mirrors/mmsegmentation
1
0
Fork 0
mirror of https://github.com/open-mmlab/mmsegmentation.git synced 2025-06-03 22:03:48 +08:00
Code Issues Projects Releases Wiki Activity

[Improvement] Add markdown linter and fix linting errors (#171)

Browse Source 
* [Improvement] Add markdown linter and fix linting errors

* fixed pip
This commit is contained in:
Jerry Jiarui XU 2020-10-07 19:50:16 +08:00 committed by GitHub
parent 118fd9ddf0
commit 0264de0bb0
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
39 changed files with 228 additions and 94 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
.github/CONTRIBUTING.md vendored
View File

@ -13,16 +13,19 @@ All kinds of contributions are welcome, including but not limited to the followi
4. create a PR
Note
- If you plan to add some new features that involve large changes, it is encouraged to open an issue for discussion first.
- If you are the author of some papers and would like to include your method to mmsegmentation,
please contact Kai Chen (chenkaidev[at]gmail[dot]com). We will much appreciate your contribution.
please contact Kai Chen (chenkaidev[at]gmail[dot]com). We will much appreciate your contribution.
## Code style
### Python
We adopt [PEP8](https://www.python.org/dev/peps/pep-0008/) as the preferred code style.
We use the following tools for linting and formatting:
- [flake8](http://flake8.pycqa.org/en/latest/): linter
- [yapf](https://github.com/google/yapf): formatter
- [isort](https://github.com/timothycrosley/isort): sort imports
@ -35,19 +38,20 @@ The config for a pre-commit hook is stored in [.pre-commit-config](../.pre-commi
After you clone the repository, you will need to install initialize pre-commit hook.
```
```shell
pip install -U pre-commit
```
From the repository folder
```
```shell
pre-commit install
```
After this on every commit check code linters and formatter will be enforced.
>Before you create a PR, make sure that your code lints and is formatted by yapf.
### C++ and CUDA
We follow the [Google C++ Style Guide](https://google.github.io/styleguide/cppguide.html).

