor: The output segmentation map. + """ + + assert self.test_cfg.mode in ['slide', 'whole'] + ori_shape = img_meta[0]['ori_shape'] + assert all(_['ori_shape'] == ori_shape for _ in img_meta) + if self.test_cfg.mode == 'slide': + seg_logit = self.slide_inference(img, img_meta, rescale) + else: + seg_logit = self.whole_inference(img, img_meta, rescale) + output = F.softmax(seg_logit, dim=1) + flip = img_meta[0]['flip'] + if flip: + flip_direction = img_meta[0]['flip_direction'] + assert flip_direction in ['horizontal', 'vertical'] + if flip_direction == 'horizontal': + output = output.flip(dims=(3, )) + elif flip_direction == 'vertical': + output = output.flip(dims=(2, )) + + return output + + def simple_test(self, img, img_meta, rescale=True): + """Simple test with single image.""" + seg_logit = self.inference(img, img_meta, rescale) + seg_pred = seg_logit.argmax(dim=1) + if torch.onnx.is_in_onnx_export(): + # our inference backend only support 4D output + seg_pred = seg_pred.unsqueeze(0) + return seg_pred + seg_pred = seg_pred.cpu().numpy() + # unravel batch dim + seg_pred = list(seg_pred) + return seg_pred + + def aug_test(self, imgs, img_metas, rescale=True): + """Test with augmentations. + + Only rescale=True is supported. + """ + # aug_test rescale all imgs back to ori_shape for now + assert rescale + # to save memory, we get augmented seg logit inplace + seg_logit = self.inference(imgs[0], img_metas[0], rescale) + for i in range(1, len(imgs)): + cur_seg_logit = self.inference(imgs[i], img_metas[i], rescale) + seg_logit += cur_seg_logit + seg_logit /= len(imgs) + seg_pred = seg_logit.argmax(dim=1) + seg_pred = seg_pred.cpu().numpy() + # unravel batch dim + seg_pred = list(seg_pred) + return seg_pred diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..fc281e3d674ac3f89afd54f9a39550c045fd9b40 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/__init__.py @@ -0,0 +1,18 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .embed import PatchEmbed +from .inverted_residual import InvertedResidual, InvertedResidualV3 +from .make_divisible import make_divisible +from .res_layer import ResLayer +from .se_layer import SELayer +from .self_attention_block import SelfAttentionBlock +from .shape_convert import (nchw2nlc2nchw, nchw_to_nlc, nlc2nchw2nlc, + nlc_to_nchw) +from .up_conv_block import UpConvBlock +from .wrappers import resize, Upsample + +__all__ = [ + 'ResLayer', 'SelfAttentionBlock', 'make_divisible', 'InvertedResidual', + 'UpConvBlock', 'InvertedResidualV3', 'SELayer', 'PatchEmbed', + 'nchw_to_nlc', 'nlc_to_nchw', 'nchw2nlc2nchw', 'nlc2nchw2nlc', + 'resize', 'Upsample' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/embed.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/embed.py new file mode 100644 index 0000000000000000000000000000000000000000..1515675e1eeec92ca1b6d417c5e202ce8aecc35f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/embed.py @@ -0,0 +1,330 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import math +from typing import Sequence + +import torch.nn as nn +import torch.nn.functional as F +from mmcv.cnn import build_conv_layer, build_norm_layer +from mmcv.runner.base_module import BaseModule +from mmcv.utils import to_2tuple + + +class AdaptivePadding(nn.Module): + """Applies padding to input (if needed) so that input can get fully covered + by filter you specified. It support two modes "same" and "corner". The + "same" mode is same with "SAME" padding mode in TensorFlow, pad zero around + input. The "corner" mode would pad zero to bottom right. + + Args: + kernel_size (int | tuple): Size of the kernel: + stride (int | tuple): Stride of the filter. Default: 1: + dilation (int | tuple): Spacing between kernel elements. + Default: 1. + padding (str): Support "same" and "corner", "corner" mode + would pad zero to bottom right, and "same" mode would + pad zero around input. Default: "corner". + Example: + >>> kernel_size = 16 + >>> stride = 16 + >>> dilation = 1 + >>> input = torch.rand(1, 1, 15, 17) + >>> adap_pad = AdaptivePadding( + >>> kernel_size=kernel_size, + >>> stride=stride, + >>> dilation=dilation, + >>> padding="corner") + >>> out = adap_pad(input) + >>> assert (out.shape[2], out.shape[3]) == (16, 32) + >>> input = torch.rand(1, 1, 16, 17) + >>> out = adap_pad(input) + >>> assert (out.shape[2], out.shape[3]) == (16, 32) + """ + + def __init__(self, kernel_size=1, stride=1, dilation=1, padding='corner'): + + super(AdaptivePadding, self).__init__() + + assert padding in ('same', 'corner') + + kernel_size = to_2tuple(kernel_size) + stride = to_2tuple(stride) + dilation = to_2tuple(dilation) + + self.padding = padding + self.kernel_size = kernel_size + self.stride = stride + self.dilation = dilation + + def get_pad_shape(self, input_shape): + input_h, input_w = input_shape + kernel_h, kernel_w = self.kernel_size + stride_h, stride_w = self.stride + output_h = math.ceil(input_h / stride_h) + output_w = math.ceil(input_w / stride_w) + pad_h = max((output_h - 1) * stride_h + + (kernel_h - 1) * self.dilation[0] + 1 - input_h, 0) + pad_w = max((output_w - 1) * stride_w + + (kernel_w - 1) * self.dilation[1] + 1 - input_w, 0) + return pad_h, pad_w + + def forward(self, x): + pad_h, pad_w = self.get_pad_shape(x.size()[-2:]) + if pad_h > 0 or pad_w > 0: + if self.padding == 'corner': + x = F.pad(x, [0, pad_w, 0, pad_h]) + elif self.padding == 'same': + x = F.pad(x, [ + pad_w // 2, pad_w - pad_w // 2, pad_h // 2, + pad_h - pad_h // 2 + ]) + return x + + +class PatchEmbed(BaseModule): + """Image to Patch Embedding. + + We use a conv layer to implement PatchEmbed. + + Args: + in_channels (int): The num of input channels. Default: 3 + embed_dims (int): The dimensions of embedding. Default: 768 + conv_type (str): The config dict for embedding + conv layer type selection. Default: "Conv2d". + kernel_size (int): The kernel_size of embedding conv. Default: 16. + stride (int, optional): The slide stride of embedding conv. + Default: None (Would be set as `kernel_size`). + padding (int | tuple | string ): The padding length of + embedding conv. When it is a string, it means the mode + of adaptive padding, support "same" and "corner" now. + Default: "corner". + dilation (int): The dilation rate of embedding conv. Default: 1. + bias (bool): Bias of embed conv. Default: True. + norm_cfg (dict, optional): Config dict for normalization layer. + Default: None. + input_size (int | tuple | None): The size of input, which will be + used to calculate the out size. Only work when `dynamic_size` + is False. Default: None. + init_cfg (`mmcv.ConfigDict`, optional): The Config for initialization. + Default: None. + """ + + def __init__(self, + in_channels=3, + embed_dims=768, + conv_type='Conv2d', + kernel_size=16, + stride=None, + padding='corner', + dilation=1, + bias=True, + norm_cfg=None, + input_size=None, + init_cfg=None): + super(PatchEmbed, self).__init__(init_cfg=init_cfg) + + self.embed_dims = embed_dims + if stride is None: + stride = kernel_size + + kernel_size = to_2tuple(kernel_size) + stride = to_2tuple(stride) + dilation = to_2tuple(dilation) + + if isinstance(padding, str): + self.adap_padding = AdaptivePadding( + kernel_size=kernel_size, + stride=stride, + dilation=dilation, + padding=padding) + # disable the padding of conv + padding = 0 + else: + self.adap_padding = None + padding = to_2tuple(padding) + + self.projection = build_conv_layer( + dict(type=conv_type), + in_channels=in_channels, + out_channels=embed_dims, + kernel_size=kernel_size, + stride=stride, + padding=padding, + dilation=dilation, + bias=bias) + + if norm_cfg is not None: + self.norm = build_norm_layer(norm_cfg, embed_dims)[1] + else: + self.norm = None + + if input_size: + input_size = to_2tuple(input_size) + # `init_out_size` would be used outside to + # calculate the num_patches + # when `use_abs_pos_embed` outside + self.init_input_size = input_size + if self.adap_padding: + pad_h, pad_w = self.adap_padding.get_pad_shape(input_size) + input_h, input_w = input_size + input_h = input_h + pad_h + input_w = input_w + pad_w + input_size = (input_h, input_w) + + # https://pytorch.org/docs/stable/generated/torch.nn.Conv2d.html + h_out = (input_size[0] + 2 * padding[0] - dilation[0] * + (kernel_size[0] - 1) - 1) // stride[0] + 1 + w_out = (input_size[1] + 2 * padding[1] - dilation[1] * + (kernel_size[1] - 1) - 1) // stride[1] + 1 + self.init_out_size = (h_out, w_out) + else: + self.init_input_size = None + self.init_out_size = None + + def forward(self, x): + """ + Args: + x (Tensor): Has shape (B, C, H, W). In most case, C is 3. + + Returns: + tuple: Contains merged results and its spatial shape. + + - x (Tensor): Has shape (B, out_h * out_w, embed_dims) + - out_size (tuple[int]): Spatial shape of x, arrange as + (out_h, out_w). + """ + + if self.adap_padding: + x = self.adap_padding(x) + + x = self.projection(x) + out_size = (x.shape[2], x.shape[3]) + x = x.flatten(2).transpose(1, 2) + if self.norm is not None: + x = self.norm(x) + return x, out_size + + +class PatchMerging(BaseModule): + """Merge patch feature map. + + This layer groups feature map by kernel_size, and applies norm and linear + layers to the grouped feature map. Our implementation uses `nn.Unfold` to + merge patch, which is about 25% faster than original implementation. + Instead, we need to modify pretrained models for compatibility. + + Args: + in_channels (int): The num of input channels. + out_channels (int): The num of output channels. + kernel_size (int | tuple, optional): the kernel size in the unfold + layer. Defaults to 2. + stride (int | tuple, optional): the stride of the sliding blocks in the + unfold layer. Default: None. (Would be set as `kernel_size`) + padding (int | tuple | string ): The padding length of + embedding conv. When it is a string, it means the mode + of adaptive padding, support "same" and "corner" now. + Default: "corner". + dilation (int | tuple, optional): dilation parameter in the unfold + layer. Default: 1. + bias (bool, optional): Whether to add bias in linear layer or not. + Defaults: False. + norm_cfg (dict, optional): Config dict for normalization layer. + Default: dict(type='LN'). + init_cfg (dict, optional): The extra config for initialization. + Default: None. + """ + + def __init__(self, + in_channels, + out_channels, + kernel_size=2, + stride=None, + padding='corner', + dilation=1, + bias=False, + norm_cfg=dict(type='LN'), + init_cfg=None): + super().__init__(init_cfg=init_cfg) + self.in_channels = in_channels + self.out_channels = out_channels + if stride: + stride = stride + else: + stride = kernel_size + + kernel_size = to_2tuple(kernel_size) + stride = to_2tuple(stride) + dilation = to_2tuple(dilation) + + if isinstance(padding, str): + self.adap_padding = AdaptivePadding( + kernel_size=kernel_size, + stride=stride, + dilation=dilation, + padding=padding) + # disable the padding of unfold + padding = 0 + else: + self.adap_padding = None + + padding = to_2tuple(padding) + self.sampler = nn.Unfold( + kernel_size=kernel_size, + dilation=dilation, + padding=padding, + stride=stride) + + sample_dim = kernel_size[0] * kernel_size[1] * in_channels + + if norm_cfg is not None: + self.norm = build_norm_layer(norm_cfg, sample_dim)[1] + else: + self.norm = None + + self.reduction = nn.Linear(sample_dim, out_channels, bias=bias) + + def forward(self, x, input_size): + """ + Args: + x (Tensor): Has shape (B, H*W, C_in). + input_size (tuple[int]): The spatial shape of x, arrange as (H, W). + Default: None. + + Returns: + tuple: Contains merged results and its spatial shape. + + - x (Tensor): Has shape (B, Merged_H * Merged_W, C_out) + - out_size (tuple[int]): Spatial shape of x, arrange as + (Merged_H, Merged_W). + """ + B, L, C = x.shape + assert isinstance(input_size, Sequence), f'Expect ' \ + f'input_size is ' \ + f'`Sequence` ' \ + f'but get {input_size}' + + H, W = input_size + assert L == H * W, 'input feature has wrong size' + + x = x.view(B, H, W, C).permute([0, 3, 1, 2]) # B, C, H, W + # Use nn.Unfold to merge patch. About 25% faster than original method, + # but need to modify pretrained model for compatibility + + if self.adap_padding: + x = self.adap_padding(x) + H, W = x.shape[-2:] + + x = self.sampler(x) + # if kernel_size=2 and stride=2, x should has shape (B, 4*C, H/2*W/2) + + out_h = (H + 2 * self.sampler.padding[0] - self.sampler.dilation[0] * + (self.sampler.kernel_size[0] - 1) - + 1) // self.sampler.stride[0] + 1 + out_w = (W + 2 * self.sampler.padding[1] - self.sampler.dilation[1] * + (self.sampler.kernel_size[1] - 1) - + 1) // self.sampler.stride[1] + 1 + + output_size = (out_h, out_w) + x = x.transpose(1, 2) # B, H/2*W/2, 4*C + x = self.norm(x) if self.norm else x + x = self.reduction(x) + return x, output_size diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/inverted_residual.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/inverted_residual.py new file mode 100644 index 0000000000000000000000000000000000000000..c9cda76822a45baa0cbd27c98f1e7196e1e26e9a --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/inverted_residual.py @@ -0,0 +1,213 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.cnn import ConvModule +from torch import nn +from torch.utils import checkpoint as cp + +from .se_layer import SELayer + + +class InvertedResidual(nn.Module): + """InvertedResidual block for MobileNetV2. + + Args: + in_channels (int): The input channels of the InvertedResidual block. + out_channels (int): The output channels of the InvertedResidual block. + stride (int): Stride of the middle (first) 3x3 convolution. + expand_ratio (int): Adjusts number of channels of the hidden layer + in InvertedResidual by this amount. + dilation (int): Dilation rate of depthwise conv. Default: 1 + conv_cfg (dict): Config dict for convolution layer. + Default: None, which means using conv2d. + norm_cfg (dict): Config dict for normalization layer. + Default: dict(type='BN'). + act_cfg (dict): Config dict for activation layer. + Default: dict(type='ReLU6'). + with_cp (bool): Use checkpoint or not. Using checkpoint will save some + memory while slowing down the training speed. Default: False. + + Returns: + Tensor: The output tensor. + """ + + def __init__(self, + in_channels, + out_channels, + stride, + expand_ratio, + dilation=1, + conv_cfg=None, + norm_cfg=dict(type='BN'), + act_cfg=dict(type='ReLU6'), + with_cp=False, + **kwargs): + super(InvertedResidual, self).__init__() + self.stride = stride + assert stride in [1, 2], f'stride must in [1, 2]. ' \ + f'But received {stride}.' + self.with_cp = with_cp + self.use_res_connect = self.stride == 1 and in_channels == out_channels + hidden_dim = int(round(in_channels * expand_ratio)) + + layers = [] + if expand_ratio != 1: + layers.append( + ConvModule( + in_channels=in_channels, + out_channels=hidden_dim, + kernel_size=1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg, + **kwargs)) + layers.extend([ + ConvModule( + in_channels=hidden_dim, + out_channels=hidden_dim, + kernel_size=3, + stride=stride, + padding=dilation, + dilation=dilation, + groups=hidden_dim, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg, + **kwargs), + ConvModule( + in_channels=hidden_dim, + out_channels=out_channels, + kernel_size=1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=None, + **kwargs) + ]) + self.conv = nn.Sequential(*layers) + + def forward(self, x): + + def _inner_forward(x): + if self.use_res_connect: + return x + self.conv(x) + else: + return self.conv(x) + + if self.with_cp and x.requires_grad: + out = cp.checkpoint(_inner_forward, x) + else: + out = _inner_forward(x) + + return out + + +class InvertedResidualV3(nn.Module): + """Inverted Residual Block for MobileNetV3. + + Args: + in_channels (int): The input channels of this Module. + out_channels (int): The output channels of this Module. + mid_channels (int): The input channels of the depthwise convolution. + kernel_size (int): The kernel size of the depthwise convolution. + Default: 3. + stride (int): The stride of the depthwise convolution. Default: 1. + se_cfg (dict): Config dict for se layer. Default: None, which means no + se layer. + with_expand_conv (bool): Use expand conv or not. If set False, + mid_channels must be the same with in_channels. Default: True. + conv_cfg (dict): Config dict for convolution layer. Default: None, + which means using conv2d. + norm_cfg (dict): Config dict for normalization layer. + Default: dict(type='BN'). + act_cfg (dict): Config dict for activation layer. + Default: dict(type='ReLU'). + with_cp (bool): Use checkpoint or not. Using checkpoint will save some + memory while slowing down the training speed. Default: False. + + Returns: + Tensor: The output tensor. + """ + + def __init__(self, + in_channels, + out_channels, + mid_channels, + kernel_size=3, + stride=1, + se_cfg=None, + with_expand_conv=True, + conv_cfg=None, + norm_cfg=dict(type='BN'), + act_cfg=dict(type='ReLU'), + with_cp=False): + super(InvertedResidualV3, self).