15
.github/ISSUE_TEMPLATE/error-report.md vendored
View File

@ -10,6 +10,7 @@ assignees: ''
Thanks for your error report and we appreciate it a lot.
**Checklist**
1. I have searched related issues but cannot get the expected help.
2. The bug has not been fixed in the latest version.
@ -17,10 +18,13 @@ Thanks for your error report and we appreciate it a lot.
A clear and concise description of what the bug is.
**Reproduction**
1. What command or script did you run?
```
A placeholder for the command.
```
```none
A placeholder for the command.
```
2. Did you make any modifications on the code or config? Did you understand what you have modified?
3. What dataset did you use?
@ -32,10 +36,13 @@ A placeholder for the command.
- Other environment variables that may be related (such as `$PATH`, `$LD_LIBRARY_PATH`, `$PYTHONPATH`, etc.)
**Error traceback**
If applicable, paste the error trackback here.
```
```none
A placeholder for trackback.
```
**Bug fix**
If you have already identified the reason, you can provide the information here. If you are willing to create a PR to fix it, please also leave a comment here and that would be much appreciated!

2
.github/ISSUE_TEMPLATE/feature_request.md vendored
View File

@ -7,7 +7,7 @@ assignees: ''
---
**Describe the feature**
# Describe the feature
**Motivation**
A clear and concise description of the motivation of the feature.

18
.github/workflows/build.yml vendored
View File

@ -11,18 +11,16 @@ jobs:
uses: actions/setup-python@v1
with:
python-version: 3.7
- name: Install linting dependencies
- name: Install pre-commit hook
run: |
python -m pip install --upgrade pip
pip install flake8 isort==4.3.21 yapf interrogate
- name: Lint with flake8
run: flake8 .
- name: Lint with isort
run: isort --recursive --check-only --diff mmseg/ tests/ examples/
- name: Format python codes with yapf
run: yapf -r -d mmseg/ tests/ examples/
pip install pre-commit
pre-commit install
- name: Linting
run: pre-commit run --all-files
- name: Check docstring
run: interrogate -v --ignore-init-method --ignore-module --ignore-nested-functions --exclude mmseg/ops --ignore-regex "__repr__" --fail-under 80 mmseg
run: |
pip install interrogate
interrogate -v --ignore-init-method --ignore-module --ignore-nested-functions --exclude mmseg/ops --ignore-regex "__repr__" --fail-under 80 mmseg
build:
env:

5
.pre-commit-config.yaml
View File

@ -28,6 +28,11 @@ repos:
args: ["--remove"]
- id: mixed-line-ending
args: ["--fix=lf"]
- repo: https://github.com/jumanjihouse/pre-commit-hooks
rev: 2.1.4
hooks:
- id: markdownlint
args: ["-r", "~MD002,~MD013,~MD029,~MD033,~MD034,~MD036"]
- repo: https://github.com/myint/docformatter
rev: v1.3.1
hooks:

2
README.md
View File

@ -54,6 +54,7 @@ Please refer to [changelog.md](docs/changelog.md) for details and release histor
Results and models are available in the [model zoo](docs/model_zoo.md).
Supported backbones:
- [x] ResNet
- [x] ResNeXt
- [x] [HRNet](configs/hrnet/README.md)
@ -61,6 +62,7 @@ Supported backbones:
- [x] [MobileNetV2](configs/mobilenet_v2/README.md)
Supported methods:
- [x] [FCN](configs/fcn)
- [x] [PSPNet](configs/pspnet)
- [x] [DeepLabV3](configs/deeplabv3)

6
configs/ann/README.md
View File

@ -1,7 +1,8 @@
# Asymmetric Non-local Neural Networks for Semantic Segmentation
## Introduction
```
```latex
@inproceedings{annn,
author = {Zhen Zhu and
Mengde Xu and
@ -18,6 +19,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| ANN | R-50-D8 | 512x1024 | 40000 | 6 | 3.71 | 77.40 | 78.57 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x1024_40k_cityscapes/ann_r50-d8_512x1024_40k_cityscapes_20200605_095211-049fc292.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x1024_40k_cityscapes/ann_r50-d8_512x1024_40k_cityscapes_20200605_095211.log.json) |
@ -30,6 +32,7 @@
| ANN | R-101-D8 | 769x769 | 80000 | - | - | 78.80 | 80.34 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_769x769_80k_cityscapes/ann_r101-d8_769x769_80k_cityscapes_20200607_013713-a9d4be8d.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_769x769_80k_cityscapes/ann_r101-d8_769x769_80k_cityscapes_20200607_013713.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| ANN | R-50-D8 | 512x512 | 80000 | 9.1 | 21.01 | 41.01 | 42.30 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_80k_ade20k/ann_r50-d8_512x512_80k_ade20k_20200615_014818-26f75e11.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_80k_ade20k/ann_r50-d8_512x512_80k_ade20k_20200615_014818.log.json) |
@ -38,6 +41,7 @@
| ANN | R-101-D8 | 512x512 | 160000 | - | - | 42.94 | 44.06 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_160k_ade20k/ann_r101-d8_512x512_160k_ade20k_20200615_231733-955eb1ec.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r101-d8_512x512_160k_ade20k/ann_r101-d8_512x512_160k_ade20k_20200615_231733.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| ANN | R-50-D8 | 512x512 | 20000 | 6 | 20.92 | 74.86 | 76.13 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_20k_voc12aug/ann_r50-d8_512x512_20k_voc12aug_20200617_222246-dfcb1c62.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ann/ann_r50-d8_512x512_20k_voc12aug/ann_r50-d8_512x512_20k_voc12aug_20200617_222246.log.json) |

6
configs/ccnet/README.md
View File

@ -1,7 +1,8 @@
# CCNet: Criss-Cross Attention for Semantic Segmentation
## Introduction
```
```latex
@article{huang2018ccnet,
title={CCNet: Criss-Cross Attention for Semantic Segmentation},
author={Huang, Zilong and Wang, Xinggang and Huang, Lichao and Huang, Chang and Wei, Yunchao and Liu, Wenyu},
@ -13,6 +14,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| CCNet | R-50-D8 | 512x1024 | 40000 | 6 | 3.32 | 77.76 | 78.87 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x1024_40k_cityscapes/ccnet_r50-d8_512x1024_40k_cityscapes_20200616_142517-4123f401.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x1024_40k_cityscapes/ccnet_r50-d8_512x1024_40k_cityscapes_20200616_142517.log.json) |
@ -25,6 +27,7 @@
| CCNet | R-101-D8 | 769x769 | 80000 | - | - | 79.45 | 80.66 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_769x769_80k_cityscapes/ccnet_r101-d8_769x769_80k_cityscapes_20200618_011502-ad3cd481.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_769x769_80k_cityscapes/ccnet_r101-d8_769x769_80k_cityscapes_20200618_011502.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| CCNet | R-50-D8 | 512x512 | 80000 | 8.8 | 20.89 | 41.78 | 42.98 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_80k_ade20k/ccnet_r50-d8_512x512_80k_ade20k_20200615_014848-aa37f61e.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_80k_ade20k/ccnet_r50-d8_512x512_80k_ade20k_20200615_014848.log.json) |
@ -33,6 +36,7 @@
| CCNet | R-101-D8 | 512x512 | 160000 | - | - | 43.71 | 45.04 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x512_160k_ade20k/ccnet_r101-d8_512x512_160k_ade20k_20200616_000644-e849e007.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r101-d8_512x512_160k_ade20k/ccnet_r101-d8_512x512_160k_ade20k_20200616_000644.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| CCNet | R-50-D8 | 512x512 | 20000 | 6 | 20.45 | 76.17 | 77.51 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_20k_voc12aug/ccnet_r50-d8_512x512_20k_voc12aug_20200617_193212-fad81784.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ccnet/ccnet_r50-d8_512x512_20k_voc12aug/ccnet_r50-d8_512x512_20k_voc12aug_20200617_193212.log.json) |

6
configs/danet/README.md
View File

@ -1,7 +1,8 @@
# Dual Attention Network for Scene Segmentation
## Introduction
```
```latex
@article{fu2018dual,
title={Dual Attention Network for Scene Segmentation},
author={Jun Fu, Jing Liu, Haijie Tian, Yong Li, Yongjun Bao, Zhiwei Fang,and Hanqing Lu},
@ -13,6 +14,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|---------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DANet | R-50-D8 | 512x1024 | 40000 | 7.4 | 2.66 | 78.74 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r50-d8_512x1024_40k_cityscapes/danet_r50-d8_512x1024_40k_cityscapes_20200605_191324-c0dbfa5f.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r50-d8_512x1024_40k_cityscapes/danet_r50-d8_512x1024_40k_cityscapes_20200605_191324.log.json) |
@ -25,6 +27,7 @@
| DANet | R-101-D8 | 769x769 | 80000 | - | - | 80.47 | 82.02 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r101-d8_769x769_80k_cityscapes/danet_r101-d8_769x769_80k_cityscapes_20200607_132918-f3a929e7.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r101-d8_769x769_80k_cityscapes/danet_r101-d8_769x769_80k_cityscapes_20200607_132918.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DANet | R-50-D8 | 512x512 | 80000 | 11.5 | 21.20 | 41.66 | 42.90 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r50-d8_512x512_80k_ade20k/danet_r50-d8_512x512_80k_ade20k_20200615_015125-edb18e08.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r50-d8_512x512_80k_ade20k/danet_r50-d8_512x512_80k_ade20k_20200615_015125.log.json) |
@ -33,6 +36,7 @@
| DANet | R-101-D8 | 512x512 | 160000 | - | - | 44.17 | 45.02 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r101-d8_512x512_160k_ade20k/danet_r101-d8_512x512_160k_ade20k_20200616_082348-23bf12f9.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r101-d8_512x512_160k_ade20k/danet_r101-d8_512x512_160k_ade20k_20200616_082348.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DANet | R-50-D8 | 512x512 | 20000 | 6.5 | 20.94 | 74.45 | 75.69 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r50-d8_512x512_20k_voc12aug/danet_r50-d8_512x512_20k_voc12aug_20200618_070026-9e9e3ab3.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/danet/danet_r50-d8_512x512_20k_voc12aug/danet_r50-d8_512x512_20k_voc12aug_20200618_070026.log.json) |

7
configs/deeplabv3/README.md
View File

@ -1,7 +1,8 @@
# Rethinking atrous convolution for semantic image segmentation
## Introduction
```
```latext
@article{chen2017rethinking,
title={Rethinking atrous convolution for semantic image segmentation},
author={Chen, Liang-Chieh and Papandreou, George and Schroff, Florian and Adam, Hartwig},
@ -15,6 +16,7 @@
Note: `D-8` here corresponding to the output stride 8 setting for DeepLab series.
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|-----------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3 | R-50-D8 | 512x1024 | 40000 | 6.1 | 2.57 | 79.09 | 80.45 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r50-d8_512x1024_40k_cityscapes/deeplabv3_r50-d8_512x1024_40k_cityscapes_20200605_022449-acadc2f8.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r50-d8_512x1024_40k_cityscapes/deeplabv3_r50-d8_512x1024_40k_cityscapes_20200605_022449.log.json) |
@ -29,6 +31,7 @@ Note: `D-8` here corresponding to the output stride 8 setting for DeepLab series
| DeepLabV3 | R-101-D16-MG124 | 512x1024 | 80000 | - | - | 78.36 | 79.84 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d16-mg124_512x1024_80k_cityscapes/deeplabv3_r101-d16-mg124_512x1024_80k_cityscapes_20200908_005644-57bb8425.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d16-mg124_512x1024_80k_cityscapes/deeplabv3_r101-d16-mg124_512x1024_80k_cityscapes-20200908_005644.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|-----------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3 | R-50-D8 | 512x512 | 80000 | 8.9 | 14.76 | 42.42 | 43.28 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r50-d8_512x512_80k_ade20k/deeplabv3_r50-d8_512x512_80k_ade20k_20200614_185028-0bb3f844.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r50-d8_512x512_80k_ade20k/deeplabv3_r50-d8_512x512_80k_ade20k_20200614_185028.log.json) |
@ -37,6 +40,7 @@ Note: `D-8` here corresponding to the output stride 8 setting for DeepLab series
| DeepLabV3 | R-101-D8 | 512x512 | 160000 | - | - | 45.00 | 46.66 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d8_512x512_160k_ade20k/deeplabv3_r101-d8_512x512_160k_ade20k_20200615_105816-b1f72b3b.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d8_512x512_160k_ade20k/deeplabv3_r101-d8_512x512_160k_ade20k_20200615_105816.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|-----------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3 | R-50-D8 | 512x512 | 20000 | 6.1 | 13.88 | 76.17 | 77.42 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r50-d8_512x512_20k_voc12aug/deeplabv3_r50-d8_512x512_20k_voc12aug_20200617_010906-596905ef.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r50-d8_512x512_20k_voc12aug/deeplabv3_r50-d8_512x512_20k_voc12aug_20200617_010906.log.json) |
@ -45,6 +49,7 @@ Note: `D-8` here corresponding to the output stride 8 setting for DeepLab series
| DeepLabV3 | R-101-D8 | 512x512 | 40000 | - | - | 77.92 | 79.18 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d8_512x512_40k_voc12aug/deeplabv3_r101-d8_512x512_40k_voc12aug_20200613_161432-0017d784.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d8_512x512_40k_voc12aug/deeplabv3_r101-d8_512x512_40k_voc12aug_20200613_161432.log.json) |
### Pascal Context
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|-----------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3 | R-101-D8 | 480x480 | 40000 | 9.2 | 7.09 | 46.55 | 47.81 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d8_480x480_40k_pascal_context/deeplabv3_r101-d8_480x480_40k_pascal_context_20200911_204118-1aa27336.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3/deeplabv3_r101-d8_480x480_40k_pascal_context/deeplabv3_r101-d8_480x480_40k_pascal_context-20200911_204118.log.json) |

7
configs/deeplabv3plus/README.md
View File

@ -1,7 +1,8 @@
# Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
## Introduction
```
```latex
@inproceedings{deeplabv3plus2018,
title={Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation},
author={Liang-Chieh Chen and Yukun Zhu and George Papandreou and Florian Schroff and Hartwig Adam},
@ -17,6 +18,7 @@ Note:
`MG-124` stands for multi-grid dilation in the last stage of ResNet.
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3+ | R-50-D8 | 512x1024 | 40000 | 7.5 | 3.94 | 79.61 | 81.01 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r50-d8_512x1024_40k_cityscapes/deeplabv3plus_r50-d8_512x1024_40k_cityscapes_20200605_094610-d222ffcd.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r50-d8_512x1024_40k_cityscapes/deeplabv3plus_r50-d8_512x1024_40k_cityscapes_20200605_094610.log.json) |
@ -31,6 +33,7 @@ Note:
| DeepLabV3+ | R-101-D16-MG124 | 512x1024 | 80000 | 9.9 | - | 79.90 | 81.33 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d16-mg124_512x1024_80k_cityscapes/deeplabv3plus_r101-d16-mg124_512x1024_80k_cityscapes_20200908_005644-ee6158e0.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d16-mg124_512x1024_80k_cityscapes/deeplabv3plus_r101-d16-mg124_512x1024_80k_cityscapes-20200908_005644.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3+ | R-50-D8 | 512x512 | 80000 | 10.6 | 21.01 | 42.72 | 43.75 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r50-d8_512x512_80k_ade20k/deeplabv3plus_r50-d8_512x512_80k_ade20k_20200614_185028-bf1400d8.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r50-d8_512x512_80k_ade20k/deeplabv3plus_r50-d8_512x512_80k_ade20k_20200614_185028.log.json) |
@ -39,6 +42,7 @@ Note:
| DeepLabV3+ | R-101-D8 | 512x512 | 160000 | - | - | 45.47 | 46.35 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d8_512x512_160k_ade20k/deeplabv3plus_r101-d8_512x512_160k_ade20k_20200615_123232-38ed86bb.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d8_512x512_160k_ade20k/deeplabv3plus_r101-d8_512x512_160k_ade20k_20200615_123232.log.json) |
#### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3+ | R-50-D8 | 512x512 | 20000 | 7.6 | 21 | 75.93 | 77.50 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r50-d8_512x512_20k_voc12aug/deeplabv3plus_r50-d8_512x512_20k_voc12aug_20200617_102323-aad58ef1.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r50-d8_512x512_20k_voc12aug/deeplabv3plus_r50-d8_512x512_20k_voc12aug_20200617_102323.log.json) |
@ -47,6 +51,7 @@ Note:
| DeepLabV3+ | R-101-D8 | 512x512 | 40000 | - | - | 78.62 | 79.53 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d8_512x512_40k_voc12aug/deeplabv3plus_r101-d8_512x512_40k_voc12aug_20200613_205333-faf03387.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d8_512x512_40k_voc12aug/deeplabv3plus_r101-d8_512x512_40k_voc12aug_20200613_205333.log.json) |
#### Pascal Context
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| DeepLabV3+ | R-101-D8 | 480x480 | 40000 | - | 9.09 | 47.30 | 48.47 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d8_480x480_40k_pascal_context/deeplabv3plus_r101-d8_480x480_40k_pascal_context_20200911_165459-d3c8a29e.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/deeplabv3plus/deeplabv3plus_r101-d8_480x480_40k_pascal_context/deeplabv3plus_r101-d8_480x480_40k_pascal_context-20200911_165459.log.json) |

4
configs/dnlnet/README.md
View File

@ -5,7 +5,8 @@
This example is to reproduce ["Disentangled Non-Local Neural Networks"](https://arxiv.org/abs/2006.06668) for semantic segmentation. It is still in progress.
## Citation
```
```latex
@misc{yin2020disentangled,
title={Disentangled Non-Local Neural Networks},
author={Minghao Yin and Zhuliang Yao and Yue Cao and Xiu Li and Zheng Zhang and Stephen Lin and Han Hu},
@ -29,7 +30,6 @@ This example is to reproduce ["Disentangled Non-Local Neural Networks"](https://
| dnl | R-50-D8 | 769x769 | 80000 | - | - | 79.36 | 80.70 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/dnlnet/dnl_r50-d8_769x769_80k_cityscapes/dnl_r50-d8_769x769_80k_cityscapes_20200820_011925-366bc4c7.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/dnlnet/dnl_r50-d8_769x769_80k_cityscapes/dnl_r50-d8_769x769_80k_cityscapes-20200820_011925.log.json) |
| dnl | R-101-D8 | 769x769 | 80000 | - | - | 79.41 | 80.68 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/dnlnet/dnl_r101-d8_769x769_80k_cityscapes/dnl_r101-d8_769x769_80k_cityscapes_20200821_051111-95ff84ab.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/dnlnet/dnl_r101-d8_769x769_80k_cityscapes/dnl_r101-d8_769x769_80k_cityscapes-20200821_051111.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |

4
configs/emanet/README.md
View File

@ -1,7 +1,8 @@
# Expectation-Maximization Attention Networks for Semantic Segmentation
## Introduction
```
```latex
@inproceedings{li2019expectation,
title={Expectation-maximization attention networks for semantic segmentation},
author={Li, Xia and Zhong, Zhisheng and Wu, Jianlong and Yang, Yibo and Lin, Zhouchen and Liu, Hong},
@ -14,6 +15,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|---------:|----------------|------:|---------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| EMANet | R-50-D8 | 512x1024 | 80000 | 5.4 | 4.58 | 77.59 | 79.44 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/emanet/emanet_r50-d8_512x1024_80k_cityscapes/emanet_r50-d8_512x1024_80k_cityscapes_20200901_100301-c43fcef1.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/emanet/emanet_r50-d8_512x1024_80k_cityscapes/emanet_r50-d8_512x1024_80k_cityscapes-20200901_100301.log.json) |

5
configs/encnet/README.md
View File

@ -1,7 +1,8 @@
# Context Encoding for Semantic Segmentation
## Introduction
```
```latex
@InProceedings{Zhang_2018_CVPR,
author = {Zhang, Hang and Dana, Kristin and Shi, Jianping and Zhang, Zhongyue and Wang, Xiaogang and Tyagi, Ambrish and Agrawal, Amit},
title = {Context Encoding for Semantic Segmentation},
@ -14,6 +15,7 @@ year = {2018}
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| encnet | R-50-D8 | 512x1024 | 40000 | 8.6 | 4.58 | 75.67 | 77.08 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/encnet/encnet_r50-d8_512x1024_40k_cityscapes/encnet_r50-d8_512x1024_40k_cityscapes_20200621_220958-68638a47.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/encnet/encnet_r50-d8_512x1024_40k_cityscapes/encnet_r50-d8_512x1024_40k_cityscapes-20200621_220958.log.json) |
@ -26,6 +28,7 @@ year = {2018}
| encnet | R-101-D8 | 769x769 | 80000 | - | - | 76.10 | 76.97 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/encnet/encnet_r101-d8_769x769_80k_cityscapes/encnet_r101-d8_769x769_80k_cityscapes_20200622_003555-470ef79d.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/encnet/encnet_r101-d8_769x769_80k_cityscapes/encnet_r101-d8_769x769_80k_cityscapes-20200622_003555.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| encnet | R-50-D8 | 512x512 | 80000 | 10.1 | 22.81 | 39.53 | 41.17 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/encnet/encnet_r50-d8_512x512_80k_ade20k/encnet_r50-d8_512x512_80k_ade20k_20200622_042412-44b46b04.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/encnet/encnet_r50-d8_512x512_80k_ade20k/encnet_r50-d8_512x512_80k_ade20k-20200622_042412.log.json) |

4
configs/fastscnn/README.md
View File

@ -1,7 +1,8 @@
# Fast-SCNN for Semantic Segmentation
## Introduction
```
```latex
@article{poudel2019fast,
title={Fast-scnn: Fast semantic segmentation network},
author={Poudel, Rudra PK and Liwicki, Stephan and Cipolla, Roberto},
@ -13,6 +14,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|-----------|-----------|--------:|----------|----------------|------:|---------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| Fast-SCNN | Fast-SCNN | 512x1024 | 80000 | 8.4 | 63.61 | 69.06 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fast_scnn/fast_scnn_4x8_80k_lr0.12_cityscapes-f5096c79.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fast_scnn/fast_scnn_4x8_80k_lr0.12_cityscapes-20200807_165744.log.json) |

7
configs/fcn/README.md
View File

@ -1,7 +1,8 @@
# Fully Convolutional Networks for Semantic Segmentation
## Introduction
```
```latex
@article{shelhamer2017fully,
title={Fully convolutional networks for semantic segmentation},
author={Shelhamer, Evan and Long, Jonathan and Darrell, Trevor},
@ -17,6 +18,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | R-50-D8 | 512x1024 | 40000 | 5.7 | 4.17 | 72.25 | 73.36 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r50-d8_512x1024_40k_cityscapes/fcn_r50-d8_512x1024_40k_cityscapes_20200604_192608-efe53f0d.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r50-d8_512x1024_40k_cityscapes/fcn_r50-d8_512x1024_40k_cityscapes_20200604_192608.log.json) |
@ -29,6 +31,7 @@
| FCN | R-101-D8 | 769x769 | 80000 | - | - | 75.52 | 76.61 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_769x769_80k_cityscapes/fcn_r101-d8_769x769_80k_cityscapes_20200606_214354-45cbac68.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_769x769_80k_cityscapes/fcn_r101-d8_769x769_80k_cityscapes_20200606_214354.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | R-50-D8 | 512x512 | 80000 | 8.5 | 23.49 | 35.94 | 37.94 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r50-d8_512x512_80k_ade20k/fcn_r50-d8_512x512_80k_ade20k_20200614_144016-f8ac5082.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r50-d8_512x512_80k_ade20k/fcn_r50-d8_512x512_80k_ade20k_20200614_144016.log.json) |
@ -37,6 +40,7 @@
| FCN | R-101-D8 | 512x512 | 160000 | - | - | 39.91 | 41.40 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_512x512_160k_ade20k/fcn_r101-d8_512x512_160k_ade20k_20200615_105816-fd192bd5.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_512x512_160k_ade20k/fcn_r101-d8_512x512_160k_ade20k_20200615_105816.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | R-50-D8 | 512x512 | 20000 | 5.7 | 23.28 | 67.08 | 69.94 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r50-d8_512x512_20k_voc12aug/fcn_r50-d8_512x512_20k_voc12aug_20200617_010715-52dc5306.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r50-d8_512x512_20k_voc12aug/fcn_r50-d8_512x512_20k_voc12aug_20200617_010715.log.json) |
@ -45,6 +49,7 @@
| FCN | R-101-D8 | 512x512 | 40000 | - | - | 69.91 | 72.38 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_512x512_40k_voc12aug/fcn_r101-d8_512x512_40k_voc12aug_20200613_161240-4c8bcefd.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_512x512_40k_voc12aug/fcn_r101-d8_512x512_40k_voc12aug_20200613_161240.log.json) |
### Pascal Context
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | R-101-D8 | 480x480 | 40000 | - | 9.93 | 44.14 | 45.67 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_480x480_40k_pascal_context/fcn_r101-d8_480x480_40k_pascal_context_20200911_212515-9b565a6d.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fcn/fcn_r101-d8_480x480_40k_pascal_context/fcn_r101-d8_480x480_40k_pascal_context-20200911_212515.log.json) |

4
configs/fp16/README.md
View File

@ -1,7 +1,8 @@
# Mixed Precision Training
## Introduction
```
```latex
@article{micikevicius2017mixed,
title={Mixed precision training},
author={Micikevicius, Paulius and Narang, Sharan and Alben, Jonah and Diamos, Gregory and Elsen, Erich and Garcia, David and Ginsburg, Boris and Houston, Michael and Kuchaiev, Oleksii and Venkatesh, Ganesh and others},
@ -13,6 +14,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | R-101-D8 | 512x1024 | 80000 | 5.50 | 2.66 | 76.80 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/fp16/fcn_r101-d8_512x1024_80k_fp16_cityscapes/fcn_r101-d8_512x1024_80k_fp16_cityscapes-50245227.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/fp16/fcn_r101-d8_512x1024_80k_fp16_cityscapes/fcn_r101-d8_512x1024_80k_fp16_cityscapes_20200717_230921.log.json) |

6
configs/gcnet/README.md
View File

@ -1,7 +1,8 @@
# GCNet: Non-local Networks Meet Squeeze-Excitation Networks and Beyond
## Introduction
```
```latex
@inproceedings{cao2019gcnet,
title={Gcnet: Non-local networks meet squeeze-excitation networks and beyond},
author={Cao, Yue and Xu, Jiarui and Lin, Stephen and Wei, Fangyun and Hu, Han},
@ -14,6 +15,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| GCNet | R-50-D8 | 512x1024 | 40000 | 5.8 | 3.93 | 77.69 | 78.56 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r50-d8_512x1024_40k_cityscapes/gcnet_r50-d8_512x1024_40k_cityscapes_20200618_074436-4b0fd17b.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r50-d8_512x1024_40k_cityscapes/gcnet_r50-d8_512x1024_40k_cityscapes_20200618_074436.log.json) |
@ -26,6 +28,7 @@
| GCNet | R-101-D8 | 769x769 | 80000 | - | - | 79.18 | 80.71 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r101-d8_769x769_80k_cityscapes/gcnet_r101-d8_769x769_80k_cityscapes_20200619_092628-8e043423.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r101-d8_769x769_80k_cityscapes/gcnet_r101-d8_769x769_80k_cityscapes_20200619_092628.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| GCNet | R-50-D8 | 512x512 | 80000 | 8.5 | 23.38 | 41.47 | 42.85 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r50-d8_512x512_80k_ade20k/gcnet_r50-d8_512x512_80k_ade20k_20200614_185146-91a6da41.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r50-d8_512x512_80k_ade20k/gcnet_r50-d8_512x512_80k_ade20k_20200614_185146.log.json) |
@ -34,6 +37,7 @@
| GCNet | R-101-D8 | 512x512 | 160000 | - | - | 43.69 | 45.21 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r101-d8_512x512_160k_ade20k/gcnet_r101-d8_512x512_160k_ade20k_20200615_225406-615528d7.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r101-d8_512x512_160k_ade20k/gcnet_r101-d8_512x512_160k_ade20k_20200615_225406.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| GCNet | R-50-D8 | 512x512 | 20000 | 5.8 | 23.35 | 76.42 | 77.51 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r50-d8_512x512_20k_voc12aug/gcnet_r50-d8_512x512_20k_voc12aug_20200617_165701-3cbfdab1.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/gcnet/gcnet_r50-d8_512x512_20k_voc12aug/gcnet_r50-d8_512x512_20k_voc12aug_20200617_165701.log.json) |

7
configs/hrnet/README.md
View File

@ -1,7 +1,8 @@
# Deep High-Resolution Representation Learning for Human Pose Estimation
## Introduction
```
```latext
@inproceedings{SunXLW19,
title={Deep High-Resolution Representation Learning for Human Pose Estimation},
author={Ke Sun and Bin Xiao and Dong Liu and Jingdong Wang},
@ -13,6 +14,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|--------------------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | HRNetV2p-W18-Small | 512x1024 | 40000 | 1.7 | 23.74 | 73.86 | 75.91 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr18s_512x1024_40k_cityscapes/fcn_hr18s_512x1024_40k_cityscapes_20200601_014216-93db27d0.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr18s_512x1024_40k_cityscapes/fcn_hr18s_512x1024_40k_cityscapes_20200601_014216.log.json) |
@ -26,6 +28,7 @@
| FCN | HRNetV2p-W48 | 512x1024 | 160000 | - | - | 80.65 | 81.92 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_512x1024_160k_cityscapes/fcn_hr48_512x1024_160k_cityscapes_20200602_190946-59b7973e.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_512x1024_160k_cityscapes/fcn_hr48_512x1024_160k_cityscapes_20200602_190946.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|--------------------|-----------|--------:|----------|----------------|------:|--------------:|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | HRNetV2p-W18-Small | 512x512 | 80000 | 3.8 | 38.66 | 31.38 | 32.45 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr18s_512x512_80k_ade20k/fcn_hr18s_512x512_80k_ade20k_20200614_144345-77fc814a.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr18s_512x512_80k_ade20k/fcn_hr18s_512x512_80k_ade20k_20200614_144345.log.json) |
@ -36,6 +39,7 @@
| FCN | HRNetV2p-W48 | 512x512 | 160000 | - | - | 42.02 | 43.86 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_512x512_160k_ade20k/fcn_hr48_512x512_160k_ade20k_20200614_214407-a52fc02c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_512x512_160k_ade20k/fcn_hr48_512x512_160k_ade20k_20200614_214407.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|--------------------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | HRNetV2p-W18-Small | 512x512 | 20000 | 1.8 | 43.36 | 65.20 | 68.55 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr18s_512x512_20k_voc12aug/fcn_hr18s_512x512_20k_voc12aug_20200617_224503-56e36088.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr18s_512x512_20k_voc12aug/fcn_hr18s_512x512_20k_voc12aug_20200617_224503.log.json) |
@ -46,6 +50,7 @@
| FCN | HRNetV2p-W48 | 512x512 | 40000 | - | - | 76.24 | 78.49 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_512x512_40k_voc12aug/fcn_hr48_512x512_40k_voc12aug_20200613_222111-1b0f18bc.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_512x512_40k_voc12aug/fcn_hr48_512x512_40k_voc12aug_20200613_222111.log.json) |
### Pascal Context
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|--------------------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | HRNetV2p-W48 | 480x480 | 40000 | 6.1 | 8.86 | 45.14 | 47.42 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_480x480_40k_pascal_context/fcn_hr48_480x480_40k_pascal_context_20200911_164852-667d00b0.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/hrnet/fcn_hr48_480x480_40k_pascal_context/fcn_hr48_480x480_40k_pascal_context-20200911_164852.log.json) |

5
configs/mobilenet_v2/README.md
View File

@ -2,7 +2,7 @@
## Introduction
```
```latex
@inproceedings{sandler2018mobilenetv2,
title={Mobilenetv2: Inverted residuals and linear bottlenecks},
author={Sandler, Mark and Howard, Andrew and Zhu, Menglong and Zhmoginov, Andrey and Chen, Liang-Chieh},
@ -12,10 +12,10 @@
}
```
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|---------:|----------------|------:|---------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | M-V2-D8 | 512x1024 | 80000 | 3.4 | 14.2 | 61.54 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/mobilenet_v2/fcn_m-v2-d8_512x1024_80k_cityscapes/fcn_m-v2-d8_512x1024_80k_cityscapes_20200825_124817-d24c28c1.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/mobilenet_v2/fcn_m-v2-d8_512x1024_80k_cityscapes/fcn_m-v2-d8_512x1024_80k_cityscapes-20200825_124817.log.json) |
@ -24,6 +24,7 @@
| DeepLabV3+ | M-V2-D8 | 512x1024 | 80000 | 5.1 | 8.4 | 75.20 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/mobilenet_v2/deeplabv3plus_m-v2-d8_512x1024_80k_cityscapes/deeplabv3plus_m-v2-d8_512x1024_80k_cityscapes_20200825_124836-d256dd4b.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/mobilenet_v2/deeplabv3plus_m-v2-d8_512x1024_80k_cityscapes/deeplabv3plus_m-v2-d8_512x1024_80k_cityscapes-20200825_124836.log.json) |
### ADE20k
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|---------:|----------------|------:|---------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | M-V2-D8 | 512x512 | 160000 | 6.5 | 64.4 | 19.71 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/mobilenet_v2/fcn_m-v2-d8_512x512_160k_ade20k/fcn_m-v2-d8_512x512_160k_ade20k_20200825_214953-c40e1095.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/mobilenet_v2/fcn_m-v2-d8_512x512_160k_ade20k/fcn_m-v2-d8_512x512_160k_ade20k-20200825_214953.log.json) |