__init__() + self.with_res_shortcut = (stride == 1 and in_channels == out_channels) + assert stride in [1, 2] + self.with_cp = with_cp + self.with_se = se_cfg is not None + self.with_expand_conv = with_expand_conv + + if self.with_se: + assert isinstance(se_cfg, dict) + if not self.with_expand_conv: + assert mid_channels == in_channels + + if self.with_expand_conv: + self.expand_conv = ConvModule( + in_channels=in_channels, + out_channels=mid_channels, + kernel_size=1, + stride=1, + padding=0, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + self.depthwise_conv = ConvModule( + in_channels=mid_channels, + out_channels=mid_channels, + kernel_size=kernel_size, + stride=stride, + padding=kernel_size // 2, + groups=mid_channels, + conv_cfg=dict( + type='Conv2dAdaptivePadding') if stride == 2 else conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + + if self.with_se: + self.se = SELayer(**se_cfg) + + self.linear_conv = ConvModule( + in_channels=mid_channels, + out_channels=out_channels, + kernel_size=1, + stride=1, + padding=0, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=None) + + def forward(self, x): + + def _inner_forward(x): + out = x + + if self.with_expand_conv: + out = self.expand_conv(out) + + out = self.depthwise_conv(out) + + if self.with_se: + out = self.se(out) + + out = self.linear_conv(out) + + if self.with_res_shortcut: + return x + out + else: + return out + + if self.with_cp and x.requires_grad: + out = cp.checkpoint(_inner_forward, x) + else: + out = _inner_forward(x) + + return out diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/make_divisible.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/make_divisible.py new file mode 100644 index 0000000000000000000000000000000000000000..ed42c2eeea2a6aed03a0be5516b8d1ef1139e486 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/make_divisible.py @@ -0,0 +1,28 @@ +# Copyright (c) OpenMMLab. All rights reserved. +def make_divisible(value, divisor, min_value=None, min_ratio=0.9): + """Make divisible function. + + This function rounds the channel number to the nearest value that can be + divisible by the divisor. It is taken from the original tf repo. It ensures + that all layers have a channel number that is divisible by divisor. It can + be seen here: https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py # noqa + + Args: + value (int): The original channel number. + divisor (int): The divisor to fully divide the channel number. + min_value (int): The minimum value of the output channel. + Default: None, means that the minimum value equal to the divisor. + min_ratio (float): The minimum ratio of the rounded channel number to + the original channel number. Default: 0.9. + + Returns: + int: The modified output channel number. + """ + + if min_value is None: + min_value = divisor + new_value = max(min_value, int(value + divisor / 2) // divisor * divisor) + # Make sure that round down does not go down by more than (1-min_ratio). + if new_value < min_ratio * value: + new_value += divisor + return new_value diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/res_layer.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/res_layer.py new file mode 100644 index 0000000000000000000000000000000000000000..190a0c5d5a63846c9a42c6622635cf2456bd6635 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/res_layer.py @@ -0,0 +1,96 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.cnn import build_conv_layer, build_norm_layer +from mmcv.runner import Sequential +from torch import nn as nn + + +class ResLayer(Sequential): + """ResLayer to build ResNet style backbone. + + Args: + block (nn.Module): block used to build ResLayer. + inplanes (int): inplanes of block. + planes (int): planes of block. + num_blocks (int): number of blocks. + stride (int): stride of the first block. Default: 1 + avg_down (bool): Use AvgPool instead of stride conv when + downsampling in the bottleneck. Default: False + conv_cfg (dict): dictionary to construct and config conv layer. + Default: None + norm_cfg (dict): dictionary to construct and config norm layer. + Default: dict(type='BN') + multi_grid (int | None): Multi grid dilation rates of last + stage. Default: None + contract_dilation (bool): Whether contract first dilation of each layer + Default: False + """ + + def __init__(self, + block, + inplanes, + planes, + num_blocks, + stride=1, + dilation=1, + avg_down=False, + conv_cfg=None, + norm_cfg=dict(type='BN'), + multi_grid=None, + contract_dilation=False, + **kwargs): + self.block = block + + downsample = None + if stride != 1 or inplanes != planes * block.expansion: + downsample = [] + conv_stride = stride + if avg_down: + conv_stride = 1 + downsample.append( + nn.AvgPool2d( + kernel_size=stride, + stride=stride, + ceil_mode=True, + count_include_pad=False)) + downsample.extend([ + build_conv_layer( + conv_cfg, + inplanes, + planes * block.expansion, + kernel_size=1, + stride=conv_stride, + bias=False), + build_norm_layer(norm_cfg, planes * block.expansion)[1] + ]) + downsample = nn.Sequential(*downsample) + + layers = [] + if multi_grid is None: + if dilation > 1 and contract_dilation: + first_dilation = dilation // 2 + else: + first_dilation = dilation + else: + first_dilation = multi_grid[0] + layers.append( + block( + inplanes=inplanes, + planes=planes, + stride=stride, + dilation=first_dilation, + downsample=downsample, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + **kwargs)) + inplanes = planes * block.expansion + for i in range(1, num_blocks): + layers.append( + block( + inplanes=inplanes, + planes=planes, + stride=1, + dilation=dilation if multi_grid is None else multi_grid[i], + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + **kwargs)) + super(ResLayer, self).__init__(*layers) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/se_layer.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/se_layer.py new file mode 100644 index 0000000000000000000000000000000000000000..16f52aa5c03b982c0d5f9ab9f145f48515a7ffa2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/se_layer.py @@ -0,0 +1,58 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import mmcv +import torch.nn as nn +from mmcv.cnn import ConvModule + +from .make_divisible import make_divisible + + +class SELayer(nn.Module): + """Squeeze-and-Excitation Module. + + Args: + channels (int): The input (and output) channels of the SE layer. + ratio (int): Squeeze ratio in SELayer, the intermediate channel will be + ``int(channels/ratio)``. Default: 16. + conv_cfg (None or dict): Config dict for convolution layer. + Default: None, which means using conv2d. + act_cfg (dict or Sequence[dict]): Config dict for activation layer. + If act_cfg is a dict, two activation layers will be configured + by this dict. If act_cfg is a sequence of dicts, the first + activation layer will be configured by the first dict and the + second activation layer will be configured by the second dict. + Default: (dict(type='ReLU'), dict(type='HSigmoid', bias=3.0, + divisor=6.0)). + """ + + def __init__(self, + channels, + ratio=16, + conv_cfg=None, + act_cfg=(dict(type='ReLU'), + dict(type='HSigmoid', bias=3.0, divisor=6.0))): + super(SELayer, self).__init__() + if isinstance(act_cfg, dict): + act_cfg = (act_cfg, act_cfg) + assert len(act_cfg) == 2 + assert mmcv.is_tuple_of(act_cfg, dict) + self.global_avgpool = nn.AdaptiveAvgPool2d(1) + self.conv1 = ConvModule( + in_channels=channels, + out_channels=make_divisible(channels // ratio, 8), + kernel_size=1, + stride=1, + conv_cfg=conv_cfg, + act_cfg=act_cfg[0]) + self.conv2 = ConvModule( + in_channels=make_divisible(channels // ratio, 8), + out_channels=channels, + kernel_size=1, + stride=1, + conv_cfg=conv_cfg, + act_cfg=act_cfg[1]) + + def forward(self, x): + out = self.global_avgpool(x) + out = self.conv1(out) + out = self.conv2(out) + return x * out diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/self_attention_block.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/self_attention_block.py new file mode 100644 index 0000000000000000000000000000000000000000..c945fa7168208fff513c4d397ad2c9a7ac4383ad --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/self_attention_block.py @@ -0,0 +1,160 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +from mmcv.cnn import ConvModule, constant_init +from torch import nn as nn +from torch.nn import functional as F + + +class SelfAttentionBlock(nn.Module): + """General self-attention block/non-local block. + + Please refer to https://arxiv.org/abs/1706.03762 for details about key, + query and value. + + Args: + key_in_channels (int): Input channels of key feature. + query_in_channels (int): Input channels of query feature. + channels (int): Output channels of key/query transform. + out_channels (int): Output channels. + share_key_query (bool): Whether share projection weight between key + and query projection. + query_downsample (nn.Module): Query downsample module. + key_downsample (nn.Module): Key downsample module. + key_query_num_convs (int): Number of convs for key/query projection. + value_num_convs (int): Number of convs for value projection. + matmul_norm (bool): Whether normalize attention map with sqrt of + channels + with_out (bool): Whether use out projection. + conv_cfg (dict|None): Config of conv layers. + norm_cfg (dict|None): Config of norm layers. + act_cfg (dict|None): Config of activation layers. + """ + + def __init__(self, key_in_channels, query_in_channels, channels, + out_channels, share_key_query, query_downsample, + key_downsample, key_query_num_convs, value_out_num_convs, + key_query_norm, value_out_norm, matmul_norm, with_out, + conv_cfg, norm_cfg, act_cfg): + super(SelfAttentionBlock, self).__init__() + if share_key_query: + assert key_in_channels == query_in_channels + self.key_in_channels = key_in_channels + self.query_in_channels = query_in_channels + self.out_channels = out_channels + self.channels = channels + self.share_key_query = share_key_query + self.conv_cfg = conv_cfg + self.norm_cfg = norm_cfg + self.act_cfg = act_cfg + self.key_project = self.build_project( + key_in_channels, + channels, + num_convs=key_query_num_convs, + use_conv_module=key_query_norm, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + if share_key_query: + self.query_project = self.key_project + else: + self.query_project = self.build_project( + query_in_channels, + channels, + num_convs=key_query_num_convs, + use_conv_module=key_query_norm, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + self.value_project = self.build_project( + key_in_channels, + channels if with_out else out_channels, + num_convs=value_out_num_convs, + use_conv_module=value_out_norm, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + if with_out: + self.out_project = self.build_project( + channels, + out_channels, + num_convs=value_out_num_convs, + use_conv_module=value_out_norm, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + else: + self.out_project = None + + self.query_downsample = query_downsample + self.key_downsample = key_downsample + self.matmul_norm = matmul_norm + + self.init_weights() + + def init_weights(self): + """Initialize weight of later layer.""" + if self.out_project is not None: + if not isinstance(self.out_project, ConvModule): + constant_init(self.out_project, 0) + + def build_project(self, in_channels, channels, num_convs, use_conv_module, + conv_cfg, norm_cfg, act_cfg): + """Build projection layer for key/query/value/out.""" + if use_conv_module: + convs = [ + ConvModule( + in_channels, + channels, + 1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + ] + for _ in range(num_convs - 1): + convs.append( + ConvModule( + channels, + channels, + 1, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg)) + else: + convs = [nn.Conv2d(in_channels, channels, 1)] + for _ in range(num_convs - 1): + convs.append(nn.Conv2d(channels, channels, 1)) + if len(convs) > 1: + convs = nn.Sequential(*convs) + else: + convs = convs[0] + return convs + + def forward(self, query_feats, key_feats): + """Forward function.""" + batch_size = query_feats.size(0) + query = self.query_project(query_feats) + if self.query_downsample is not None: + query = self.query_downsample(query) + query = query.reshape(*query.shape[:2], -1) + query = query.permute(0, 2, 1).contiguous() + + key = self.key_project(key_feats) + value = self.value_project(key_feats) + if self.key_downsample is not None: + key = self.key_downsample(key) + value = self.key_downsample(value) + key = key.reshape(*key.shape[:2], -1) + value = value.reshape(*value.shape[:2], -1) + value = value.permute(0, 2, 1).contiguous() + + sim_map = torch.matmul(query, key) + if self.matmul_norm: + sim_map = (self.channels**-.5) * sim_map + sim_map = F.softmax(sim_map, dim=-1) + + context = torch.matmul(sim_map, value) + context = context.permute(0, 2, 1).contiguous() + context = context.reshape(batch_size, -1, *query_feats.shape[2:]) + if self.out_project is not None: + context = self.out_project(context) + return context diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/shape_convert.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/shape_convert.py new file mode 100644 index 0000000000000000000000000000000000000000..cce1e220b645d4b02df1ec2d9ed3137c8acba707 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/shape_convert.py @@ -0,0 +1,107 @@ +# Copyright (c) OpenMMLab. All rights reserved. +def nlc_to_nchw(x, hw_shape): + """Convert [N, L, C] shape tensor to [N, C, H, W] shape tensor. + + Args: + x (Tensor): The input tensor of shape [N, L, C] before conversion. + hw_shape (Sequence[int]): The height and width of output feature map. + + Returns: + Tensor: The output tensor of shape [N, C, H, W] after conversion. + """ + H, W = hw_shape + assert len(x.shape) == 3 + B, L, C = x.shape + assert L == H * W, 'The seq_len doesn\'t match H, W' + return x.transpose(1, 2).reshape(B, C, H, W) + + +def nchw_to_nlc(x): + """Flatten [N, C, H, W] shape tensor to [N, L, C] shape tensor. + + Args: + x (Tensor): The input tensor of shape [N, C, H, W] before conversion. + + Returns: + Tensor: The output tensor of shape [N, L, C] after conversion. + """ + assert len(x.shape) == 4 + return x.flatten(2).transpose(1, 2).contiguous() + + +def nchw2nlc2nchw(module, x, contiguous=False, **kwargs): + """Flatten [N, C, H, W] shape tensor `x` to [N, L, C] shape tensor. Use the + reshaped tensor as the input of `module`, and the convert the output of + `module`, whose shape is. + + [N, L, C], to [N, C, H, W]. + + Args: + module (Callable): A callable object the takes a tensor + with shape [N, L, C] as input. + x (Tensor): The input tensor of shape [N, C, H, W]. + contiguous: + contiguous (Bool): Whether to make the tensor contiguous + after each shape transform. + + Returns: + Tensor: The output tensor of shape [N, C, H, W]. + + Example: + >>> import torch + >>> import torch.nn as nn + >>> norm = nn.LayerNorm(4) + >>> feature_map = torch.rand(4, 4, 5, 5) + >>> output = nchw2nlc2nchw(norm, feature_map) + """ + B, C, H, W = x.shape + if not contiguous: + x = x.flatten(2).transpose(1, 2) + x = module(x, **kwargs) + x = x.transpose(1, 2).reshape(B, C, H, W) + else: + x = x.flatten(2).transpose(1, 2).contiguous() + x = module(x, **kwargs) + x = x.transpose(1, 2).reshape(B, C, H, W).contiguous() + return x + + +def nlc2nchw2nlc(module, x, hw_shape, contiguous=False, **kwargs): + """Convert [N, L, C] shape tensor `x` to [N, C, H, W] shape tensor. Use the + reshaped tensor as the input of `module`, and convert the output of + `module`, whose shape is. + + [N, C, H, W], to [N, L, C]. + + Args: + module (Callable): A callable object the takes a tensor + with shape [N, C, H, W] as input. + x (Tensor): The input tensor of shape [N, L, C]. + hw_shape: (Sequence[int]): The height and width of the + feature map with shape [N, C, H, W]. + contiguous (Bool): Whether to make the tensor contiguous + after each shape transform. + + Returns: + Tensor: The output tensor of shape [N, L, C]. + + Example: + >>> import torch + >>> import torch.nn as nn + >>> conv = nn.Conv2d(16, 16, 3, 1, 1) + >>> feature_map = torch.rand(4, 25, 16) + >>> output = nlc2nchw2nlc(conv, feature_map, (5, 5)) + """ + H, W = hw_shape + assert len(x.shape) == 3 + B, L, C = x.shape + assert L == H * W, 'The seq_len doesn\'t match H, W' + if not contiguous: + x = x.transpose(1, 2).reshape(B, C, H, W) + x = module(x, **kwargs) + x = x.flatten(2).transpose(1, 2) + else: + x = x.transpose(1, 2).reshape(B, C, H, W).contiguous() + x = module(x, **kwargs) + x = x.flatten(2).transpose(1, 2).contiguous() + return x diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/up_conv_block.