6
configs/nonlocal_net/README.md
View File

@ -1,7 +1,8 @@
# Non-local Neural Networks
## Introduction
```
```latex
@inproceedings{wang2018non,
title={Non-local neural networks},
author={Wang, Xiaolong and Girshick, Ross and Gupta, Abhinav and He, Kaiming},
@ -14,6 +15,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|----------|----------|-----------|--------:|----------|----------------|------:|---------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| NonLocal | R-50-D8 | 512x1024 | 40000 | 7.4 | 2.72 | 78.24 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r50-d8_512x1024_40k_cityscapes/nonlocal_r50-d8_512x1024_40k_cityscapes_20200605_210748-c75e81e3.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r50-d8_512x1024_40k_cityscapes/nonlocal_r50-d8_512x1024_40k_cityscapes_20200605_210748.log.json) |
@ -26,6 +28,7 @@
| NonLocal | R-101-D8 | 769x769 | 80000 | - | - | 79.40 | 80.85 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r101-d8_769x769_80k_cityscapes/nonlocal_r101-d8_769x769_80k_cityscapes_20200607_183428-0e1fa4f9.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r101-d8_769x769_80k_cityscapes/nonlocal_r101-d8_769x769_80k_cityscapes_20200607_183428.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|----------|----------|-----------|--------:|----------|----------------|------:|--------------:|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| NonLocal | R-50-D8 | 512x512 | 80000 | 9.1 | 21.37 | 40.75 | 42.05 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r50-d8_512x512_80k_ade20k/nonlocal_r50-d8_512x512_80k_ade20k_20200615_015801-5ae0aa33.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r50-d8_512x512_80k_ade20k/nonlocal_r50-d8_512x512_80k_ade20k_20200615_015801.log.json) |
@ -34,6 +37,7 @@
| NonLocal | R-101-D8 | 512x512 | 160000 | - | - | 43.36 | 44.83 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r101-d8_512x512_160k_ade20k/nonlocal_r101-d8_512x512_160k_ade20k_20200616_003422-affd0f8d.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r101-d8_512x512_160k_ade20k/nonlocal_r101-d8_512x512_160k_ade20k_20200616_003422.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|----------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| NonLocal | R-50-D8 | 512x512 | 20000 | 6.4 | 21.21 | 76.20 | 77.12 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r50-d8_512x512_20k_voc12aug/nonlocal_r50-d8_512x512_20k_voc12aug_20200617_222613-07f2a57c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/nonlocal_net/nonlocal_r50-d8_512x512_20k_voc12aug/nonlocal_r50-d8_512x512_20k_voc12aug_20200617_222613.log.json) |

7
configs/ocrnet/README.md
View File

@ -2,7 +2,7 @@
## Introduction
```
```latex
@article{YuanW18,
title={Ocnet: Object context network for scene parsing},
author={Yuhui Yuan and Jingdong Wang},
@ -23,6 +23,7 @@
### Cityscapes
#### HRNet backbone
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|--------------------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| OCRNet | HRNetV2p-W18-Small | 512x1024 | 40000 | 3.5 | 10.45 | 74.30 | 75.95 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18s_512x1024_40k_cityscapes/ocrnet_hr18s_512x1024_40k_cityscapes_20200601_033304-fa2436c2.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18s_512x1024_40k_cityscapes/ocrnet_hr18s_512x1024_40k_cityscapes_20200601_033304.log.json) |
@ -35,7 +36,6 @@
| OCRNet | HRNetV2p-W18 | 512x1024 | 160000 | - | - | 79.47 | 80.91 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18_512x1024_160k_cityscapes/ocrnet_hr18_512x1024_160k_cityscapes_20200602_191001-b9172d0c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18_512x1024_160k_cityscapes/ocrnet_hr18_512x1024_160k_cityscapes_20200602_191001.log.json) |
| OCRNet | HRNetV2p-W48 | 512x1024 | 160000 | - | - | 81.35 | 82.70 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr48_512x1024_160k_cityscapes/ocrnet_hr48_512x1024_160k_cityscapes_20200602_191037-dfbf1b0c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr48_512x1024_160k_cityscapes/ocrnet_hr48_512x1024_160k_cityscapes_20200602_191037.log.json) |
#### ResNet backbone
| Method | Backbone | Crop Size | Batch Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
@ -44,8 +44,8 @@
| OCRNet | R-101-D8 | 512x1024 | 16 | 40000 | 8.8 | 3.02 | 80.30 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_r101-d8_512x1024_40k_b16_cityscapes/ocrnet_r101-d8_512x1024_40k_b16_cityscapes-db500f80.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_r101-d8_512x1024_40k_b16_cityscapes/ocrnet_r101-d8_512x1024_40k_b16_cityscapes_20200723_193726.log.json) |
| OCRNet | R-101-D8 | 512x1024 | 16 | 80000 | 8.8 | 3.02 | 80.81 | - | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_r101-d8_512x1024_80k_b16_cityscapes/ocrnet_r101-d8_512x1024_80k_b16_cityscapes-78688424.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_r101-d8_512x1024_80k_b16_cityscapes/ocrnet_r101-d8_512x1024_80k_b16_cityscapes_20200723_192421.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|--------------------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| OCRNet | HRNetV2p-W18-Small | 512x512 | 80000 | 6.7 | 28.98 | 35.06 | 35.80 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18s_512x512_80k_ade20k/ocrnet_hr18s_512x512_80k_ade20k_20200615_055600-e80b62af.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18s_512x512_80k_ade20k/ocrnet_hr18s_512x512_80k_ade20k_20200615_055600.log.json) |
@ -56,6 +56,7 @@
| OCRNet | HRNetV2p-W48 | 512x512 | 160000 | - | - | 43.25 | 44.88 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr48_512x512_160k_ade20k/ocrnet_hr48_512x512_160k_ade20k_20200615_184705-a073726d.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr48_512x512_160k_ade20k/ocrnet_hr48_512x512_160k_ade20k_20200615_184705.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|--------------------|-----------|--------:|----------|----------------|------:|--------------:|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| OCRNet | HRNetV2p-W18-Small | 512x512 | 20000 | 3.5 | 31.55 | 71.70 | 73.84 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18s_512x512_20k_voc12aug/ocrnet_hr18s_512x512_20k_voc12aug_20200617_233913-02b04fcb.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/ocrnet/ocrnet_hr18s_512x512_20k_voc12aug/ocrnet_hr18s_512x512_20k_voc12aug_20200617_233913.log.json) |

6
configs/ocrnet/ocrnet_r101-d8_512x1024_40k_b16_cityscapes.py
View File

@ -1,8 +1,6 @@
_base_ = [
'../_base_/models/ocrnet_r50-d8.py',
'../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py',
'../_base_/schedules/schedule_40k.py'
'../_base_/models/ocrnet_r50-d8.py', '../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py', '../_base_/schedules/schedule_40k.py'
]
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))
optimizer = dict(lr=0.02)

6
configs/ocrnet/ocrnet_r101-d8_512x1024_40k_b8_cityscapes.py
View File

@ -1,7 +1,5 @@
_base_ = [
'../_base_/models/ocrnet_r50-d8.py',
'../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py',
'../_base_/schedules/schedule_40k.py'
'../_base_/models/ocrnet_r50-d8.py', '../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py', '../_base_/schedules/schedule_40k.py'
]
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))

6
configs/ocrnet/ocrnet_r101-d8_512x1024_80k_b16_cityscapes.py
View File

@ -1,8 +1,6 @@
_base_ = [
'../_base_/models/ocrnet_r50-d8.py',
'../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py',
'../_base_/schedules/schedule_80k.py'
'../_base_/models/ocrnet_r50-d8.py', '../_base_/datasets/cityscapes.py',
'../_base_/default_runtime.py', '../_base_/schedules/schedule_80k.py'
]
model = dict(pretrained='open-mmlab://resnet101_v1c', backbone=dict(depth=101))
optimizer = dict(lr=0.02)