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/up_conv_block.py new file mode 100644 index 0000000000000000000000000000000000000000..d8396d9c2cc77135946ae4b43620e7b583915421 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/up_conv_block.py @@ -0,0 +1,102 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +import torch.nn as nn +from mmcv.cnn import ConvModule, build_upsample_layer + + +class UpConvBlock(nn.Module): + """Upsample convolution block in decoder for UNet. + + This upsample convolution block consists of one upsample module + followed by one convolution block. The upsample module expands the + high-level low-resolution feature map and the convolution block fuses + the upsampled high-level low-resolution feature map and the low-level + high-resolution feature map from encoder. + + Args: + conv_block (nn.Sequential): Sequential of convolutional layers. + in_channels (int): Number of input channels of the high-level + skip_channels (int): Number of input channels of the low-level + high-resolution feature map from encoder. + out_channels (int): Number of output channels. + num_convs (int): Number of convolutional layers in the conv_block. + Default: 2. + stride (int): Stride of convolutional layer in conv_block. Default: 1. + dilation (int): Dilation rate of convolutional layer in conv_block. + Default: 1. + with_cp (bool): Use checkpoint or not. Using checkpoint will save some + memory while slowing down the training speed. Default: False. + conv_cfg (dict | None): Config dict for convolution layer. + Default: None. + norm_cfg (dict | None): Config dict for normalization layer. + Default: dict(type='BN'). + act_cfg (dict | None): Config dict for activation layer in ConvModule. + Default: dict(type='ReLU'). + upsample_cfg (dict): The upsample config of the upsample module in + decoder. Default: dict(type='InterpConv'). If the size of + high-level feature map is the same as that of skip feature map + (low-level feature map from encoder), it does not need upsample the + high-level feature map and the upsample_cfg is None. + dcn (bool): Use deformable convolution in convolutional layer or not. + Default: None. + plugins (dict): plugins for convolutional layers. Default: None. + """ + + def __init__(self, + conv_block, + in_channels, + skip_channels, + out_channels, + num_convs=2, + stride=1, + dilation=1, + with_cp=False, + conv_cfg=None, + norm_cfg=dict(type='BN'), + act_cfg=dict(type='ReLU'), + upsample_cfg=dict(type='InterpConv'), + dcn=None, + plugins=None): + super(UpConvBlock, self).__init__() + assert dcn is None, 'Not implemented yet.' + assert plugins is None, 'Not implemented yet.' + + self.conv_block = conv_block( + in_channels=2 * skip_channels, + out_channels=out_channels, + num_convs=num_convs, + stride=stride, + dilation=dilation, + with_cp=with_cp, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg, + dcn=None, + plugins=None) + if upsample_cfg is not None: + self.upsample = build_upsample_layer( + cfg=upsample_cfg, + in_channels=in_channels, + out_channels=skip_channels, + with_cp=with_cp, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + else: + self.upsample = ConvModule( + in_channels, + skip_channels, + kernel_size=1, + stride=1, + padding=0, + conv_cfg=conv_cfg, + norm_cfg=norm_cfg, + act_cfg=act_cfg) + + def forward(self, skip, x): + """Forward function.""" + + x = self.upsample(x) + out = torch.cat([skip, x], dim=1) + out = self.conv_block(out) + + return out diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/wrappers.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/wrappers.py new file mode 100644 index 0000000000000000000000000000000000000000..abbd0c029623b4f480a067e4b78adfec234ef8d0 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/models/utils/wrappers.py @@ -0,0 +1,51 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch.nn as nn +import torch.nn.functional as F + + +def resize(input, + size=None, + scale_factor=None, + mode='nearest', + align_corners=None, + warning=True): + if warning: + if size is not None and align_corners: + input_h, input_w = tuple(int(x) for x in input.shape[2:]) + output_h, output_w = tuple(int(x) for x in size) + if output_h > input_h or output_w > output_h: + if ((output_h > 1 and output_w > 1 and input_h > 1 + and input_w > 1) and (output_h - 1) % (input_h - 1) + and (output_w - 1) % (input_w - 1)): + warnings.warn( + f'When align_corners={align_corners}, ' + 'the output would more aligned if ' + f'input size {(input_h, input_w)} is `x+1` and ' + f'out size {(output_h, output_w)} is `nx+1`') + return F.interpolate(input, size, scale_factor, mode, align_corners) + + +class Upsample(nn.Module): + + def __init__(self, + size=None, + scale_factor=None, + mode='nearest', + align_corners=None): + super().__init__() + self.size = size + if isinstance(scale_factor, tuple): + self.scale_factor = tuple(float(factor) for factor in scale_factor) + else: + self.scale_factor = float(scale_factor) if scale_factor else None + self.mode = mode + self.align_corners = align_corners + + def forward(self, x): + if not self.size: + size = [int(t * self.scale_factor) for t in x.shape[-2:]] + else: + size = self.size + return resize(x, size, None, self.mode, self.align_corners) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..bc075cd4eb7af7a8d2ad146233d9d5973e7f036d --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/__init__.py @@ -0,0 +1,5 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .encoding import Encoding +from .wrappers import Upsample, resize + +__all__ = ['Upsample', 'resize', 'Encoding'] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/encoding.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/encoding.py new file mode 100644 index 0000000000000000000000000000000000000000..f397cc54e84feb6a9979c767d8602601d5613542 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/encoding.py @@ -0,0 +1,75 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import torch +from torch import nn +from torch.nn import functional as F + + +class Encoding(nn.Module): + """Encoding Layer: a learnable residual encoder. + + Input is of shape (batch_size, channels, height, width). + Output is of shape (batch_size, num_codes, channels). + + Args: + channels: dimension of the features or feature channels + num_codes: number of code words + """ + + def __init__(self, channels, num_codes): + super(Encoding, self).__init__() + # init codewords and smoothing factor + self.channels, self.num_codes = channels, num_codes + std = 1. / ((num_codes * channels)**0.5) + # [num_codes, channels] + self.codewords = nn.Parameter( + torch.empty(num_codes, channels, + dtype=torch.float).uniform_(-std, std), + requires_grad=True) + # [num_codes] + self.scale = nn.Parameter( + torch.empty(num_codes, dtype=torch.float).uniform_(-1, 0), + requires_grad=True) + + @staticmethod + def scaled_l2(x, codewords, scale): + num_codes, channels = codewords.size() + batch_size = x.size(0) + reshaped_scale = scale.view((1, 1, num_codes)) + expanded_x = x.unsqueeze(2).expand( + (batch_size, x.size(1), num_codes, channels)) + reshaped_codewords = codewords.view((1, 1, num_codes, channels)) + + scaled_l2_norm = reshaped_scale * ( + expanded_x - reshaped_codewords).pow(2).sum(dim=3) + return scaled_l2_norm + + @staticmethod + def aggregate(assignment_weights, x, codewords): + num_codes, channels = codewords.size() + reshaped_codewords = codewords.view((1, 1, num_codes, channels)) + batch_size = x.size(0) + + expanded_x = x.unsqueeze(2).expand( + (batch_size, x.size(1), num_codes, channels)) + encoded_feat = (assignment_weights.unsqueeze(3) * + (expanded_x - reshaped_codewords)).sum(dim=1) + return encoded_feat + + def forward(self, x): + assert x.dim() == 4 and x.size(1) == self.channels + # [batch_size, channels, height, width] + batch_size = x.size(0) + # [batch_size, height x width, channels] + x = x.view(batch_size, self.channels, -1).transpose(1, 2).contiguous() + # assignment_weights: [batch_size, channels, num_codes] + assignment_weights = F.softmax( + self.scaled_l2(x, self.codewords, self.scale), dim=2) + # aggregate + encoded_feat = self.aggregate(assignment_weights, x, self.codewords) + return encoded_feat + + def __repr__(self): + repr_str = self.__class__.__name__ + repr_str += f'(Nx{self.channels}xHxW =>Nx{self.num_codes}' \ + f'x{self.channels})' + return repr_str diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/wrappers.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/wrappers.py new file mode 100644 index 0000000000000000000000000000000000000000..ce67e4bebe1ed463072858f97dd950e596ca6a28 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/ops/wrappers.py @@ -0,0 +1,51 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import warnings + +import torch.nn as nn +import torch.nn.functional as F + + +def resize(input, + size=None, + scale_factor=None, + mode='nearest', + align_corners=None, + warning=True): + if warning: + if size is not None and align_corners: + input_h, input_w = tuple(int(x) for x in input.shape[2:]) + output_h, output_w = tuple(int(x) for x in size) + if output_h > input_h or output_w > output_h: + if ((output_h > 1 and output_w > 1 and input_h > 1 + and input_w > 1) and (output_h - 1) % (input_h - 1) + and (output_w - 1) % (input_w - 1)): + warnings.warn( + f'When align_corners={align_corners}, ' + 'the output would more aligned if ' + f'input size {(input_h, input_w)} is `x+1` and ' + f'out size {(output_h, output_w)} is `nx+1`') + return F.interpolate(input, size, scale_factor, mode, align_corners) + + +class Upsample(nn.Module): + + def __init__(self, + size=None, + scale_factor=None, + mode='nearest', + align_corners=None): + super(Upsample, self).__init__() + self.size = size + if isinstance(scale_factor, tuple): + self.scale_factor = tuple(float(factor) for factor in scale_factor) + else: + self.scale_factor = float(scale_factor) if scale_factor else None + self.mode = mode + self.align_corners = align_corners + + def forward(self, x): + if not self.size: + size = [int(t * self.scale_factor) for t in x.shape[-2:]] + else: + size = self.size + return resize(x, size, None, self.mode, self.align_corners) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/registry/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/registry/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..ee514d1a2a2bdd54a0a9b017ec227160ee502be5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/registry/__init__.py @@ -0,0 +1,15 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .registry import (DATA_SAMPLERS, DATASETS, EVALUATOR, HOOKS, INFERENCERS, + LOG_PROCESSORS, LOOPS, METRICS, MODEL_WRAPPERS, MODELS, + OPTIM_WRAPPER_CONSTRUCTORS, OPTIM_WRAPPERS, OPTIMIZERS, + PARAM_SCHEDULERS, RUNNER_CONSTRUCTORS, RUNNERS, + TASK_UTILS, TRANSFORMS, VISBACKENDS, VISUALIZERS, + WEIGHT_INITIALIZERS) + +__all__ = [ + 'HOOKS', 'DATASETS', 'DATA_SAMPLERS', 'TRANSFORMS', 'MODELS', + 'WEIGHT_INITIALIZERS', 'OPTIMIZERS', 'OPTIM_WRAPPER_CONSTRUCTORS', + 'TASK_UTILS', 'PARAM_SCHEDULERS', 'METRICS', 'MODEL_WRAPPERS', + 'VISBACKENDS', 'VISUALIZERS', 'RUNNERS', 'RUNNER_CONSTRUCTORS', 'LOOPS', + 'EVALUATOR', 'LOG_PROCESSORS', 'OPTIM_WRAPPERS', 'INFERENCERS' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/registry/registry.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/registry/registry.py new file mode 100644 index 0000000000000000000000000000000000000000..1e423980d1a450a2cda5cd2baad5513e05382fc5 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/registry/registry.py @@ -0,0 +1,116 @@ +# Copyright (c) OpenMMLab. All rights reserved. +"""MMSegmentation provides 21 registry nodes to support using modules across +projects. Each node is a child of the root registry in MMEngine. + +More details can be found at +https://mmengine.readthedocs.io/en/latest/advanced_tutorials/registry.html. +""" + +from mmengine.registry import DATA_SAMPLERS as MMENGINE_DATA_SAMPLERS +from mmengine.registry import DATASETS as MMENGINE_DATASETS +from mmengine.registry import EVALUATOR as MMENGINE_EVALUATOR +from mmengine.registry import HOOKS as MMENGINE_HOOKS +from mmengine.registry import INFERENCERS as MMENGINE_INFERENCERS +from mmengine.registry import LOG_PROCESSORS as MMENGINE_LOG_PROCESSORS +from mmengine.registry import LOOPS as MMENGINE_LOOPS +from mmengine.registry import METRICS as MMENGINE_METRICS +from mmengine.registry import MODEL_WRAPPERS as MMENGINE_MODEL_WRAPPERS +from mmengine.registry import MODELS as MMENGINE_MODELS +from mmengine.registry import \ + OPTIM_WRAPPER_CONSTRUCTORS as MMENGINE_OPTIM_WRAPPER_CONSTRUCTORS +from mmengine.registry import OPTIM_WRAPPERS as MMENGINE_OPTIM_WRAPPERS +from mmengine.registry import OPTIMIZERS as MMENGINE_OPTIMIZERS +from mmengine.registry import PARAM_SCHEDULERS as MMENGINE_PARAM_SCHEDULERS +from mmengine.registry import \ + RUNNER_CONSTRUCTORS as MMENGINE_RUNNER_CONSTRUCTORS +from mmengine.registry import RUNNERS as MMENGINE_RUNNERS +from mmengine.registry import TASK_UTILS as MMENGINE_TASK_UTILS +from mmengine.registry import TRANSFORMS as MMENGINE_TRANSFORMS +from mmengine.registry import VISBACKENDS as MMENGINE_VISBACKENDS +from mmengine.registry import VISUALIZERS as MMENGINE_VISUALIZERS +from mmengine.registry import \ + WEIGHT_INITIALIZERS as MMENGINE_WEIGHT_INITIALIZERS +from mmengine.registry import Registry + +# manage all kinds of runners like `EpochBasedRunner` and `IterBasedRunner` +RUNNERS = Registry('runner', parent=MMENGINE_RUNNERS) +# manage runner constructors that define how to initialize runners +RUNNER_CONSTRUCTORS = Registry( + 'runner constructor', parent=MMENGINE_RUNNER_CONSTRUCTORS) +# manage all kinds of loops like `EpochBasedTrainLoop` +LOOPS = Registry('loop', parent=MMENGINE_LOOPS) +# manage all kinds of hooks like `CheckpointHook` +HOOKS = Registry( + 'hook', parent=MMENGINE_HOOKS, locations=['mmseg.engine.hooks']) + +# manage data-related modules +DATASETS = Registry( + 'dataset', parent=MMENGINE_DATASETS, locations=['mmseg.datasets']) +DATA_SAMPLERS = Registry('data sampler', parent=MMENGINE_DATA_SAMPLERS) +TRANSFORMS = Registry( + 'transform', + parent=MMENGINE_TRANSFORMS, + locations=['mmseg.datasets.transforms']) + +# mangage all kinds of modules inheriting `nn.Module` +MODELS = Registry('model', parent=MMENGINE_MODELS, locations=['mmseg.models']) +# mangage all kinds of model wrappers like 'MMDistributedDataParallel' +MODEL_WRAPPERS = Registry( + 'model_wrapper', + parent=MMENGINE_MODEL_WRAPPERS, + locations=['mmseg.models']) +# mangage all kinds of weight initialization modules like `Uniform` +WEIGHT_INITIALIZERS = Registry( + 'weight initializer', + parent=MMENGINE_WEIGHT_INITIALIZERS, + locations=['mmseg.models']) + +# mangage all kinds of optimizers like `SGD` and `Adam` +OPTIMIZERS = Registry( + 'optimizer', + parent=MMENGINE_OPTIMIZERS, + locations=['mmseg.engine.optimizers']) +# manage optimizer wrapper +OPTIM_WRAPPERS = Registry( + 'optim_wrapper', + parent=MMENGINE_OPTIM_WRAPPERS, + locations=['mmseg.engine.optimizers']) +# manage constructors that customize the optimization hyperparameters. +OPTIM_WRAPPER_CONSTRUCTORS = Registry( + 'optimizer wrapper constructor', + parent=MMENGINE_OPTIM_WRAPPER_CONSTRUCTORS, + locations=['mmseg.engine.optimizers']) +# mangage all kinds of parameter schedulers like `MultiStepLR` +PARAM_SCHEDULERS = Registry( + 'parameter scheduler', parent=MMENGINE_PARAM_SCHEDULERS) + +# manage all kinds of metrics +METRICS = Registry( + 'metric', parent=MMENGINE_METRICS, locations=['mmseg.evaluation']) +# manage evaluator +EVALUATOR = Registry( + 'evaluator', parent=MMENGINE_EVALUATOR, locations=['mmseg.evaluation']) + +# manage task-specific modules like ohem pixel sampler +TASK_UTILS = Registry( + 'task util', parent=MMENGINE_TASK_UTILS, locations=['mmseg.models']) + +# manage visualizer +VISUALIZERS = Registry( + 'visualizer', + parent=MMENGINE_VISUALIZERS, + locations=['mmseg.visualization']) +# manage visualizer backend +VISBACKENDS = Registry( + 'vis_backend', + parent=MMENGINE_VISBACKENDS, + locations=['mmseg.visualization']) + +# manage logprocessor +LOG_PROCESSORS = Registry( + 'log_processor', + parent=MMENGINE_LOG_PROCESSORS, + locations=['mmseg.visualization']) + +# manage inferencer +INFERENCERS = Registry('inferencer', parent=MMENGINE_INFERENCERS) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/__init__.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..ed002c7de4436bd95947a762a5916cd44ff67128 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/__init__.py @@ -0,0 +1,10 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from .collect_env import collect_env +from .logger import get_root_logger +from .misc import find_latest_checkpoint +from .set_env import setup_multi_processes + +__all__ = [ + 'get_root_logger', 'collect_env', 'find_latest_checkpoint', + 'setup_multi_processes' +] diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/collect_env.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/collect_env.py new file mode 100644 index 0000000000000000000000000000000000000000..3379ecb06bb059f2694c7ce9e1f10fa6c4f938b9 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/collect_env.py @@ -0,0 +1,18 @@ +# Copyright (c) OpenMMLab. All rights reserved. +from mmcv.utils import collect_env as collect_base_env +from mmcv.utils import get_git_hash + +import mmseg + + +def collect_env(): + """Collect the information of the running environments.""" + env_info = collect_base_env() + env_info['MMSegmentation'] = f'{mmseg.__version__}+{get_git_hash()[:7]}' + + return env_info + + +if __name__ == '__main__': + for name, val in collect_env().items(): + print('{}: {}'.format(name, val)) diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/logger.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/logger.py new file mode 100644 index 0000000000000000000000000000000000000000..0cb3c78d6da0d6c850a4603a255ecdea93b15d2b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/logger.py @@ -0,0 +1,28 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import logging + +from mmcv.utils import get_logger + + +def get_root_logger(log_file=None, log_level=logging.INFO): + """Get the root logger. + + The logger will be initialized if it has not been initialized. By default a + StreamHandler will be added. If `log_file` is specified, a FileHandler will + also be added. The name of the root logger is the top-level package name, + e.g., "mmseg". + + Args: + log_file (str | None): The log filename. If specified, a FileHandler + will be added to the root logger. + log_level (int): The root logger level. Note that only the process of + rank 0 is affected, while other processes will set the level to + "Error" and be silent most of the time. + + Returns: + logging.Logger: The root logger. + """ + + logger = get_logger(name='mmseg', log_file=log_file, log_level=log_level) + + return logger diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/misc.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/misc.py new file mode 100644 index 0000000000000000000000000000000000000000..bd1b6b163c21ed8ef071a97f42b7fd8997e78215 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/misc.py @@ -0,0 +1,41 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import glob +import os.path as osp +import warnings + + +def find_latest_checkpoint(path, suffix='pth'): + """This function is for finding the latest checkpoint. + + It will be used when automatically resume, modified from + https://github.com/open-mmlab/mmdetection/blob/dev-v2.20.0/mmdet/utils/misc.py + + Args: + path (str): The path to find checkpoints. + suffix (str): File extension for the checkpoint. Defaults to pth. + + Returns: + latest_path(str | None): File path of the latest checkpoint. + """ + if not osp.exists(path): + warnings.warn("The path of the checkpoints doesn't exist.") + return None + if osp.exists(osp.join(path, f'latest.{suffix}')): + return osp.join(path, f'latest.{suffix}') + + checkpoints = glob.glob(osp.join(path, f'*.{suffix}')) + if len(checkpoints) == 0: + warnings.warn('The are no checkpoints in the path') + return None + latest = -1 + latest_path = '' + for checkpoint in checkpoints: + if len(checkpoint) < len(latest_path): + continue + # `count` is iteration number, as checkpoints are saved as + # 'iter_xx.pth' or 'epoch_xx.pth' and xx is iteration number. + count = int(osp.basename(checkpoint).split('_')[-1].split('.')[0]) + if count > latest: + latest = count + latest_path = checkpoint + return latest_path diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/set_env.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/set_env.py new file mode 100644 index 0000000000000000000000000000000000000000..b2d3aaf14b2f9706a561fb1af3129126a9647592 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/utils/set_env.py @@ -0,0 +1,55 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import os +import platform + +import cv2 +import torch.multiprocessing as mp + +from ..utils import get_root_logger + + +def setup_multi_processes(cfg): + """Setup multi-processing environment variables.""" + logger = get_root_logger() + + # set multi-process start method + if platform.system() != 'Windows': + mp_start_method = cfg.get('mp_start_method', None) + current_method = mp.get_start_method(allow_none=True) + if mp_start_method in ('fork', 'spawn', 'forkserver'): + logger.info( + f'Multi-processing start method `{mp_start_method}` is ' + f'different from the previous setting `{current_method}`.' + f'It will be force set to `{mp_start_method}`.') + mp.set_start_method(mp_start_method, force=True) + else: + logger.info( + f'Multi-processing start method is `{mp_start_method}`') + + # disable opencv multithreading to avoid system being overloaded + opencv_num_threads = cfg.get('opencv_num_threads', None) + if isinstance(opencv_num_threads, int): + logger.info(f'OpenCV num_threads is `{opencv_num_threads}`') + cv2.setNumThreads(opencv_num_threads) + else: + logger.info(f'OpenCV num_threads is `{cv2.getNumThreads}') + + if cfg.data.workers_per_gpu > 1: + # setup OMP threads + # This code is referred from https://github.com/pytorch/pytorch/blob/master/torch/distributed/run.py # noqa + omp_num_threads = cfg.get('omp_num_threads', None) + if 'OMP_NUM_THREADS' not in os.environ: + if isinstance(omp_num_threads, int): + logger.info(f'OMP num threads is {omp_num_threads}') + os.environ['OMP_NUM_THREADS'] = str(omp_num_threads) + else: + logger.info(f'OMP num threads is {os.environ["OMP_NUM_THREADS"] }') + + # setup MKL threads + if 'MKL_NUM_THREADS' not in os.environ: + mkl_num_threads = cfg.get('mkl_num_threads', None) + if isinstance(mkl_num_threads, int): + logger.info(f'MKL num threads is {mkl_num_threads}') + os.environ['MKL_NUM_THREADS'] = str(mkl_num_threads) + else: + logger.info(f'MKL num threads is {os.environ["MKL_NUM_THREADS"]}') diff --git a/cv/semantic_segmentation/att_unet/pytorch/mmseg/version.py b/cv/semantic_segmentation/att_unet/pytorch/mmseg/version.py new file mode 100644 index 0000000000000000000000000000000000000000..e05146f0a07709d5f51efed652b4cb9af900c0d2 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/mmseg/version.py @@ -0,0 +1,18 @@ +# Copyright (c) Open-MMLab. All rights reserved. + +__version__ = '0.24.1' + + +def parse_version_info(version_str): + version_info = [] + for x in version_str.split('.'): + if x.isdigit(): + version_info.append(int(x)) + elif x.find('rc') != -1: + patch_version = x.split('rc') + version_info.append(int(patch_version[0])) + version_info.append(f'rc{patch_version[1]}') + return tuple(version_info) + + +version_info = parse_version_info(__version__) diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..a50fc71db2a2131dc8907faae06b1591e61530dc --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements.txt @@ -0,0 +1,4 @@ +-r requirements/mmcv/runtime.txt +-r requirements/apcnet.txt +opencv-python +cityscapesscripts \ No newline at end of file diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/apcnet.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/apcnet.txt new file mode 100644 index 0000000000000000000000000000000000000000..2712f504c7836788ab022e2e70b596bc9f94e753 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/apcnet.txt @@ -0,0 +1,4 @@ +matplotlib +numpy +packaging +prettytable diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/build.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/build.txt new file mode 100644 index 0000000000000000000000000000000000000000..abf514853e58db1b0903721c7624cb313bf3aa57 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/build.txt @@ -0,0 +1 @@ +pytest-runner diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/docs.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/docs.txt new file mode 100644 index 0000000000000000000000000000000000000000..6a5319af3a33c9ee0a7caa63e381be1400a7ba99 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/docs.txt @@ -0,0 +1,8 @@ +docutils==0.16.0 +myst-parser +opencv-python +-e git+https://github.com/open-mmlab/pytorch_sphinx_theme.git#egg=pytorch_sphinx_theme +sphinx==4.0.2 +sphinx-copybutton +sphinx_markdown_tables +torch diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/optional.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/optional.txt new file mode 100644 index 0000000000000000000000000000000000000000..63730036fd34e349d7856c5401d395262f99db16 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/optional.txt @@ -0,0 +1 @@ +ninja diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/requirements.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..6f00e6303e23466bc99b4fa8c7549c6f224b3b25 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/requirements.txt @@ -0,0 +1,4 @@ +-r requirements/mmcv/build.txt +-r requirements/mmcv/optional.txt +-r requirements/mmcv/runtime.txt +-r requirements/mmcv/test.txt diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/runtime.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/runtime.txt new file mode 100644 index 0000000000000000000000000000000000000000..66e90d6748971e14a852a7008d9301aea262606c --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/runtime.txt @@ -0,0 +1,7 @@ +addict +numpy +packaging +Pillow +pyyaml +regex;sys_platform=='win32' +yapf diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/test.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/test.txt new file mode 100644 index 0000000000000000000000000000000000000000..6d9d17b98e1e9f193a82ef1fb2cc34b32571c925 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/mmcv/test.txt @@ -0,0 +1,9 @@ +coverage +lmdb +onnx==1.7.0; python_version < '3.10' +onnxoptimizer; python_version < '3.10' +onnxruntime>=1.8.0; python_version < '3.10' +pytest +PyTurboJPEG +scipy +tifffile diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/mminstall.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/mminstall.txt new file mode 100644 index 0000000000000000000000000000000000000000..bd43faf87e177562f7b45fc0c6c3dbbf694794e8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/mminstall.txt @@ -0,0 +1,2 @@ +mmcls>=0.20.1 +mmcv-full>=1.4.4,<=1.6.0 diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/optional.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/optional.txt new file mode 100644 index 0000000000000000000000000000000000000000..47fa5933159eb068640f6b45281d36ab4af294cb --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/optional.txt @@ -0,0 +1 @@ +cityscapesscripts diff --git a/cv/semantic_segmentation/att_unet/pytorch/requirements/runtime.txt b/cv/semantic_segmentation/att_unet/pytorch/requirements/runtime.txt new file mode 100644 index 0000000000000000000000000000000000000000..520408fe8b322c0cb288160014e4e996ffa8f457 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/requirements/runtime.txt @@ -0,0 +1,5 @@ +matplotlib +mmcls>=0.20.1 +numpy +packaging +prettytable diff --git a/cv/semantic_segmentation/att_unet/pytorch/setup.py b/cv/semantic_segmentation/att_unet/pytorch/setup.py new file mode 100644 index 0000000000000000000000000000000000000000..d409997dbca22e6abee8bc9e1198f565bed320cf --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/setup.py @@ -0,0 +1,258 @@ +import glob +import os +import platform +import re +import warnings +from pkg_resources import DistributionNotFound, get_distribution +from setuptools import find_packages, setup + +EXT_TYPE = '' +try: + import torch + if torch.__version__ == 'parrots': + from parrots.utils.build_extension import BuildExtension + EXT_TYPE = 'parrots' + elif (hasattr(torch, 'is_mlu_available') and torch.is_mlu_available()) or \ + os.getenv('FORCE_MLU', '0') == '1': + from torch_mlu.utils.cpp_extension import BuildExtension + EXT_TYPE = 'pytorch' + else: + from torch.utils.cpp_extension import BuildExtension + EXT_TYPE = 'pytorch' + cmd_class = {'build_ext': BuildExtension} +except ModuleNotFoundError: + cmd_class = {} + print('Skip building ext ops due to the absence of torch.') + + +def choose_requirement(primary, secondary): + """If some version of primary requirement installed, return primary, else + return secondary.""" + try: + name = re.split(r'[!<>=]', primary)[0] + get_distribution(name) + except DistributionNotFound: + return secondary + + return str(primary) + + +def get_version(): + version_file = 'mmcv/version.py' + with open(version_file, 'r', encoding='utf-8') as f: + exec(compile(f.read(), version_file, 'exec')) + return locals()['__version__'] + + +def parse_requirements(fname='requirements/mmcv/runtime.txt', with_version=True): + """Parse the package dependencies listed in a requirements file but strips + specific versioning information. + + Args: + fname (str): path to requirements file + with_version (bool, default=False): if True include version specs + + Returns: + List[str]: list of requirements items + + CommandLine: + python -c "import setup; print(setup.parse_requirements())" + """ + import sys + from os.path import exists + require_fpath = fname + + def parse_line(line): + """Parse information from a line in a requirements text file.""" + if line.startswith('-r '): + # Allow specifying requirements in other files + target = line.split(' ')[1] + for info in parse_require_file(target): + yield info + else: + info = {'line': line} + if line.startswith('-e '): + info['package'] = line.split('#egg=')[1] + else: + # Remove versioning from the package + pat = '(' + '|'.join(['>=', '==', '>']) + ')' + parts = re.split(pat, line, maxsplit=1) + parts = [p.strip() for p in parts] + + info['package'] = parts[0] + if len(parts) > 1: + op, rest = parts[1:] + if ';' in rest: + # Handle platform specific dependencies + # http://setuptools.readthedocs.io/en/latest/setuptools.html#declaring-platform-specific-dependencies + version, platform_deps = map(str.strip, + rest.split(';')) + info['platform_deps'] = platform_deps + else: + version = rest # NOQA + info['version'] = (op, version) + yield info + + def parse_require_file(fpath): + with open(fpath, 'r') as f: + for line in f.readlines(): + line = line.strip() + if line and not line.startswith('#'): + for info in parse_line(line): + yield info + + def gen_packages_items(): + if exists(require_fpath): + for info in parse_require_file(require_fpath): + parts = [info['package']] + if with_version and 'version' in info: + parts.extend(info['version']) + if not sys.version.startswith('3.4'): + # apparently package_deps are broken in 3.4 + platform_deps = info.get('platform_deps') + if platform_deps is not None: + parts.append(';' + platform_deps) + item = ''.join(parts) + yield item + + packages = list(gen_packages_items()) + return packages + + +install_requires = parse_requirements() + +try: + # OpenCV installed via conda. + import cv2 # NOQA: F401 + major, minor, *rest = cv2.