3
configs/point_rend/README.md
View File

@ -1,6 +1,7 @@
# PointRend: Image Segmentation as Rendering
## Introduction
```
@misc{alex2019pointrend,
title={PointRend: Image Segmentation as Rendering},
@ -15,12 +16,14 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|-----------|----------|-----------|--------:|---------:|----------------|------:|---------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PointRend | R-50 | 512x1024 | 80000 | 3.1 | 8.48 | 76.47 | 78.13 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/point_rend/pointrend_r50_512x1024_80k_cityscapes/pointrend_r50_512x1024_80k_cityscapes_20200711_015821-bb1ff523.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/point_rend/pointrend_r50_512x1024_80k_cityscapes/pointrend_r50_512x1024_80k_cityscapes-20200715_214714.log.json) |
| PointRend | R-101 | 512x1024 | 80000 | 4.2 | 7.00 | 78.30 | 79.97 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/point_rend/pointrend_r101_512x1024_80k_cityscapes/pointrend_r101_512x1024_80k_cityscapes_20200711_170850-d0ca84be.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/point_rend/pointrend_r101_512x1024_80k_cityscapes/pointrend_r101_512x1024_80k_cityscapes-20200715_214824.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|-----------|----------|-----------|--------:|---------:|----------------|------:|---------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PointRend | R-50 | 512x512 | 160000 | 5.1 | 17.31 | 37.64 | 39.17 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/point_rend/pointrend_r50_512x512_160k_ade20k/pointrend_r50_512x512_160k_ade20k_20200807_232644-ac3febf2.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/point_rend/pointrend_r50_512x512_160k_ade20k/pointrend_r50_512x512_160k_ade20k-20200807_232644.log.json) |

6
configs/psanet/README.md
View File

@ -1,7 +1,8 @@
# PSANet: Point-wise Spatial Attention Network for Scene Parsing
## Introduction
```
```latex
@inproceedings{zhao2018psanet,
title={Psanet: Point-wise spatial attention network for scene parsing},
author={Zhao, Hengshuang and Zhang, Yi and Liu, Shu and Shi, Jianping and Change Loy, Chen and Lin, Dahua and Jia, Jiaya},
@ -14,6 +15,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PSANet | R-50-D8 | 512x1024 | 40000 | 7 | 3.17 | 77.63 | 79.04 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r50-d8_512x1024_40k_cityscapes/psanet_r50-d8_512x1024_40k_cityscapes_20200606_103117-99fac37c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r50-d8_512x1024_40k_cityscapes/psanet_r50-d8_512x1024_40k_cityscapes_20200606_103117.log.json) |
@ -26,6 +28,7 @@
| PSANet | R-101-D8 | 769x769 | 80000 | - | - | 79.69 | 80.89 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r101-d8_769x769_80k_cityscapes/psanet_r101-d8_769x769_80k_cityscapes_20200606_214550-7665827b.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r101-d8_769x769_80k_cityscapes/psanet_r101-d8_769x769_80k_cityscapes_20200606_214550.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PSANet | R-50-D8 | 512x512 | 80000 | 9 | 18.91 | 41.14 | 41.91 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r50-d8_512x512_80k_ade20k/psanet_r50-d8_512x512_80k_ade20k_20200614_144141-835e4b97.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r50-d8_512x512_80k_ade20k/psanet_r50-d8_512x512_80k_ade20k_20200614_144141.log.json) |
@ -34,6 +37,7 @@
| PSANet | R-101-D8 | 512x512 | 160000 | - | - | 43.74 | 45.38 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r101-d8_512x512_160k_ade20k/psanet_r101-d8_512x512_160k_ade20k_20200615_161537-dbfa564c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r101-d8_512x512_160k_ade20k/psanet_r101-d8_512x512_160k_ade20k_20200615_161537.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PSANet | R-50-D8 | 512x512 | 20000 | 6.9 | 18.24 | 76.39 | 77.34 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r50-d8_512x512_20k_voc12aug/psanet_r50-d8_512x512_20k_voc12aug_20200617_102413-2f1bbaa1.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/psanet/psanet_r50-d8_512x512_20k_voc12aug/psanet_r50-d8_512x512_20k_voc12aug_20200617_102413.log.json) |

7
configs/pspnet/README.md
View File

@ -1,7 +1,8 @@
# Pyramid Scene Parsing Network
## Introduction
```
```latex
@inproceedings{zhao2017pspnet,
title={Pyramid Scene Parsing Network},
author={Zhao, Hengshuang and Shi, Jianping and Qi, Xiaojuan and Wang, Xiaogang and Jia, Jiaya},
@ -13,6 +14,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PSPNet | R-50-D8 | 512x1024 | 40000 | 6.1 | 4.07 | 77.85 | 79.18 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r50-d8_512x1024_40k_cityscapes/pspnet_r50-d8_512x1024_40k_cityscapes_20200605_003338-2966598c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r50-d8_512x1024_40k_cityscapes/pspnet_r50-d8_512x1024_40k_cityscapes_20200605_003338.log.json) |
@ -25,6 +27,7 @@
| PSPNet | R-101-D8 | 769x769 | 80000 | - | - | 79.77 | 81.06 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_769x769_80k_cityscapes/pspnet_r101-d8_769x769_80k_cityscapes_20200606_225055-dba412fa.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_769x769_80k_cityscapes/pspnet_r101-d8_769x769_80k_cityscapes_20200606_225055.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PSPNet | R-50-D8 | 512x512 | 80000 | 8.5 | 23.53 | 41.13 | 41.94 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r50-d8_512x512_80k_ade20k/pspnet_r50-d8_512x512_80k_ade20k_20200615_014128-15a8b914.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r50-d8_512x512_80k_ade20k/pspnet_r50-d8_512x512_80k_ade20k_20200615_014128.log.json) |
@ -33,6 +36,7 @@
| PSPNet | R-101-D8 | 512x512 | 160000 | - | - | 44.39 | 45.35 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_512x512_160k_ade20k/pspnet_r101-d8_512x512_160k_ade20k_20200615_100650-967c316f.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_512x512_160k_ade20k/pspnet_r101-d8_512x512_160k_ade20k_20200615_100650.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PSPNet | R-50-D8 | 512x512 | 20000 | 6.1 | 23.59 | 76.78 | 77.61 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r50-d8_512x512_20k_voc12aug/pspnet_r50-d8_512x512_20k_voc12aug_20200617_101958-ed5dfbd9.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r50-d8_512x512_20k_voc12aug/pspnet_r50-d8_512x512_20k_voc12aug_20200617_101958.log.json) |
@ -41,6 +45,7 @@
| PSPNet | R-101-D8 | 512x512 | 40000 | - | - | 78.52 | 79.57 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_512x512_40k_voc12aug/pspnet_r101-d8_512x512_40k_voc12aug_20200613_161222-bc933b18.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_512x512_40k_voc12aug/pspnet_r101-d8_512x512_40k_voc12aug_20200613_161222.log.json) |
### Pascal Context
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| PSPNet | R-101-D8 | 480x480 | 40000 | 8.8 | 9.68 | 46.60 | 47.78 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_480x480_40k_pascal_context/pspnet_r101-d8_480x480_40k_pascal_context_20200911_211210-bf0f5d7c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/pspnet/pspnet_r101-d8_480x480_40k_pascal_context/pspnet_r101-d8_480x480_40k_pascal_context-20200911_211210.log.json) |

4
configs/resnest/README.md
View File

@ -2,7 +2,7 @@
## Introduction
```
```latex
@article{zhang2020resnest,
title={ResNeSt: Split-Attention Networks},
author={Zhang, Hang and Wu, Chongruo and Zhang, Zhongyue and Zhu, Yi and Zhang, Zhi and Lin, Haibin and Sun, Yue and He, Tong and Muller, Jonas and Manmatha, R. and Li, Mu and Smola, Alexander},
@ -14,6 +14,7 @@ year={2020}
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|---------:|----------------|------:|---------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | S-101-D8 | 512x1024 | 80000 | 11.4 | 2.39 | 77.56 | 78.98 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/resnest/fcn_s101-d8_512x1024_80k_cityscapes/fcn_s101-d8_512x1024_80k_cityscapes_20200807_140631-f8d155b3.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/resnest/fcn_s101-d8_512x1024_80k_cityscapes/fcn_s101-d8_512x1024_80k_cityscapes-20200807_140631.log.json) |
@ -22,6 +23,7 @@ year={2020}
| DeepLabV3+ | S-101-D8 | 512x1024 | 80000 | 13.2 | 2.36 | 79.62 | 80.27 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/resnest/deeplabv3plus_s101-d8_512x1024_80k_cityscapes/deeplabv3plus_s101-d8_512x1024_80k_cityscapes_20200807_144429-1239eb43.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/resnest/deeplabv3plus_s101-d8_512x1024_80k_cityscapes/deeplabv3plus_s101-d8_512x1024_80k_cityscapes-20200807_144429.log.json) |
### ADE20k
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|------------|----------|-----------|--------:|---------:|----------------|------:|---------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FCN | S-101-D8 | 512x512 | 160000 | 14.2 | 12.86 | 45.62 | 46.16 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/resnest/fcn_s101-d8_512x512_160k_ade20k/fcn_s101-d8_512x512_160k_ade20k_20200807_145416-d3160329.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/resnest/fcn_s101-d8_512x512_160k_ade20k/fcn_s101-d8_512x512_160k_ade20k-20200807_145416.log.json) |

5
configs/sem_fpn/README.md
View File

@ -1,7 +1,8 @@
# Panoptic Feature Pyramid Networks
## Introduction
```
```latex
@article{Kirillov_2019,
title={Panoptic Feature Pyramid Networks},
ISBN={9781728132938},
@ -18,12 +19,14 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|---------:|----------------|------:|---------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FPN | R-50 | 512x1024 | 80000 | 2.8 | 13.54 | 74.52 | 76.08 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/sem_fpn/fpn_r50_512x1024_80k_cityscapes/fpn_r50_512x1024_80k_cityscapes_20200717_021437-94018a0d.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/sem_fpn/fpn_r50_512x1024_80k_cityscapes/fpn_r50_512x1024_80k_cityscapes-20200717_021437.log.json) |
| FPN | R-101 | 512x1024 | 80000 | 3.9 | 10.29 | 75.80 | 77.40 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/sem_fpn/fpn_r101_512x1024_80k_cityscapes/fpn_r101_512x1024_80k_cityscapes_20200717_012416-c5800d4c.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/sem_fpn/fpn_r101_512x1024_80k_cityscapes/fpn_r101_512x1024_80k_cityscapes-20200717_012416.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|--------|----------|-----------|--------:|---------:|----------------|------:|---------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| FPN | R-50 | 512x512 | 160000 | 4.9 | 55.77 | 37.49 | 39.09 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/sem_fpn/fpn_r50_512x512_160k_ade20k/fpn_r50_512x512_160k_ade20k_20200718_131734-5b5a6ab9.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/sem_fpn/fpn_r50_512x512_160k_ade20k/fpn_r50_512x512_160k_ade20k-20200718_131734.log.json) |

6
configs/upernet/README.md
View File

@ -1,7 +1,8 @@
# Unified Perceptual Parsing for Scene Understanding
## Introduction
```
```latex
@inproceedings{xiao2018unified,
title={Unified perceptual parsing for scene understanding},
author={Xiao, Tete and Liu, Yingcheng and Zhou, Bolei and Jiang, Yuning and Sun, Jian},
@ -14,6 +15,7 @@
## Results and models
### Cityscapes
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|---------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| UPerNet | R-50 | 512x1024 | 40000 | 6.4 | 4.25 | 77.10 | 78.37 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r50_512x1024_40k_cityscapes/upernet_r50_512x1024_40k_cityscapes_20200605_094827-aa54cb54.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r50_512x1024_40k_cityscapes/upernet_r50_512x1024_40k_cityscapes_20200605_094827.log.json) |
@ -26,6 +28,7 @@
| UPerNet | R-101 | 769x769 | 80000 | - | - | 80.10 | 81.49 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r101_769x769_80k_cityscapes/upernet_r101_769x769_80k_cityscapes_20200607_001014-082fc334.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r101_769x769_80k_cityscapes/upernet_r101_769x769_80k_cityscapes_20200607_001014.log.json) |
### ADE20K
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|---------|----------|-----------|--------:|----------|----------------|------:|--------------:|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| UPerNet | R-50 | 512x512 | 80000 | 8.1 | 23.40 | 40.70 | 41.81 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r50_512x512_80k_ade20k/upernet_r50_512x512_80k_ade20k_20200614_144127-ecc8377b.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r50_512x512_80k_ade20k/upernet_r50_512x512_80k_ade20k_20200614_144127.log.json) |
@ -34,6 +37,7 @@
| UPerNet | R-101 | 512x512 | 160000 | - | - | 43.82 | 44.85 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r101_512x512_160k_ade20k/upernet_r101_512x512_160k_ade20k_20200615_161951-91b32684.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r101_512x512_160k_ade20k/upernet_r101_512x512_160k_ade20k_20200615_161951.log.json) |
### Pascal VOC 2012 + Aug
| Method | Backbone | Crop Size | Lr schd | Mem (GB) | Inf time (fps) | mIoU | mIoU(ms+flip) | download |
|---------|----------|-----------|--------:|----------|----------------|------:|--------------:|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| UPerNet | R-50 | 512x512 | 20000 | 6.4 | 23.17 | 74.82 | 76.35 | [model](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r50_512x512_20k_voc12aug/upernet_r50_512x512_20k_voc12aug_20200617_165330-5b5890a7.pth) | [log](https://download.openmmlab.com/mmsegmentation/v0.5/upernet/upernet_r50_512x512_20k_voc12aug/upernet_r50_512x512_20k_voc12aug_20200617_165330.log.json) |

17
docs/changelog.md
View File

@ -1,15 +1,17 @@
## Changelog
### V0.6 (10/09/2020)
**Highlights**
- Support new methods i.e. MobileNetV2, EMANet, DNL, PointRend, Semantic FPN, Fast-SCNN,
ResNeSt.
- Support new methods i.e. MobileNetV2, EMANet, DNL, PointRend, Semantic FPN, Fast-SCNN, ResNeSt.
**Bug Fixes**
- Fixed sliding inference ONNX export ([#90](https://github.com/open-mmlab/mmsegmentation/pull/90))
**New Features**
- Support MobileNet v2 ([#86](https://github.com/open-mmlab/mmsegmentation/pull/86))
- Support EMANet ([#34](https://github.com/open-mmlab/mmsegmentation/pull/34))
- Support DNL ([#37](https://github.com/open-mmlab/mmsegmentation/pull/37))
@ -20,6 +22,7 @@ ResNeSt.
- Support ONNX export (experimental) ([#12](https://github.com/open-mmlab/mmsegmentation/pull/12))
**Improvements**
- Support Upsample in ONNX ([#100](https://github.com/open-mmlab/mmsegmentation/pull/100))
- Support Windows install (experimental) ([#75](https://github.com/open-mmlab/mmsegmentation/pull/75))
- Add more OCRNet results ([#20](https://github.com/open-mmlab/mmsegmentation/pull/20))
@ -27,15 +30,23 @@ ResNeSt.
- Get version and githash automatically ([#55](https://github.com/open-mmlab/mmsegmentation/pull/55))
### v0.5.1 (11/08/2020)
**Highlights**
- Support FP16 and more generalized OHEM
**Bug Fixes**
- Fixed Pascal VOC conversion script (#19)
- Fixed OHEM weight assign bug (#54)
- Fixed palette type when palette is not given (#27)
**New Features**
- Support FP16 (#21)
- Generalized OHEM (#54)
**Improvements**
- Add load-from flag (#33)
- Fixed training tricks doc about different learning rates of model (#26)

4
docs/config.md
View File

@ -1,4 +1,5 @@
# Config System
We incorporate modular and inheritance design into our config system, which is convenient to conduct various experiments.
If you wish to inspect the config file, you may run `python tools/print_config.py /PATH/TO/CONFIG` to see the complete config.
You may also pass `--options xxx.yyy=zzz` to see updated config.
@ -325,6 +326,7 @@ The `_delete_=True` would replace all old keys in `backbone` field with new keys
Some intermediate variables are used in the configs files, like `train_pipeline`/`test_pipeline` in datasets.
It's worth noting that when modifying intermediate variables in the children configs, user need to pass the intermediate variables into corresponding fields again.
For example, we would like to change multi scale strategy to train/test a PSPNet. `train_pipeline`/`test_pipeline` are intermediate variable we would like modify.
```python
_base_ = '../pspnet/psp_r50_512x1024_40ki_cityscapes.py'
crop_size = (512, 1024)
@ -362,9 +364,11 @@ data = dict(
val=dict(pipeline=test_pipeline),
test=dict(pipeline=test_pipeline))
```
We first define the new `train_pipeline`/`test_pipeline` and pass them into `data`.
Similarly, if we would like to switch from `SyncBN` to `BN` or `MMSyncBN`, we need to substitute every `norm_cfg` in the config.
```python
_base_ = '../pspnet/psp_r50_512x1024_40ki_cityscpaes.py'
norm_cfg = dict(type='BN', requires_grad=True)