__version__.split('.') + if int(major) < 3: + raise RuntimeError( + f'OpenCV >=3 is required but {cv2.__version__} is installed') +except ImportError: + # If first not installed install second package + CHOOSE_INSTALL_REQUIRES = [('opencv-python-headless>=3', + 'opencv-python>=3')] + for main, secondary in CHOOSE_INSTALL_REQUIRES: + install_requires.append(choose_requirement(main, secondary)) + + +def get_extensions(): + extensions = [] + + if EXT_TYPE == 'pytorch': + ext_name = 'mmcv._ext' + from torch.utils.cpp_extension import CppExtension, CUDAExtension + + # prevent ninja from using too many resources + try: + import psutil + num_cpu = len(psutil.Process().cpu_affinity()) + cpu_use = max(4, num_cpu - 1) + except (ModuleNotFoundError, AttributeError): + cpu_use = 4 + + os.environ.setdefault('MAX_JOBS', str(cpu_use)) + define_macros = [] + + # Before PyTorch1.8.0, when compiling CUDA code, `cxx` is a + # required key passed to PyTorch. Even if there is no flag passed + # to cxx, users also need to pass an empty list to PyTorch. + # Since PyTorch1.8.0, it has a default value so users do not need + # to pass an empty list anymore. + # More details at https://github.com/pytorch/pytorch/pull/45956 + extra_compile_args = {'cxx': []} + + # Since the PR (https://github.com/open-mmlab/mmcv/pull/1463) uses + # c++14 features, the argument ['std=c++14'] must be added here. + # However, in the windows environment, some standard libraries + # will depend on c++17 or higher. In fact, for the windows + # environment, the compiler will choose the appropriate compiler + # to compile those cpp files, so there is no need to add the + # argument + if platform.system() != 'Windows': + extra_compile_args['cxx'] = ['-std=c++14'] + + include_dirs = [] + + is_rocm_pytorch = False + try: + from torch.utils.cpp_extension import ROCM_HOME + is_rocm_pytorch = True if ((torch.version.hip is not None) and + (ROCM_HOME is not None)) else False + except ImportError: + pass + + if is_rocm_pytorch or torch.cuda.is_available() or os.getenv( + 'FORCE_CUDA', '0') == '1': + if is_rocm_pytorch: + define_macros += [('HIP_DIFF', None)] + define_macros += [('MMCV_WITH_CUDA', None)] + cuda_args = os.getenv('MMCV_CUDA_ARGS') + extra_compile_args['nvcc'] = [cuda_args] if cuda_args else [] + op_files = glob.glob('./mmcv/ops/csrc/pytorch/*.cpp') + \ + glob.glob('./mmcv/ops/csrc/pytorch/cpu/*.cpp') + \ + glob.glob('./mmcv/ops/csrc/pytorch/cuda/*.cu') + \ + glob.glob('./mmcv/ops/csrc/pytorch/cuda/*.cpp') + extension = CUDAExtension + include_dirs.append(os.path.abspath('./mmcv/ops/csrc/common')) + include_dirs.append(os.path.abspath('./mmcv/ops/csrc/common/cuda')) + else: + print(f'Compiling {ext_name} only with CPU') + op_files = glob.glob('./mmcv/ops/csrc/pytorch/*.cpp') + \ + glob.glob('./mmcv/ops/csrc/pytorch/cpu/*.cpp') + extension = CppExtension + include_dirs.append(os.path.abspath('./mmcv/ops/csrc/common')) + + # Since the PR (https://github.com/open-mmlab/mmcv/pull/1463) uses + # c++14 features, the argument ['std=c++14'] must be added here. + # However, in the windows environment, some standard libraries + # will depend on c++17 or higher. In fact, for the windows + # environment, the compiler will choose the appropriate compiler + # to compile those cpp files, so there is no need to add the + # argument + if 'nvcc' in extra_compile_args and platform.system() != 'Windows': + extra_compile_args['nvcc'] += ['-std=c++14'] + + ext_ops = extension( + name=ext_name, + sources=op_files, + include_dirs=include_dirs, + define_macros=define_macros, + extra_compile_args=extra_compile_args) + extensions.append(ext_ops) + + return extensions + + +setup( + name='mmcv' if os.getenv('MMCV_WITH_OPS', '0') == '0' else 'mmcv-full', + version=get_version(), + description='OpenMMLab Computer Vision Foundation', + keywords='computer vision', + packages=find_packages(), + include_package_data=True, + classifiers=[ + 'Development Status :: 4 - Beta', + 'License :: OSI Approved :: Apache Software License', + 'Operating System :: OS Independent', + 'Programming Language :: Python :: 3', + 'Programming Language :: Python :: 3.6', + 'Programming Language :: Python :: 3.7', + 'Programming Language :: Python :: 3.8', + 'Programming Language :: Python :: 3.9', + 'Programming Language :: Python :: 3.10', + 'Topic :: Utilities', + ], + url='https://github.com/open-mmlab/mmcv', + author='MMCV Contributors', + author_email='openmmlab@gmail.com', + install_requires=install_requires, + extras_require={ + 'all': parse_requirements('requirements/mmcv/requirements.txt'), + 'tests': parse_requirements('requirements/mmcv/test.txt'), + 'build': parse_requirements('requirements/mmcv/build.txt'), + 'optional': parse_requirements('requirements/mmcv/optional.txt'), + }, + ext_modules=get_extensions(), + cmdclass=cmd_class, + zip_safe=False) diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/analyze_logs.py b/cv/semantic_segmentation/att_unet/pytorch/tools/analyze_logs.py new file mode 100644 index 0000000000000000000000000000000000000000..e2127d4d627a1044f2f2b3719c38fda7bc8c928f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/analyze_logs.py @@ -0,0 +1,128 @@ +# Copyright (c) OpenMMLab. All rights reserved. +"""Modified from https://github.com/open- +mmlab/mmdetection/blob/master/tools/analysis_tools/analyze_logs.py.""" +import argparse +import json +from collections import defaultdict + +import matplotlib.pyplot as plt +import seaborn as sns + + +def plot_curve(log_dicts, args): + if args.backend is not None: + plt.switch_backend(args.backend) + sns.set_style(args.style) + # if legend is None, use {filename}_{key} as legend + legend = args.legend + if legend is None: + legend = [] + for json_log in args.json_logs: + for metric in args.keys: + legend.append(f'{json_log}_{metric}') + assert len(legend) == (len(args.json_logs) * len(args.keys)) + metrics = args.keys + + num_metrics = len(metrics) + for i, log_dict in enumerate(log_dicts): + epochs = list(log_dict.keys()) + for j, metric in enumerate(metrics): + print(f'plot curve of {args.json_logs[i]}, metric is {metric}') + plot_epochs = [] + plot_iters = [] + plot_values = [] + # In some log files exist lines of validation, + # `mode` list is used to only collect iter number + # of training line. + for epoch in epochs: + epoch_logs = log_dict[epoch] + if metric not in epoch_logs.keys(): + continue + if metric in ['mIoU', 'mAcc', 'aAcc']: + plot_epochs.append(epoch) + plot_values.append(epoch_logs[metric][0]) + else: + for idx in range(len(epoch_logs[metric])): + if epoch_logs['mode'][idx] == 'train': + plot_iters.append(epoch_logs['iter'][idx]) + plot_values.append(epoch_logs[metric][idx]) + ax = plt.gca() + label = legend[i * num_metrics + j] + if metric in ['mIoU', 'mAcc', 'aAcc']: + ax.set_xticks(plot_epochs) + plt.xlabel('epoch') + plt.plot(plot_epochs, plot_values, label=label, marker='o') + else: + plt.xlabel('iter') + plt.plot(plot_iters, plot_values, label=label, linewidth=0.5) + plt.legend() + if args.title is not None: + plt.title(args.title) + if args.out is None: + plt.show() + else: + print(f'save curve to: {args.out}') + plt.savefig(args.out) + plt.cla() + + +def parse_args(): + parser = argparse.ArgumentParser(description='Analyze Json Log') + parser.add_argument( + 'json_logs', + type=str, + nargs='+', + help='path of train log in json format') + parser.add_argument( + '--keys', + type=str, + nargs='+', + default=['mIoU'], + help='the metric that you want to plot') + parser.add_argument('--title', type=str, help='title of figure') + parser.add_argument( + '--legend', + type=str, + nargs='+', + default=None, + help='legend of each plot') + parser.add_argument( + '--backend', type=str, default=None, help='backend of plt') + parser.add_argument( + '--style', type=str, default='dark', help='style of plt') + parser.add_argument('--out', type=str, default=None) + args = parser.parse_args() + return args + + +def load_json_logs(json_logs): + # load and convert json_logs to log_dict, key is epoch, value is a sub dict + # keys of sub dict is different metrics + # value of sub dict is a list of corresponding values of all iterations + log_dicts = [dict() for _ in json_logs] + for json_log, log_dict in zip(json_logs, log_dicts): + with open(json_log, 'r') as log_file: + for line in log_file: + log = json.loads(line.strip()) + # skip lines without `epoch` field + if 'epoch' not in log: + continue + epoch = log.pop('epoch') + if epoch not in log_dict: + log_dict[epoch] = defaultdict(list) + for k, v in log.items(): + log_dict[epoch][k].append(v) + return log_dicts + + +def main(): + args = parse_args() + json_logs = args.json_logs + for json_log in json_logs: + assert json_log.endswith('.json') + log_dicts = load_json_logs(json_logs) + plot_curve(log_dicts, args) + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/benchmark.py b/cv/semantic_segmentation/att_unet/pytorch/tools/benchmark.py new file mode 100644 index 0000000000000000000000000000000000000000..f6d6888482a1f389f2dbf0e6c3d4e58542229b71 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/benchmark.py @@ -0,0 +1,120 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os.path as osp +import time + +import mmcv +import numpy as np +import torch +from mmcv import Config +from mmcv.parallel import MMDataParallel +from mmcv.runner import load_checkpoint, wrap_fp16_model + +from mmseg.datasets import build_dataloader, build_dataset +from mmseg.models import build_segmentor + + +def parse_args(): + parser = argparse.ArgumentParser(description='MMSeg benchmark a model') + parser.add_argument('config', help='test config file path') + parser.add_argument('checkpoint', help='checkpoint file') + parser.add_argument( + '--log-interval', type=int, default=50, help='interval of logging') + parser.add_argument( + '--work-dir', + help=('if specified, the results will be dumped ' + 'into the directory as json')) + parser.add_argument('--repeat-times', type=int, default=1) + args = parser.parse_args() + return args + + +def main(): + args = parse_args() + + cfg = Config.fromfile(args.config) + timestamp = time.strftime('%Y%m%d_%H%M%S', time.localtime()) + if args.work_dir is not None: + mmcv.mkdir_or_exist(osp.abspath(args.work_dir)) + json_file = osp.join(args.work_dir, f'fps_{timestamp}.json') + else: + # use config filename as default work_dir if cfg.work_dir is None + work_dir = osp.join('./work_dirs', + osp.splitext(osp.basename(args.config))[0]) + mmcv.mkdir_or_exist(osp.abspath(work_dir)) + json_file = osp.join(work_dir, f'fps_{timestamp}.json') + + repeat_times = args.repeat_times + # set cudnn_benchmark + torch.backends.cudnn.benchmark = False + cfg.model.pretrained = None + cfg.data.test.test_mode = True + + benchmark_dict = dict(config=args.config, unit='img / s') + overall_fps_list = [] + for time_index in range(repeat_times): + print(f'Run {time_index + 1}:') + # build the dataloader + # TODO: support multiple images per gpu (only minor changes are needed) + dataset = build_dataset(cfg.data.test) + data_loader = build_dataloader( + dataset, + samples_per_gpu=1, + workers_per_gpu=cfg.data.workers_per_gpu, + dist=False, + shuffle=False) + + # build the model and load checkpoint + cfg.model.train_cfg = None + model = build_segmentor(cfg.model, test_cfg=cfg.get('test_cfg')) + fp16_cfg = cfg.get('fp16', None) + if fp16_cfg is not None: + wrap_fp16_model(model) + if 'checkpoint' in args and osp.exists(args.checkpoint): + load_checkpoint(model, args.checkpoint, map_location='cpu') + + model = MMDataParallel(model, device_ids=[0]) + + model.eval() + + # the first several iterations may be very slow so skip them + num_warmup = 5 + pure_inf_time = 0 + total_iters = 200 + + # benchmark with 200 image and take the average + for i, data in enumerate(data_loader): + + torch.cuda.synchronize() + start_time = time.perf_counter() + + with torch.no_grad(): + model(return_loss=False, rescale=True, **data) + + torch.cuda.synchronize() + elapsed = time.perf_counter() - start_time + + if i >= num_warmup: + pure_inf_time += elapsed + if (i + 1) % args.log_interval == 0: + fps = (i + 1 - num_warmup) / pure_inf_time + print(f'Done image [{i + 1:<3}/ {total_iters}], ' + f'fps: {fps:.2f} img / s') + + if (i + 1) == total_iters: + fps = (i + 1 - num_warmup) / pure_inf_time + print(f'Overall fps: {fps:.2f} img / s\n') + benchmark_dict[f'overall_fps_{time_index + 1}'] = round(fps, 2) + overall_fps_list.append(fps) + break + benchmark_dict['average_fps'] = round(np.mean(overall_fps_list), 2) + benchmark_dict['fps_variance'] = round(np.var(overall_fps_list), 4) + print(f'Average fps of {repeat_times} evaluations: ' + f'{benchmark_dict["average_fps"]}') + print(f'The variance of {repeat_times} evaluations: ' + f'{benchmark_dict["fps_variance"]}') + mmcv.dump(benchmark_dict, json_file, indent=4) + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/confusion_matrix.py b/cv/semantic_segmentation/att_unet/pytorch/tools/confusion_matrix.py new file mode 100644 index 0000000000000000000000000000000000000000..2c5b64cf4e8214d17a7fb4b600e97a392a3aa297 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/confusion_matrix.py @@ -0,0 +1,184 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os + +import matplotlib.pyplot as plt +import mmcv +import numpy as np +from matplotlib.ticker import MultipleLocator +from mmcv import Config, DictAction + +from mmseg.datasets import build_dataset + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Generate confusion matrix from segmentation results') + parser.add_argument('config', help='test config file path') + parser.add_argument( + 'prediction_path', help='prediction path where test .pkl result') + parser.add_argument( + 'save_dir', help='directory where confusion matrix will be saved') + parser.add_argument( + '--show', action='store_true', help='show confusion matrix') + parser.add_argument( + '--color-theme', + default='winter', + help='theme of the matrix color map') + parser.add_argument( + '--title', + default='Normalized Confusion Matrix', + help='title of the matrix color map') + parser.add_argument( + '--cfg-options', + nargs='+', + action=DictAction, + help='override some settings in the used config, the key-value pair ' + 'in xxx=yyy format will be merged into config file. If the value to ' + 'be overwritten is a list, it should be like key="[a,b]" or key=a,b ' + 'It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" ' + 'Note that the quotation marks are necessary and that no white space ' + 'is allowed.') + args = parser.parse_args() + return args + + +def calculate_confusion_matrix(dataset, results): + """Calculate the confusion matrix. + + Args: + dataset (Dataset): Test or val dataset. + results (list[ndarray]): A list of segmentation results in each image. + """ + n = len(dataset.CLASSES) + confusion_matrix = np.zeros(shape=[n, n]) + assert len(dataset) == len(results) + prog_bar = mmcv.ProgressBar(len(results)) + for idx, per_img_res in enumerate(results): + res_segm = per_img_res + gt_segm = dataset.get_gt_seg_map_by_idx(idx) + inds = n * gt_segm + res_segm + inds = inds.flatten() + mat = np.bincount(inds, minlength=n**2).reshape(n, n) + confusion_matrix += mat + prog_bar.update() + return confusion_matrix + + +def plot_confusion_matrix(confusion_matrix, + labels, + save_dir=None, + show=True, + title='Normalized Confusion Matrix', + color_theme='winter'): + """Draw confusion matrix with matplotlib. + + Args: + confusion_matrix (ndarray): The confusion matrix. + labels (list[str]): List of class names. + save_dir (str|optional): If set, save the confusion matrix plot to the + given path. Default: None. + show (bool): Whether to show the plot. Default: True. + title (str): Title of the plot. Default: `Normalized Confusion Matrix`. + color_theme (str): Theme of the matrix color map. Default: `winter`. + """ + # normalize the confusion matrix + per_label_sums = confusion_matrix.sum(axis=1)[:, np.newaxis] + confusion_matrix = \ + confusion_matrix.astype(np.float32) / per_label_sums * 100 + + num_classes = len(labels) + fig, ax = plt.subplots( + figsize=(2 * num_classes, 2 * num_classes * 0.8), dpi=180) + cmap = plt.get_cmap(color_theme) + im = ax.imshow(confusion_matrix, cmap=cmap) + plt.colorbar(mappable=im, ax=ax) + + title_font = {'weight': 'bold', 'size': 12} + ax.set_title(title, fontdict=title_font) + label_font = {'size': 10} + plt.ylabel('Ground Truth Label', fontdict=label_font) + plt.xlabel('Prediction Label', fontdict=label_font) + + # draw locator + xmajor_locator = MultipleLocator(1) + xminor_locator = MultipleLocator(0.5) + ax.xaxis.set_major_locator(xmajor_locator) + ax.xaxis.set_minor_locator(xminor_locator) + ymajor_locator = MultipleLocator(1) + yminor_locator = MultipleLocator(0.5) + ax.yaxis.set_major_locator(ymajor_locator) + ax.yaxis.set_minor_locator(yminor_locator) + + # draw grid + ax.grid(True, which='minor', linestyle='-') + + # draw label + ax.set_xticks(np.arange(num_classes)) + ax.set_yticks(np.arange(num_classes)) + ax.set_xticklabels(labels) + ax.set_yticklabels(labels) + + ax.tick_params( + axis='x', bottom=False, top=True, labelbottom=False, labeltop=True) + plt.setp( + ax.get_xticklabels(), rotation=45, ha='left', rotation_mode='anchor') + + # draw confusion matrix value + for i in range(num_classes): + for j in range(num_classes): + ax.text( + j, + i, + '{}%'.format( + round(confusion_matrix[i, j], 2 + ) if not np.isnan(confusion_matrix[i, j]) else -1), + ha='center', + va='center', + color='w', + size=7) + + ax.set_ylim(len(confusion_matrix) - 0.5, -0.5) # matplotlib>3.1.1 + + fig.tight_layout() + if save_dir is not None: + plt.savefig( + os.path.join(save_dir, 'confusion_matrix.png'), format='png') + if show: + plt.show() + + +def main(): + args = parse_args() + + cfg = Config.fromfile(args.config) + if args.cfg_options is not None: + cfg.merge_from_dict(args.cfg_options) + + results = mmcv.load(args.prediction_path) + + assert isinstance(results, list) + if isinstance(results[0], np.ndarray): + pass + else: + raise TypeError('invalid type of prediction results') + + if isinstance(cfg.data.test, dict): + cfg.data.test.test_mode = True + elif isinstance(cfg.data.test, list): + for ds_cfg in cfg.data.test: + ds_cfg.test_mode = True + + dataset = build_dataset(cfg.data.test) + confusion_matrix = calculate_confusion_matrix(dataset, results) + plot_confusion_matrix( + confusion_matrix, + dataset.CLASSES, + save_dir=args.save_dir, + show=args.show, + title=args.title, + color_theme=args.color_theme) + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/cityscapes.py b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/cityscapes.py new file mode 100644 index 0000000000000000000000000000000000000000..17b61684787bf1e482bf912c1c988446ad320597 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/cityscapes.py @@ -0,0 +1,56 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os.path as osp + +import mmcv +from cityscapesscripts.preparation.json2labelImg import json2labelImg + + +def convert_json_to_label(json_file): + label_file = json_file.replace('_polygons.json', '_labelTrainIds.png') + json2labelImg(json_file, label_file, 'trainIds') + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Convert Cityscapes annotations to TrainIds') + parser.add_argument('cityscapes_path', help='cityscapes data path') + parser.add_argument('--gt-dir', default='gtFine', type=str) + parser.add_argument('-o', '--out-dir', help='output path') + parser.add_argument( + '--nproc', default=1, type=int, help='number of process') + args = parser.parse_args() + return args + + +def main(): + args = parse_args() + cityscapes_path = args.cityscapes_path + out_dir = args.out_dir if args.out_dir else cityscapes_path + mmcv.mkdir_or_exist(out_dir) + + gt_dir = osp.join(cityscapes_path, args.gt_dir) + + poly_files = [] + for poly in mmcv.scandir(gt_dir, '_polygons.json', recursive=True): + poly_file = osp.join(gt_dir, poly) + poly_files.append(poly_file) + if args.nproc > 1: + mmcv.track_parallel_progress(convert_json_to_label, poly_files, + args.nproc) + else: + mmcv.track_progress(convert_json_to_label, poly_files) + + split_names = ['train', 'val', 'test'] + + for split in split_names: + filenames = [] + for poly in mmcv.scandir( + osp.join(gt_dir, split), '_polygons.json', recursive=True): + filenames.append(poly.replace('_gtFine_polygons.json', '')) + with open(osp.join(out_dir, f'{split}.txt'), 'w') as f: + f.writelines(f + '\n' for f in filenames) + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/coco_stuff10k.py b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/coco_stuff10k.py new file mode 100644 index 0000000000000000000000000000000000000000..374f81970378f865a12307dedf6507ba61314e0f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/coco_stuff10k.py @@ -0,0 +1,307 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os.path as osp +import shutil +from functools import partial + +import mmcv +import numpy as np +from PIL import Image +from scipy.io import loadmat + +COCO_LEN = 10000 + +clsID_to_trID = { + 0: 0, + 1: 1, + 2: 2, + 3: 3, + 4: 4, + 5: 5, + 6: 6, + 7: 7, + 8: 8, + 9: 9, + 10: 10, + 11: 11, + 13: 12, + 14: 13, + 15: 14, + 16: 15, + 17: 16, + 18: 17, + 19: 18, + 20: 19, + 21: 20, + 22: 21, + 23: 22, + 24: 23, + 25: 24, + 27: 25, + 28: 26, + 31: 27, + 32: 28, + 33: 29, + 34: 30, + 35: 31, + 36: 32, + 37: 33, + 38: 34, + 39: 35, + 40: 36, + 41: 37, + 42: 38, + 43: 39, + 44: 40, + 46: 41, + 47: 42, + 48: 43, + 49: 44, + 50: 45, + 51: 46, + 52: 47, + 53: 48, + 54: 49, + 55: 50, + 56: 51, + 57: 52, + 58: 53, + 59: 54, + 60: 55, + 61: 56, + 62: 57, + 63: 58, + 64: 59, + 65: 60, + 67: 61, + 70: 62, + 72: 63, + 73: 64, + 74: 65, + 75: 66, + 76: 67, + 77: 68, + 78: 69, + 79: 70, + 80: 71, + 81: 72, + 82: 73, + 84: 74, + 85: 75, + 86: 76, + 87: 77, + 88: 78, + 89: 79, + 90: 80, + 92: 81, + 93: 82, + 94: 83, + 95: 84, + 96: 85, + 97: 86, + 98: 87, + 99: 88, + 100: 89, + 101: 90, + 102: 91, + 103: 92, + 104: 93, + 105: 94, + 106: 95, + 107: 96, + 108: 97, + 109: 98, + 110: 99, + 111: 100, + 112: 101, + 113: 102, + 114: 103, + 115: 104, + 116: 105, + 117: 106, + 118: 107, + 119: 108, + 120: 109, + 121: 110, + 122: 111, + 123: 112, + 124: 113, + 125: 114, + 126: 115, + 127: 116, + 128: 117, + 129: 118, + 130: 119, + 131: 120, + 132: 121, + 133: 122, + 134: 123, + 135: 124, + 136: 125, + 137: 126, + 138: 127, + 139: 128, + 140: 129, + 141: 130, + 142: 131, + 143: 132, + 144: 133, + 145: 134, + 146: 135, + 147: 136, + 148: 137, + 149: 138, + 150: 139, + 151: 140, + 152: 141, + 153: 142, + 154: 143, + 155: 144, + 156: 145, + 157: 146, + 158: 147, + 159: 148, + 160: 149, + 161: 150, + 162: 151, + 163: 152, + 164: 153, + 165: 154, + 166: 155, + 167: 156, + 168: 157, + 169: 158, + 170: 159, + 171: 160, + 172: 161, + 173: 162, + 174: 163, + 175: 164, + 176: 165, + 177: 166, + 178: 167, + 179: 168, + 180: 169, + 181: 170, + 182: 171 +} + + +def convert_to_trainID(tuple_path, in_img_dir, in_ann_dir, out_img_dir, + out_mask_dir, is_train): + imgpath, maskpath = tuple_path + shutil.copyfile( + osp.join(in_img_dir, imgpath), + osp.join(out_img_dir, 'train2014', imgpath) if is_train else osp.join( + out_img_dir, 'test2014', imgpath)) + annotate = loadmat(osp.join(in_ann_dir, maskpath)) + mask = annotate['S'].astype(np.uint8) + mask_copy = mask.copy() + for clsID, trID in clsID_to_trID.items(): + mask_copy[mask == clsID] = trID + seg_filename = osp.join(out_mask_dir, 'train2014', + maskpath.split('.')[0] + + '_labelTrainIds.png') if is_train else osp.join( + out_mask_dir, 'test2014', + maskpath.split('.')[0] + '_labelTrainIds.png') + Image.fromarray(mask_copy).save(seg_filename, 'PNG') + + +def generate_coco_list(folder): + train_list = osp.join(folder, 'imageLists', 'train.txt') + test_list = osp.join(folder, 'imageLists', 'test.txt') + train_paths = [] + test_paths = [] + + with open(train_list) as f: + for filename in f: + basename = filename.strip() + imgpath = basename + '.jpg' + maskpath = basename + '.mat' + train_paths.append((imgpath, maskpath)) + + with open(test_list) as f: + for filename in f: + basename = filename.strip() + imgpath = basename + '.jpg' + maskpath = basename + '.mat' + test_paths.append((imgpath, maskpath)) + + return train_paths, test_paths + + +def parse_args(): + parser = argparse.ArgumentParser( + description=\ + 'Convert COCO Stuff 10k annotations to mmsegmentation format') # noqa + parser.add_argument('coco_path', help='coco stuff path') + parser.add_argument('-o', '--out_dir', help='output path') + parser.add_argument( + '--nproc', default=16, type=int, help='number of process') + args = parser.parse_args() + return args + + +def main(): + args = parse_args() + coco_path = args.coco_path + nproc = args.nproc + + out_dir = args.out_dir or coco_path + out_img_dir = osp.join(out_dir, 'images') + out_mask_dir = osp.join(out_dir, 'annotations') + + mmcv.mkdir_or_exist(osp.join(out_img_dir, 'train2014')) + mmcv.mkdir_or_exist(osp.join(out_img_dir, 'test2014')) + mmcv.mkdir_or_exist(osp.join(out_mask_dir, 'train2014')) + mmcv.mkdir_or_exist(osp.join(out_mask_dir, 'test2014')) + + train_list, test_list = generate_coco_list(coco_path) + assert (len(train_list) + + len(test_list)) == COCO_LEN, 'Wrong length of list {} & {}'.format( + len(train_list), len(test_list)) + + if args.nproc > 1: + mmcv.track_parallel_progress( + partial( + convert_to_trainID, + in_img_dir=osp.join(coco_path, 'images'), + in_ann_dir=osp.join(coco_path, 'annotations'), + out_img_dir=out_img_dir, + out_mask_dir=out_mask_dir, + is_train=True), + train_list, + nproc=nproc) + mmcv.track_parallel_progress( + partial( + convert_to_trainID, + in_img_dir=osp.join(coco_path, 'images'), + in_ann_dir=osp.join(coco_path, 'annotations'), + out_img_dir=out_img_dir, + out_mask_dir=out_mask_dir, + is_train=False), + test_list, + nproc=nproc) + else: + mmcv.track_progress( + partial( + convert_to_trainID, + in_img_dir=osp.join(coco_path, 'images'), + in_ann_dir=osp.join(coco_path, 'annotations'), + out_img_dir=out_img_dir, + out_mask_dir=out_mask_dir, + is_train=True), train_list) + mmcv.track_progress( + partial( + convert_to_trainID, + in_img_dir=osp.join(coco_path, 'images'), + in_ann_dir=osp.join(coco_path, 'annotations'), + out_img_dir=out_img_dir, + out_mask_dir=out_mask_dir, + is_train=False), test_list) + + print('Done!') + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/coco_stuff164k.py b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/coco_stuff164k.py new file mode 100644 index 0000000000000000000000000000000000000000..6d8e2f2a315ea1bc63ec6ee1d9f37a72995330ce --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/coco_stuff164k.py @@ -0,0 +1,264 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os.path as osp +import shutil +from functools import partial +from glob import glob + +import mmcv +import numpy as np +from PIL import Image + +COCO_LEN = 123287 + +clsID_to_trID = { + 0: 0, + 1: 1, + 2: 2, + 3: 3, + 4: 4, + 5: 5, + 6: 6, + 7: 7, + 8: 8, + 9: 9, + 10: 10, + 12: 11, + 13: 12, + 14: 13, + 15: 14, + 16: 15, + 17: 16, + 18: 17, + 19: 18, + 20: 19, + 21: 20, + 22: 21, + 23: 22, + 24: 23, + 26: 24, + 27: 25, + 30: 26, + 31: 27, + 32: 28, + 33: 29, + 34: 30, + 35: 31, + 36: 32, + 37: 33, + 38: 34, + 39: 35, + 40: 36, + 41: 37, + 42: 38, + 43: 39, + 45: 40, + 46: 41, + 47: 42, + 48: 43, + 49: 44, + 50: 45, + 51: 46, + 52: 47, + 53: 48, + 54: 49, + 55: 50, + 56: 51, + 57: 52, + 58: 53, + 59: 54, + 60: 55, + 61: 56, + 62: 57, + 63: 58, + 64: 59, + 66: 60, + 69: 61, + 71: 62, + 72: 63, + 73: 64, + 74: 65, + 75: 66, + 76: 67, + 77: 68, + 78: 69, + 79: 70, + 80: 71, + 81: 72, + 83: 73, + 84: 74, + 85: 75, + 86: 76, + 87: 77, + 88: 78, + 89: 79, + 91: 80, + 92: 81, + 93: 82, + 94: 83, + 95: 84, + 96: 85, + 97: 86, + 98: 87, + 99: 88, + 100: 89, + 101: 90, + 102: 91, + 103: 92, + 104: 93, + 105: 94, + 106: 95, + 107: 96, + 108: 97, + 109: 98, + 110: 99, + 111: 100, + 112: 101, + 113: 102, + 114: 103, + 115: 104, + 116: 105, + 117: 106, + 118: 107, + 119: 108, + 120: 109, + 121: 110, + 122: 111, + 123: 112, + 124: 113, + 125: 114, + 126: 115, + 127: 116, + 128: 117, + 129: 118, + 130: 119, + 131: 120, + 132: 121, + 133: 122, + 134: 123, + 135: 124, + 136: 125, + 137: 126, + 138: 127, + 139: 128, + 140: 129, + 141: 130, + 142: 131, + 143: 132, + 144: 133, + 145: 134, + 146: 135, + 147: 136, + 148: 137, + 149: 138, + 150: 139, + 151: 140, + 152: 141, + 153: 142, + 154: 143, + 155: 144, + 156: 145, + 157: 146, + 158: 147, + 159: 148, + 160: 149, + 161: 150, + 162: 151, + 163: 152, + 164: 153, + 165: 154, + 166: 155, + 167: 156, + 168: 157, + 169: 158, + 170: 159, + 171: 160, + 172: 161, + 173: 162, + 174: 163, + 175: 164, + 176: 165, + 177: 166, + 178: 167, + 179: 168, + 180: 169, + 181: 170, + 255: 255 +} + + +def convert_to_trainID(maskpath, out_mask_dir, is_train): + mask = np.array(Image.open(maskpath)) + mask_copy = mask.copy() + for clsID, trID in clsID_to_trID.items(): + mask_copy[mask == clsID] = trID + seg_filename = osp.join( + out_mask_dir, 'train2017', + osp.basename(maskpath).split('.')[0] + + '_labelTrainIds.png') if is_train else osp.join( + out_mask_dir, 'val2017', + osp.basename(maskpath).split('.')