21
docs/getting_started.md
View File

@ -8,7 +8,7 @@ For installation instructions, please see [install.md](install.md).
It is recommended to symlink the dataset root to `$MMSEGMENTATION/data`.
If your folder structure is different, you may need to change the corresponding paths in config files.
```
```none
mmsegmentation
├── mmseg
├── tools
@ -50,21 +50,25 @@ mmsegmentation
```
### Cityscapes
The data could be found [here](https://www.cityscapes-dataset.com/downloads/) after registration.
By convention, `**labelTrainIds.png` are used for cityscapes training.
We provided a [scripts](https://github.com/open-mmlab/mmsegmentation/blob/master/tools/convert_datasets/cityscapes.py) based on [cityscapesscripts](https://github.com/mcordts/cityscapesScripts)
to generate `**labelTrainIds.png`.
```shell
# --nproc means 8 process for conversion, which could be omitted as well.
python tools/convert_datasets/cityscapes.py data/cityscapes --nproc 8
```
### Pascal VOC
Pascal VOC 2012 could be downloaded from [here](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/VOCtrainval_11-May-2012.tar).
Beside, most recent works on Pascal VOC dataset usually exploit extra augmentation data, which could be found [here](http://www.eecs.berkeley.edu/Research/Projects/CS/vision/grouping/semantic_contours/benchmark.tgz).
If you would like to use augmented VOC dataset, please run following command to convert augmentation annotations into proper format.
```shell
# --nproc means 8 process for conversion, which could be omitted as well.
python tools/convert_datasets/voc_aug.py data/VOCdevkit data/VOCdevkit/VOCaug --nproc 8
@ -72,12 +76,13 @@ python tools/convert_datasets/voc_aug.py data/VOCdevkit data/VOCdevkit/VOCaug --
Please refer to [concat dataset](https://github.com/open-mmlab/mmsegmentation/blob/master/docs/tutorials/new_dataset.md#concatenate-dataset) for details about how to concatenate them and train them together.
### ADE20K
The training and validation set of ADE20K could be download from this [link](http://data.csail.mit.edu/places/ADEchallenge/ADEChallengeData2016.zip).
We may also download test set from [here](http://data.csail.mit.edu/places/ADEchallenge/ADEChallengeData2016.zip).
### Pascal Context
The training and validation set of Pascal Context could be download from [here](http://host.robots.ox.ac.uk/pascal/VOC/voc2010/VOCtrainval_03-May-2010.tar). You may also download test set from [here](http://host.robots.ox.ac.uk:8080/eval/downloads/VOC2010test.tar) after registration.
To split the training and validation set from original dataset, you may download trainval_merged.json from [here](https://codalabuser.blob.core.windows.net/public/trainval_merged.json).
@ -110,12 +115,12 @@ python tools/test.py ${CONFIG_FILE} ${CHECKPOINT_FILE} [--out ${RESULT_FILE}] [-
```
Optional arguments:
- `RESULT_FILE`: Filename of the output results in pickle format. If not specified, the results will not be saved to a file.
- `EVAL_METRICS`: Items to be evaluated on the results. Allowed values depend on the dataset, e.g., `mIoU` is available for all dataset. Cityscapes could be evaluated by `cityscapes` as well as standard `mIoU` metrics.
- `--show`: If specified, segmentation results will be plotted on the images and shown in a new window. It is only applicable to single GPU testing and used for debugging and visualization. Please make sure that GUI is available in your environment, otherwise you may encounter the error like `cannot connect to X server`.
- `--show-dir`: If specified, segmentation results will be plotted on the images and saved to the specified directory. It is only applicable to single GPU testing and used for debugging and visualization. You do NOT need a GUI available in your environment for using this option.
Examples:
Assume that you have already downloaded the checkpoints to the directory `checkpoints/`.
@ -151,6 +156,7 @@ Assume that you have already downloaded the checkpoints to the directory `checkp
checkpoints/pspnet_r50-d8_512x1024_40k_cityscapes_20200605_003338-2966598c.pth \
4 --out results.pkl --eval mIoU cityscapes
```
Note: There is some gap (~0.1%) between cityscapes mIoU and our mIoU. The reason is that cityscapes average each class with class size by default.
We use the simple version without average for all datasets.
@ -164,6 +170,7 @@ Assume that you have already downloaded the checkpoints to the directory `checkp
img_dir='leftImg8bit/test',
ann_dir='gtFine/test'))
```
Then run test.
```shell
@ -175,7 +182,6 @@ Assume that you have already downloaded the checkpoints to the directory `checkp
You will get png files under `./pspnet_test_results` directory.
You may run `zip -r results.zip pspnet_test_results/` and submit the zip file to [evaluation server](https://www.cityscapes-dataset.com/submit/).
### Image demo
We provide a demo script to test a single image.
@ -191,7 +197,6 @@ python demo/image_demo.py demo/demo.jpg configs/pspnet/pspnet_r50-d8_512x1024_40
checkpoints/pspnet_r50-d8_512x1024_40k_cityscapes_20200605_003338-2966598c.pth --device cuda:0 --palette cityscapes
```
### High-level APIs for testing images
Here is an example of building the model and test given images.
@ -223,7 +228,6 @@ for frame in video:
A notebook demo can be found in [demo/inference_demo.ipynb](../demo/inference_demo.ipynb).
## Train a model
MMSegmentation implements distributed training and non-distributed training,
@ -233,6 +237,7 @@ All outputs (log files and checkpoints) will be saved to the working directory,
which is specified by `work_dir` in the config file.
By default we evaluate the model on the validation set after some iterations, you can change the evaluation interval by adding the interval argument in the training config.
```python
evaluation = dict(interval=4000) # This evaluate the model per 4000 iterations.
```
@ -264,6 +269,7 @@ Optional arguments are:
- `--load-from ${CHECKPOINT_FILE}`: Load weights from a checkpoint file (to start finetuning for another task).
Difference between `resume-from` and `load-from`:
- `resume-from` loads both the model weights and optimizer state including the iteration number.
- `load-from` loads only the model weights, starts the training from iteration 0.
@ -301,7 +307,6 @@ CUDA_VISIBLE_DEVICES=4,5,6,7 PORT=29501 ./tools/dist_train.sh ${CONFIG_FILE} 4
If you use `slurm_train.sh` to launch training jobs, you can set the port in commands with environment variable `MASTER_PORT`.
```shell
MASTER_PORT=29500 ./tools/slurm_train.sh ${PARTITION} ${JOB_NAME} ${CONFIG_FILE}
MASTER_PORT=29501 ./tools/slurm_train.sh ${PARTITION} ${JOB_NAME} ${CONFIG_FILE}
@ -321,7 +326,7 @@ python tools/get_flops.py ${CONFIG_FILE} [--shape ${INPUT_SHAPE}]
You will get the result like this.
```
```none
==============================
Input shape: (3, 2048, 1024)
Flops: 1429.68 GMac