[0] + '_labelTrainIds.png') + Image.fromarray(mask_copy).save(seg_filename, 'PNG') + + +def parse_args(): + parser = argparse.ArgumentParser( + description=\ + 'Convert COCO Stuff 164k annotations to mmsegmentation format') # noqa + parser.add_argument('coco_path', help='coco stuff path') + parser.add_argument('-o', '--out_dir', help='output path') + parser.add_argument( + '--nproc', default=16, type=int, help='number of process') + args = parser.parse_args() + return args + + +def main(): + args = parse_args() + coco_path = args.coco_path + nproc = args.nproc + + out_dir = args.out_dir or coco_path + out_img_dir = osp.join(out_dir, 'images') + out_mask_dir = osp.join(out_dir, 'annotations') + + mmcv.mkdir_or_exist(osp.join(out_mask_dir, 'train2017')) + mmcv.mkdir_or_exist(osp.join(out_mask_dir, 'val2017')) + + if out_dir != coco_path: + shutil.copytree(osp.join(coco_path, 'images'), out_img_dir) + + train_list = glob(osp.join(coco_path, 'annotations', 'train2017', '*.png')) + train_list = [file for file in train_list if '_labelTrainIds' not in file] + test_list = glob(osp.join(coco_path, 'annotations', 'val2017', '*.png')) + test_list = [file for file in test_list if '_labelTrainIds' not in file] + assert (len(train_list) + + len(test_list)) == COCO_LEN, 'Wrong length of list {} & {}'.format( + len(train_list), len(test_list)) + + if args.nproc > 1: + mmcv.track_parallel_progress( + partial( + convert_to_trainID, out_mask_dir=out_mask_dir, is_train=True), + train_list, + nproc=nproc) + mmcv.track_parallel_progress( + partial( + convert_to_trainID, out_mask_dir=out_mask_dir, is_train=False), + test_list, + nproc=nproc) + else: + mmcv.track_progress( + partial( + convert_to_trainID, out_mask_dir=out_mask_dir, is_train=True), + train_list) + mmcv.track_progress( + partial( + convert_to_trainID, out_mask_dir=out_mask_dir, is_train=False), + test_list) + + print('Done!') + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/pascal_context.py b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/pascal_context.py new file mode 100644 index 0000000000000000000000000000000000000000..03b79d51869d5c675b750e2b901c820af907de0c --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/pascal_context.py @@ -0,0 +1,87 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os.path as osp +from functools import partial + +import mmcv +import numpy as np +from detail import Detail +from PIL import Image + +_mapping = np.sort( + np.array([ + 0, 2, 259, 260, 415, 324, 9, 258, 144, 18, 19, 22, 23, 397, 25, 284, + 158, 159, 416, 33, 162, 420, 454, 295, 296, 427, 44, 45, 46, 308, 59, + 440, 445, 31, 232, 65, 354, 424, 68, 326, 72, 458, 34, 207, 80, 355, + 85, 347, 220, 349, 360, 98, 187, 104, 105, 366, 189, 368, 113, 115 + ])) +_key = np.array(range(len(_mapping))).astype('uint8') + + +def generate_labels(img_id, detail, out_dir): + + def _class_to_index(mask, _mapping, _key): + # assert the values + values = np.unique(mask) + for i in range(len(values)): + assert (values[i] in _mapping) + index = np.digitize(mask.ravel(), _mapping, right=True) + return _key[index].reshape(mask.shape) + + mask = Image.fromarray( + _class_to_index(detail.getMask(img_id), _mapping=_mapping, _key=_key)) + filename = img_id['file_name'] + mask.save(osp.join(out_dir, filename.replace('jpg', 'png'))) + return osp.splitext(osp.basename(filename))[0] + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Convert PASCAL VOC annotations to mmsegmentation format') + parser.add_argument('devkit_path', help='pascal voc devkit path') + parser.add_argument('json_path', help='annoation json filepath') + parser.add_argument('-o', '--out_dir', help='output path') + args = parser.parse_args() + return args + + +def main(): + args = parse_args() + devkit_path = args.devkit_path + if args.out_dir is None: + out_dir = osp.join(devkit_path, 'VOC2010', 'SegmentationClassContext') + else: + out_dir = args.out_dir + json_path = args.json_path + mmcv.mkdir_or_exist(out_dir) + img_dir = osp.join(devkit_path, 'VOC2010', 'JPEGImages') + + train_detail = Detail(json_path, img_dir, 'train') + train_ids = train_detail.getImgs() + + val_detail = Detail(json_path, img_dir, 'val') + val_ids = val_detail.getImgs() + + mmcv.mkdir_or_exist( + osp.join(devkit_path, 'VOC2010/ImageSets/SegmentationContext')) + + train_list = mmcv.track_progress( + partial(generate_labels, detail=train_detail, out_dir=out_dir), + train_ids) + with open( + osp.join(devkit_path, 'VOC2010/ImageSets/SegmentationContext', + 'train.txt'), 'w') as f: + f.writelines(line + '\n' for line in sorted(train_list)) + + val_list = mmcv.track_progress( + partial(generate_labels, detail=val_detail, out_dir=out_dir), val_ids) + with open( + osp.join(devkit_path, 'VOC2010/ImageSets/SegmentationContext', + 'val.txt'), 'w') as f: + f.writelines(line + '\n' for line in sorted(val_list)) + + print('Done!') + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/voc_aug.py b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/voc_aug.py new file mode 100644 index 0000000000000000000000000000000000000000..1d42c2704703356b059873ad4904d9681ed9cce8 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/convert_datasets/voc_aug.py @@ -0,0 +1,92 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os.path as osp +from functools import partial + +import mmcv +import numpy as np +from PIL import Image +from scipy.io import loadmat + +AUG_LEN = 10582 + + +def convert_mat(mat_file, in_dir, out_dir): + data = loadmat(osp.join(in_dir, mat_file)) + mask = data['GTcls'][0]['Segmentation'][0].astype(np.uint8) + seg_filename = osp.join(out_dir, mat_file.replace('.mat', '.png')) + Image.fromarray(mask).save(seg_filename, 'PNG') + + +def generate_aug_list(merged_list, excluded_list): + return list(set(merged_list) - set(excluded_list)) + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Convert PASCAL VOC annotations to mmsegmentation format') + parser.add_argument('devkit_path', help='pascal voc devkit path') + parser.add_argument('aug_path', help='pascal voc aug path') + parser.add_argument('-o', '--out_dir', help='output path') + parser.add_argument( + '--nproc', default=1, type=int, help='number of process') + args = parser.parse_args() + return args + + +def main(): + args = parse_args() + devkit_path = args.devkit_path + aug_path = args.aug_path + nproc = args.nproc + if args.out_dir is None: + out_dir = osp.join(devkit_path, 'VOC2012', 'SegmentationClassAug') + else: + out_dir = args.out_dir + mmcv.mkdir_or_exist(out_dir) + in_dir = osp.join(aug_path, 'dataset', 'cls') + + mmcv.track_parallel_progress( + partial(convert_mat, in_dir=in_dir, out_dir=out_dir), + list(mmcv.scandir(in_dir, suffix='.mat')), + nproc=nproc) + + full_aug_list = [] + with open(osp.join(aug_path, 'dataset', 'train.txt')) as f: + full_aug_list += [line.strip() for line in f] + with open(osp.join(aug_path, 'dataset', 'val.txt')) as f: + full_aug_list += [line.strip() for line in f] + + with open( + osp.join(devkit_path, 'VOC2012/ImageSets/Segmentation', + 'train.txt')) as f: + ori_train_list = [line.strip() for line in f] + with open( + osp.join(devkit_path, 'VOC2012/ImageSets/Segmentation', + 'val.txt')) as f: + val_list = [line.strip() for line in f] + + aug_train_list = generate_aug_list(ori_train_list + full_aug_list, + val_list) + assert len(aug_train_list) == AUG_LEN, 'len(aug_train_list) != {}'.format( + AUG_LEN) + + with open( + osp.join(devkit_path, 'VOC2012/ImageSets/Segmentation', + 'trainaug.txt'), 'w') as f: + f.writelines(line + '\n' for line in aug_train_list) + + aug_list = generate_aug_list(full_aug_list, ori_train_list + val_list) + assert len(aug_list) == AUG_LEN - len( + ori_train_list), 'len(aug_list) != {}'.format(AUG_LEN - + len(ori_train_list)) + with open( + osp.join(devkit_path, 'VOC2012/ImageSets/Segmentation', 'aug.txt'), + 'w') as f: + f.writelines(line + '\n' for line in aug_list) + + print('Done!') + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/get_flops.py b/cv/semantic_segmentation/att_unet/pytorch/tools/get_flops.py new file mode 100644 index 0000000000000000000000000000000000000000..e30c36fdfc603bb247047088660a4966b5e2330b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/get_flops.py @@ -0,0 +1,60 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse + +from mmcv import Config +from mmcv.cnn import get_model_complexity_info + +from mmseg.models import build_segmentor + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Get the FLOPs of a segmentor') + parser.add_argument('config', help='train config file path') + parser.add_argument( + '--shape', + type=int, + nargs='+', + default=[2048, 1024], + help='input image size') + args = parser.parse_args() + return args + + +def main(): + + args = parse_args() + + if len(args.shape) == 1: + input_shape = (3, args.shape[0], args.shape[0]) + elif len(args.shape) == 2: + input_shape = (3, ) + tuple(args.shape) + else: + raise ValueError('invalid input shape') + + cfg = Config.fromfile(args.config) + cfg.model.pretrained = None + model = build_segmentor( + cfg.model, + train_cfg=cfg.get('train_cfg'), + test_cfg=cfg.get('test_cfg')).cuda() + model.eval() + + if hasattr(model, 'forward_dummy'): + model.forward = model.forward_dummy + else: + raise NotImplementedError( + 'FLOPs counter is currently not currently supported with {}'. + format(model.__class__.__name__)) + + flops, params = get_model_complexity_info(model, input_shape) + split_line = '=' * 30 + print('{0}\nInput shape: {1}\nFlops: {2}\nParams: {3}\n{0}'.format( + split_line, input_shape, flops, params)) + print('!!!Please be cautious if you use the results in papers. ' + 'You may need to check if all ops are supported and verify that the ' + 'flops computation is correct.') + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/print_config.py b/cv/semantic_segmentation/att_unet/pytorch/tools/print_config.py new file mode 100644 index 0000000000000000000000000000000000000000..3f9c08dd968535e14cd653b7fb9562e51bb482cf --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/print_config.py @@ -0,0 +1,69 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import warnings + +from mmcv import Config, DictAction + +from mmseg.apis import init_segmentor + + +def parse_args(): + parser = argparse.ArgumentParser(description='Print the whole config') + parser.add_argument('config', help='config file path') + parser.add_argument( + '--graph', action='store_true', help='print the models graph') + parser.add_argument( + '--options', + nargs='+', + action=DictAction, + help="--options is deprecated in favor of --cfg_options' and it will " + 'not be supported in version v0.22.0. Override some settings in the ' + 'used config, the key-value pair in xxx=yyy format will be merged ' + 'into config file. If the value to be overwritten is a list, it ' + 'should be like key="[a,b]" or key=a,b It also allows nested ' + 'list/tuple values, e.g. key="[(a,b),(c,d)]" Note that the quotation ' + 'marks are necessary and that no white space is allowed.') + parser.add_argument( + '--cfg-options', + nargs='+', + action=DictAction, + help='override some settings in the used config, the key-value pair ' + 'in xxx=yyy format will be merged into config file. If the value to ' + 'be overwritten is a list, it should be like key="[a,b]" or key=a,b ' + 'It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" ' + 'Note that the quotation marks are necessary and that no white space ' + 'is allowed.') + args = parser.parse_args() + + if args.options and args.cfg_options: + raise ValueError( + '--options and --cfg-options cannot be both ' + 'specified, --options is deprecated in favor of --cfg-options. ' + '--options will not be supported in version v0.22.0.') + if args.options: + warnings.warn('--options is deprecated in favor of --cfg-options, ' + '--options will not be supported in version v0.22.0.') + args.cfg_options = args.options + + return args + + +def main(): + args = parse_args() + + cfg = Config.fromfile(args.config) + if args.cfg_options is not None: + cfg.merge_from_dict(args.cfg_options) + print(f'Config:\n{cfg.pretty_text}') + # dump config + cfg.dump('example.py') + # dump models graph + if args.graph: + model = init_segmentor(args.config, device='cpu') + print(f'Model graph:\n{str(model)}') + with open('example-graph.txt', 'w') as f: + f.writelines(str(model)) + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/slurm_test.sh b/cv/semantic_segmentation/att_unet/pytorch/tools/slurm_test.sh new file mode 100755 index 0000000000000000000000000000000000000000..4e6f7bf4e33267f269cf0f455924cb70166ccd4b --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/slurm_test.sh @@ -0,0 +1,24 @@ +#!/usr/bin/env bash + +set -x + +PARTITION=$1 +JOB_NAME=$2 +CONFIG=$3 +CHECKPOINT=$4 +GPUS=${GPUS:-4} +GPUS_PER_NODE=${GPUS_PER_NODE:-4} +CPUS_PER_TASK=${CPUS_PER_TASK:-5} +PY_ARGS=${@:5} +SRUN_ARGS=${SRUN_ARGS:-""} + +PYTHONPATH="$(dirname $0)/..":$PYTHONPATH \ +srun -p ${PARTITION} \ + --job-name=${JOB_NAME} \ + --gres=gpu:${GPUS_PER_NODE} \ + --ntasks=${GPUS} \ + --ntasks-per-node=${GPUS_PER_NODE} \ + --cpus-per-task=${CPUS_PER_TASK} \ + --kill-on-bad-exit=1 \ + ${SRUN_ARGS} \ + python -u tools/test.py ${CONFIG} ${CHECKPOINT} --launcher="slurm" ${PY_ARGS} diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/slurm_train.sh b/cv/semantic_segmentation/att_unet/pytorch/tools/slurm_train.sh new file mode 100755 index 0000000000000000000000000000000000000000..ab232105f0309c720ed81a522eca14b6fbd64afd --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/slurm_train.sh @@ -0,0 +1,23 @@ +#!/usr/bin/env bash + +set -x + +PARTITION=$1 +JOB_NAME=$2 +CONFIG=$3 +GPUS=${GPUS:-4} +GPUS_PER_NODE=${GPUS_PER_NODE:-4} +CPUS_PER_TASK=${CPUS_PER_TASK:-5} +SRUN_ARGS=${SRUN_ARGS:-""} +PY_ARGS=${@:4} + +PYTHONPATH="$(dirname $0)/..":$PYTHONPATH \ +srun -p ${PARTITION} \ + --job-name=${JOB_NAME} \ + --gres=gpu:${GPUS_PER_NODE} \ + --ntasks=${GPUS} \ + --ntasks-per-node=${GPUS_PER_NODE} \ + --cpus-per-task=${CPUS_PER_TASK} \ + --kill-on-bad-exit=1 \ + ${SRUN_ARGS} \ + python -u tools/train.py ${CONFIG} --launcher="slurm" ${PY_ARGS} diff --git a/cv/semantic_segmentation/att_unet/pytorch/tools/test.py b/cv/semantic_segmentation/att_unet/pytorch/tools/test.py new file mode 100644 index 0000000000000000000000000000000000000000..12892ec9b164554b856065637b7f5621783ff21f --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/tools/test.py @@ -0,0 +1,319 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import os +import os.path as osp +import shutil +import time +import warnings + +import mmcv +import torch +from mmcv.cnn.utils import revert_sync_batchnorm +from mmcv.parallel import MMDataParallel, MMDistributedDataParallel +from mmcv.runner import (get_dist_info, init_dist, load_checkpoint, + wrap_fp16_model) +from mmcv.utils import DictAction + +from mmseg import digit_version +from mmseg.apis import multi_gpu_test, single_gpu_test +from mmseg.datasets import build_dataloader, build_dataset +from mmseg.models import build_segmentor +from mmseg.utils import setup_multi_processes + + +def parse_args(): + parser = argparse.ArgumentParser( + description='mmseg test (and eval) a model') + parser.add_argument('config', help='test config file path') + parser.add_argument('checkpoint', help='checkpoint file') + parser.add_argument( + '--work-dir', + help=('if specified, the evaluation metric results will be dumped' + 'into the directory as json')) + parser.add_argument( + '--aug-test', action='store_true', help='Use Flip and Multi scale aug') + parser.add_argument('--out', help='output result file in pickle format') + parser.add_argument( + '--format-only', + action='store_true', + help='Format the output results without perform evaluation. It is' + 'useful when you want to format the result to a specific format and ' + 'submit it to the test server') + parser.add_argument( + '--eval', + type=str, + nargs='+', + help='evaluation metrics, which depends on the dataset, e.g., "mIoU"' + ' for generic datasets, and "cityscapes" for Cityscapes') + parser.add_argument('--show', action='store_true', help='show results') + parser.add_argument( + '--show-dir', help='directory where painted images will be saved') + parser.add_argument( + '--gpu-collect', + action='store_true', + help='whether to use gpu to collect results.') + parser.add_argument( + '--gpu-id', + type=int, + default=0, + help='id of gpu to use ' + '(only applicable to non-distributed testing)') + parser.add_argument( + '--tmpdir', + help='tmp directory used for collecting results from multiple ' + 'workers, available when gpu_collect is not specified') + parser.add_argument( + '--options', + nargs='+', + action=DictAction, + help="--options is deprecated in favor of --cfg_options' and it will " + 'not be supported in version v0.22.0. Override some settings in the ' + 'used config, the key-value pair in xxx=yyy format will be merged ' + 'into config file. If the value to be overwritten is a list, it ' + 'should be like key="[a,b]" or key=a,b It also allows nested ' + 'list/tuple values, e.g. key="[(a,b),(c,d)]" Note that the quotation ' + 'marks are necessary and that no white space is allowed.') + parser.add_argument( + '--cfg-options', + nargs='+', + action=DictAction, + help='override some settings in the used config, the key-value pair ' + 'in xxx=yyy format will be merged into config file. If the value to ' + 'be overwritten is a list, it should be like key="[a,b]" or key=a,b ' + 'It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" ' + 'Note that the quotation marks are necessary and that no white space ' + 'is allowed.') + parser.add_argument( + '--eval-options', + nargs='+', + action=DictAction, + help='custom options for evaluation') + parser.add_argument( + '--launcher', + choices=['none', 'pytorch', 'slurm', 'mpi'], + default='none', + help='job launcher') + parser.add_argument( + '--opacity', + type=float, + default=0.5, + help='Opacity of painted segmentation map. In (0, 1] range.') + parser.add_argument('--local_rank', type=int, default=0) + args = parser.parse_args() + if 'LOCAL_RANK' not in os.environ: + os.environ['LOCAL_RANK'] = str(args.local_rank) + + if args.options and args.cfg_options: + raise ValueError( + '--options and --cfg-options cannot be both ' + 'specified, --options is deprecated in favor of --cfg-options. ' + '--options will not be supported in version v0.22.0.') + if args.options: + warnings.warn('--options is deprecated in favor of --cfg-options. ' + '--options will not be supported in version v0.22.0.') + args.cfg_options = args.options + + return args + + +def main(): + args = parse_args() + assert args.out or args.eval or args.format_only or args.show \ + or args.show_dir, \ + ('Please specify at least one operation (save/eval/format/show the ' + 'results / save the results) with the argument "--out", "--eval"' + ', "--format-only", "--show" or "--show-dir"') + + if args.eval and args.format_only: + raise ValueError('--eval and --format_only cannot be both specified') + + if args.out is not None and not args.out.endswith(('.pkl', '.pickle')): + raise ValueError('The output file must be a pkl file.') + + cfg = mmcv.Config.fromfile(args.config) + if args.cfg_options is not None: + cfg.merge_from_dict(args.cfg_options) + + # set multi-process settings + setup_multi_processes(cfg) + + # set cudnn_benchmark + if cfg.get('cudnn_benchmark', False): + torch.backends.cudnn.benchmark = True + if args.aug_test: + # hard code index + cfg.data.test.pipeline[1].img_ratios = [ + 0.5, 0.75, 1.0, 1.25, 1.5, 1.75 + ] + cfg.data.test.pipeline[1].flip = True + cfg.model.pretrained = None + cfg.data.test.test_mode = True + + if args.gpu_id is not None: + cfg.gpu_ids = [args.gpu_id] + + # init distributed env first, since logger depends on the dist info. + if args.launcher == 'none': + cfg.gpu_ids = [args.gpu_id] + distributed = False + if len(cfg.gpu_ids) > 1: + warnings.warn(f'The gpu-ids is reset from {cfg.gpu_ids} to ' + f'{cfg.gpu_ids[0:1]} to avoid potential error in ' + 'non-distribute testing time.') + cfg.gpu_ids = cfg.gpu_ids[0:1] + else: + distributed = True + init_dist(args.launcher, **cfg.dist_params) + + rank, _ = get_dist_info() + # allows not to create + if args.work_dir is not None and rank == 0: + mmcv.mkdir_or_exist(osp.abspath(args.work_dir)) + timestamp = time.strftime('%Y%m%d_%H%M%S', time.localtime()) + if args.aug_test: + json_file = osp.join(args.work_dir, + f'eval_multi_scale_{timestamp}.json') + else: + json_file = osp.join(args.work_dir, + f'eval_single_scale_{timestamp}.json') + elif rank == 0: + work_dir = osp.join('./work_dirs', + osp.splitext(osp.basename(args.config))[0]) + mmcv.mkdir_or_exist(osp.abspath(work_dir)) + timestamp = time.strftime('%Y%m%d_%H%M%S', time.localtime()) + if args.aug_test: + json_file = osp.join(work_dir, + f'eval_multi_scale_{timestamp}.json') + else: + json_file = osp.join(work_dir, + f'eval_single_scale_{timestamp}.json') + + # build the dataloader + # TODO: support multiple images per gpu (only minor changes are needed) + dataset = build_dataset(cfg.data.test) + # The default loader config + loader_cfg = dict( + # cfg.gpus will be ignored if distributed + num_gpus=len(cfg.gpu_ids), + dist=distributed, + shuffle=False) + # The overall dataloader settings + loader_cfg.update({ + k: v + for k, v in cfg.data.items() if k not in [ + 'train', 'val', 'test', 'train_dataloader', 'val_dataloader', + 'test_dataloader' + ] + }) + test_loader_cfg = { + **loader_cfg, + 'samples_per_gpu': 1, + 'shuffle': False, # Not shuffle by default + **cfg.data.get('test_dataloader', {}) + } + # build the dataloader + data_loader = build_dataloader(dataset, **test_loader_cfg) + + # build the model and load checkpoint + cfg.model.train_cfg = None + model = build_segmentor(cfg.model, test_cfg=cfg.get('test_cfg')) + fp16_cfg = cfg.get('fp16', None) + if fp16_cfg is not None: + wrap_fp16_model(model) + checkpoint = load_checkpoint(model, args.checkpoint, map_location='cpu') + if 'CLASSES' in checkpoint.get('meta', {}): + model.CLASSES = checkpoint['meta']['CLASSES'] + else: + print('"CLASSES" not found in meta, use dataset.CLASSES instead') + model.CLASSES = dataset.CLASSES + if 'PALETTE' in checkpoint.get('meta', {}): + model.PALETTE = checkpoint['meta']['PALETTE'] + else: + print('"PALETTE" not found in meta, use dataset.PALETTE instead') + model.PALETTE = dataset.PALETTE + + # clean gpu memory when starting a new evaluation. + torch.cuda.empty_cache() + eval_kwargs = {} if args.eval_options is None else args.eval_options + + # Deprecated + efficient_test = eval_kwargs.get('efficient_test', False) + if efficient_test: + warnings.warn( + '``efficient_test=True`` does not have effect in tools/test.py, ' + 'the evaluation and format results are CPU memory efficient by ' + 'default') + + eval_on_format_results = ( + args.eval is not None and 'cityscapes' in args.eval) + if eval_on_format_results: + assert len(args.eval) == 1, 'eval on format results is not ' \ + 'applicable for metrics other than ' \ + 'cityscapes' + if args.format_only or eval_on_format_results: + if 'imgfile_prefix' in eval_kwargs: + tmpdir = eval_kwargs['imgfile_prefix'] + else: + tmpdir = '.format_cityscapes' + eval_kwargs.setdefault('imgfile_prefix', tmpdir) + mmcv.mkdir_or_exist(tmpdir) + else: + tmpdir = None + + if not distributed: + warnings.warn( + 'SyncBN is only supported with DDP. To be compatible with DP, ' + 'we convert SyncBN to BN. Please use dist_train.sh which can ' + 'avoid this error.') + if not torch.cuda.is_available(): + assert digit_version(mmcv.__version__) >= digit_version('1.4.4'), \ + 'Please use MMCV >= 1.4.4 for CPU training!' + model = revert_sync_batchnorm(model) + model = MMDataParallel(model, device_ids=cfg.gpu_ids) + results = single_gpu_test( + model, + data_loader, + args.show, + args.show_dir, + False, + args.opacity, + pre_eval=args.eval is not None and not eval_on_format_results, + format_only=args.format_only or eval_on_format_results, + format_args=eval_kwargs) + else: + model = MMDistributedDataParallel( + model.cuda(), + device_ids=[torch.cuda.current_device()], + broadcast_buffers=False) + results = multi_gpu_test( + model, + data_loader, + args.tmpdir, + args.gpu_collect, + False, + pre_eval=args.eval is not None and not eval_on_format_results, + format_only=args.format_only or eval_on_format_results, + format_args=eval_kwargs) + + rank, _ = get_dist_info() + if rank == 0: + if args.out: + warnings.warn( + 'The behavior of ``args.out`` has been changed since MMSeg ' + 'v0.16, the pickled outputs could be seg map as type of ' + 'np.array, pre-eval results or file paths for ' + '``dataset.format_results()``.') + print(f'\nwriting results to {args.out}') + mmcv.dump(results, args.out) + if args.eval: + eval_kwargs.update(metric=args.eval) + metric = dataset.evaluate(results, **eval_kwargs) + metric_dict = dict(config=args.config, metric=metric) + mmcv.dump(metric_dict, json_file, indent=4) + if tmpdir is not None and eval_on_format_results: + # remove tmp dir when cityscapes evaluation + shutil.rmtree(tmpdir) + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/train.py b/cv/semantic_segmentation/att_unet/pytorch/train.py new file mode 100644 index 0000000000000000000000000000000000000000..e198dd60f69a10aed53421c81b1ee5dc22685800 --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/train.py @@ -0,0 +1,243 @@ +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import copy +import os +import os.path as osp +import time +import warnings + +import mmcv +import torch +import torch.distributed as dist +from mmcv.cnn.utils import revert_sync_batchnorm +from mmcv.runner import get_dist_info, init_dist +from mmcv.utils import Config, DictAction, get_git_hash + +from mmseg import __version__ +from mmseg.apis import init_random_seed, set_random_seed, train_segmentor +from mmseg.datasets import build_dataset +from mmseg.models import build_segmentor +from mmseg.utils import collect_env, get_root_logger, setup_multi_processes + + +def parse_args(): + parser = argparse.ArgumentParser(description='Train a segmentor') + parser.add_argument('config', help='train config file path') + parser.add_argument('--work-dir', help='the dir to save logs and models') + parser.add_argument( + '--load-from', help='the checkpoint file to load weights from') + parser.add_argument( + '--resume-from', help='the checkpoint file to resume from') + parser.add_argument( + '--no-validate', + action='store_true', + help='whether not to evaluate the checkpoint during training') + group_gpus = parser.add_mutually_exclusive_group() + group_gpus.add_argument( + '--gpus', + type=int, + help='(Deprecated, please use --gpu-id) number of gpus to use ' + '(only applicable to non-distributed training)') + group_gpus.add_argument( + '--gpu-ids', + type=int, + nargs='+', + help='(Deprecated, please use --gpu-id) ids of gpus to use ' + '(only applicable to non-distributed training)') + group_gpus.add_argument( + '--gpu-id', + type=int, + default=0, + help='id of gpu to use ' + '(only applicable to non-distributed training)') + parser.add_argument('--seed', type=int, default=None, help='random seed') + parser.add_argument( + '--diff_seed', + action='store_true', + help='Whether or not set different seeds for different ranks') + parser.add_argument( + '--deterministic', + action='store_true', + help='whether to set deterministic options for CUDNN backend.') + parser.add_argument( + '--options', + nargs='+', + action=DictAction, + help="--options is deprecated in favor of --cfg_options' and it will " + 'not be supported in version v0.22.0. Override some settings in the ' + 'used config, the key-value pair in xxx=yyy format will be merged ' + 'into config file. If the value to be overwritten is a list, it ' + 'should be like key="[a,b]" or key=a,b It also allows nested ' + 'list/tuple values, e.g. key="[(a,b),(c,d)]" Note that the quotation ' + 'marks are necessary and that no white space is allowed.') + parser.add_argument( + '--cfg-options', + nargs='+', + action=DictAction, + help='override some settings in the used config, the key-value pair ' + 'in xxx=yyy format will be merged into config file. If the value to ' + 'be overwritten is a list, it should be like key="[a,b]" or key=a,b ' + 'It also allows nested list/tuple values, e.g. key="[(a,b),(c,d)]" ' + 'Note that the quotation marks are necessary and that no white space ' + 'is allowed.') + parser.add_argument( + '--launcher', + choices=['none', 'pytorch', 'slurm', 'mpi'], + default='none', + help='job launcher') + parser.add_argument('--local_rank', type=int, default=0) + parser.add_argument('--dist_backend', type=str, default=None) + parser.add_argument( + '--auto-resume', + action='store_true', + help='resume from the latest checkpoint automatically.') + args = parser.parse_args() + if 'LOCAL_RANK' not in os.environ: + os.environ['LOCAL_RANK'] = str(args.local_rank) + + if args.options and args.cfg_options: + raise ValueError( + '--options and --cfg-options cannot be both ' + 'specified, --options is deprecated in favor of --cfg-options. ' + '--options will not be supported in version v0.22.0.') + if args.options: + warnings.warn('--options is deprecated in favor of --cfg-options. ' + '--options will not be supported in version v0.22.0.') + args.cfg_options = args.options + + return args + + +def main(): + args = parse_args() + + cfg = Config.fromfile(args.config) + if args.cfg_options is not None: + cfg.merge_from_dict(args.cfg_options) + + # set cudnn_benchmark + if cfg.get('cudnn_benchmark', False): + torch.backends.cudnn.benchmark = True + + # work_dir is determined in this priority: CLI > segment in file > filename + if args.work_dir is not None: + # update configs according to CLI args if args.work_dir is not None + cfg.work_dir = args.work_dir + elif cfg.get('work_dir', None) is None: + # use config filename as default work_dir if cfg.work_dir is None + cfg.work_dir = osp.join('./work_dirs', + osp.splitext(osp.basename(args.config))[0]) + if args.load_from is not None: + cfg.load_from = args.load_from + if args.resume_from is not None: + cfg.resume_from = args.resume_from + if args.gpus is not None: + cfg.gpu_ids = range(1) + warnings.warn('`--gpus` is deprecated because we only support ' + 'single GPU mode in non-distributed training. ' + 'Use `gpus=1` now.') + if args.gpu_ids is not None: + cfg.gpu_ids = args.gpu_ids[0:1] + warnings.warn('`--gpu-ids` is deprecated, please use `--gpu-id`. ' + 'Because we only support single GPU mode in ' + 'non-distributed training. Use the first GPU ' + 'in `gpu_ids` now.') + if args.gpus is None and args.gpu_ids is None: + cfg.gpu_ids = [args.gpu_id] + + cfg.auto_resume = args.auto_resume + + # init distributed env first, since logger depends on the dist info. + if args.launcher == 'none': + distributed = False + else: + distributed = True + if args.dist_backend is not None: + cfg.dist_params.backend = args.dist_backend + init_dist(args.launcher, **cfg.dist_params) + # gpu_ids is used to calculate iter when resuming checkpoint + _, world_size = get_dist_info() + cfg.gpu_ids = range(world_size) + + # create work_dir + mmcv.mkdir_or_exist(osp.abspath(cfg.work_dir)) + # dump config + cfg.dump(osp.join(cfg.work_dir, osp.basename(args.config))) + # init the logger before other steps + timestamp = time.strftime('%Y%m%d_%H%M%S', time.localtime()) + log_file = osp.join(cfg.work_dir, f'{timestamp}.log') + logger = get_root_logger(log_file=log_file, log_level=cfg.log_level) + + # set multi-process settings + setup_multi_processes(cfg) + + # init the meta dict to record some important information such as + # environment info and seed, which will be logged + meta = dict() + # log env info + env_info_dict = collect_env() + env_info = '\n'.join([f'{k}: {v}' for k, v in env_info_dict.items()]) + dash_line = '-' * 60 + '\n' + logger.info('Environment info:\n' + dash_line + env_info + '\n' + + dash_line) + meta['env_info'] = env_info + + # log some basic info + logger.info(f'Distributed training: {distributed}') + logger.info(f'Config:\n{cfg.pretty_text}') + + # set random seeds + seed = init_random_seed(args.seed) + seed = seed + dist.get_rank() if args.diff_seed else seed + logger.info(f'Set random seed to {seed}, ' + f'deterministic: {args.deterministic}') + set_random_seed(seed, deterministic=args.deterministic) + cfg.seed = seed + meta['seed'] = seed + meta['exp_name'] = osp.basename(args.config) + + model = build_segmentor( + cfg.model, + train_cfg=cfg.get('train_cfg'), + test_cfg=cfg.get('test_cfg')) + model.init_weights() + + # SyncBN is not support for DP + if not distributed: + warnings.warn( + 'SyncBN is only supported with DDP. To be compatible with DP, ' + 'we convert SyncBN to BN. Please use dist_train.sh which can ' + 'avoid this error.') + model = revert_sync_batchnorm(model) + + logger.info(model) + + datasets = [build_dataset(cfg.data.train)] + if len(cfg.workflow) == 2: + val_dataset = copy.deepcopy(cfg.data.val) + val_dataset.pipeline = cfg.data.train.pipeline + datasets.append(build_dataset(val_dataset)) + if cfg.checkpoint_config is not None: + # save mmseg version, config file content and class names in + # checkpoints as meta data + cfg.checkpoint_config.meta = dict( + mmseg_version=f'{__version__}+{get_git_hash()[:7]}', + config=cfg.pretty_text, + CLASSES=datasets[0].CLASSES, + PALETTE=datasets[0].PALETTE) + # add an attribute for visualization convenience + model.CLASSES = datasets[0].CLASSES + # passing checkpoint meta for saving best checkpoint + meta.update(cfg.checkpoint_config.meta) + train_segmentor( + model, + datasets, + cfg, + distributed=distributed, + validate=(not args.no_validate), + timestamp=timestamp, + meta=meta) + + +if __name__ == '__main__': + main() diff --git a/cv/semantic_segmentation/att_unet/pytorch/train_dist.sh b/cv/semantic_segmentation/att_unet/pytorch/train_dist.sh new file mode 100755 index 0000000000000000000000000000000000000000..d09675129188cd65caedfa5ab70c83f93f1488ee --- /dev/null +++ b/cv/semantic_segmentation/att_unet/pytorch/train_dist.sh @@ -0,0 +1,19 @@ +# Copyright (c) 2023, Shanghai Iluvatar CoreX Semiconductor Co., Ltd. + +CONFIG=$1 +GPUS=$2 +NNODES=${NNODES:-1} +NODE_RANK=${NODE_RANK:-0} +PORT=${PORT:-29500} +MASTER_ADDR=${MASTER_ADDR:-"127.0.0.1"} + +PYTHONPATH="$(dirname $0)/..":$PYTHONPATH \ +python3 -m torch.distributed.launch \ + --nnodes=$NNODES \ + --node_rank=$NODE_RANK \ + --master_addr=$MASTER_ADDR \ + --nproc_per_node=$GPUS \ + --master_port=$PORT \ + $(dirname "$0")/train.py \ + $CONFIG \ + --launcher pytorch ${@:3}