13
docs/install.md
View File

@ -54,6 +54,7 @@ pip install -e .
```
Or simply:
```shell
pip install mmcv
```
@ -73,6 +74,7 @@ pip install git+https://github.com/open-mmlab/mmsegmentation.git # install the m
```
Instead, if you would like to install MMSegmentation in `dev` mode, run following
```shell
git clone https://github.com/open-mmlab/mmsegmentation.git
cd mmsegmentation
@ -82,19 +84,15 @@ pip install -e . # or "python setup.py develop"
Note:
1. When training or testing models on Windows, please ensure that all the '\\' in paths are replaced with '/'. Add .replace('\\', '/') to your python code wherever path strings occur.
2. The `version+git_hash` will also be saved in trained models meta, e.g. 0.5.0+c415a2e.
3. When MMsegmentation is installed on `dev` mode, any local modifications made to the code will take effect without the need to reinstall it.
4. If you would like to use `opencv-python-headless` instead of `opencv-python`,
you can install it before installing MMCV.
you can install it before installing MMCV.
5. Some dependencies are optional. Simply running `pip install -e .` will only install the minimum runtime requirements.
To use optional dependencies like `cityscapessripts` either install them manually with `pip install -r requirements/optional.txt` or specify desired extras when calling `pip` (e.g. `pip install -e .[optional]`). Valid keys for the extras field are: `all`, `tests`, `build`, and `optional`.
To use optional dependencies like `cityscapessripts` either install them manually with `pip install -r requirements/optional.txt` or specify desired extras when calling `pip` (e.g. `pip install -e .[optional]`). Valid keys for the extras field are: `all`, `tests`, `build`, and `optional`.
## A from-scratch setup script
### Linux
Here is a full script for setting up mmsegmentation with conda and link the dataset path (supposing that your dataset path is $DATA_ROOT).
@ -114,6 +112,7 @@ ln -s $DATA_ROOT data
```
### Windows(Experimental)
Here is a full script for setting up mmsegmentation with conda and link the dataset path (supposing that your dataset path is
%DATA_ROOT%. Notice: It must be an absolute path).

34
docs/model_zoo.md
View File

@ -4,24 +4,26 @@
* We use distributed training with 4 GPUs by default.
* All pytorch-style pretrained backbones on ImageNet are train by ourselves, with the same procedure in the [paper](https://arxiv.org/pdf/1812.01187.pdf).
Our ResNet style backbone are based on ResNetV1c variant, where the 7x7 conv in the input stem is replaced with three 3x3 convs.
Our ResNet style backbone are based on ResNetV1c variant, where the 7x7 conv in the input stem is replaced with three 3x3 convs.
* For the consistency across different hardwares, we report the GPU memory as the maximum value of `torch.cuda.max_memory_allocated()` for all 4 GPUs with `torch.backends.cudnn.benchmark=False`.
Note that this value is usually less than what `nvidia-smi` shows.
* We report the inference time as the total time of network forwarding and post-processing, excluding the data loading time.
Results are obtained with the script `tools/benchmark.py` which computes the average time on 200 images with `torch.backends.cudnn.benchmark=False`.
Results are obtained with the script `tools/benchmark.py` which computes the average time on 200 images with `torch.backends.cudnn.benchmark=False`.
* There are two inference modes in this framework.
* `slide` mode: The `test_cfg` will be like `dict(mode='slide', crop_size=(769, 769), stride=(513, 513))`.
In this mode, multiple patches will be cropped from input image, passed into network individually.
The crop size and stride between patches are specified by `crop_size` and `stride`.
The overlapping area will be merged by average
* `whole` mode: The `test_cfg` will be like `dict(mode='whole')`.
* `slide` mode: The `test_cfg` will be like `dict(mode='slide', crop_size=(769, 769), stride=(513, 513))`.
In this mode, the whole imaged will be passed into network directly.
In this mode, multiple patches will be cropped from input image, passed into network individually.
The crop size and stride between patches are specified by `crop_size` and `stride`.
The overlapping area will be merged by average
* `whole` mode: The `test_cfg` will be like `dict(mode='whole')`.
In this mode, the whole imaged will be passed into network directly.
By default, we use `slide` inference for 769x769 trained model, `whole` inference for the rest.
* For input size of 8x+1 (e.g. 769), `align_corner=True` is adopted as a traditional practice.
Otherwise, for input size of 8x (e.g. 512, 1024), `align_corner=False` is adopted.
Otherwise, for input size of 8x (e.g. 512, 1024), `align_corner=False` is adopted.
## Baselines
@ -117,16 +119,16 @@ Please refer [Mixed Precision (FP16) Training](https://github.com/open-mmlab/mms
### Hardware
- 8 NVIDIA Tesla V100 (32G) GPUs
- Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz
* 8 NVIDIA Tesla V100 (32G) GPUs
* Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz
### Software environment
- Python 3.7
- PyTorch 1.5
- CUDA 10.1
- CUDNN 7.6.03
- NCCL 2.4.08
* Python 3.7
* PyTorch 1.5
* CUDA 10.1
* CUDNN 7.6.03
* NCCL 2.4.08
### Training speed

15
docs/tutorials/data_pipeline.md
View File

@ -56,56 +56,71 @@ For each operation, we list the related dict fields that are added/updated/remov
### Data loading
`LoadImageFromFile`
- add: img, img_shape, ori_shape
`LoadAnnotations`
- add: gt_semantic_seg, seg_fields
### Pre-processing
`Resize`
- add: scale, scale_idx, pad_shape, scale_factor, keep_ratio
- update: img, img_shape, *seg_fields
`RandomFlip`
- add: flip
- update: img, *seg_fields
`Pad`
- add: pad_fixed_size, pad_size_divisor
- update: img, pad_shape, *seg_fields
`RandomCrop`
- update: img, pad_shape, *seg_fields
`Normalize`
- add: img_norm_cfg
- update: img
`SegRescale`
- update: gt_semantic_seg
`PhotoMetricDistortion`
- update: img
### Formatting
`ToTensor`
- update: specified by `keys`.
`ImageToTensor`
- update: specified by `keys`.
`Transpose`
- update: specified by `keys`.
`ToDataContainer`
- update: specified by `fields`.
`DefaultFormatBundle`
- update: img, gt_semantic_seg
`Collect`
- add: img_meta (the keys of img_meta is specified by `meta_keys`)
- remove: all other keys except for those specified by `keys`

14
docs/tutorials/new_dataset.md
View File

@ -5,7 +5,8 @@
The simplest way is to convert your dataset to organize your data into folders.
An example of file structure is as followed.
```
```none
├── data
│ ├── my_dataset
│ │ ├── img_dir
@ -22,16 +23,19 @@ An example of file structure is as followed.
│ │ │ ├── val
```
A training pair will consist of the files with same suffix in img_dir/ann_dir.
If `split` argument is given, only part of the files in img_dir/ann_dir will be loaded.
We may specify the prefix of files we would like to be included in the split txt.
More specifically, for a split txt like following,
```
```none
xxx
zzz
```
Only
`data/my_dataset/img_dir/train/xxx{img_suffix}`,
`data/my_dataset/img_dir/train/zzz{img_suffix}`,
@ -50,6 +54,7 @@ Currently it supports to concat and repeat datasets.
We use `RepeatDataset` as wrapper to repeat the dataset.
For example, suppose the original dataset is `Dataset_A`, to repeat it, the config looks like the following
```python
dataset_A_train = dict(
type='RepeatDataset',
@ -70,6 +75,7 @@ There 2 ways to concatenate the dataset.
you can concatenate the dataset configs like the following.
1. You may concatenate two `ann_dir`.
```python
dataset_A_train = dict(
type='Dataset_A',
@ -78,6 +84,7 @@ There 2 ways to concatenate the dataset.
pipeline=train_pipeline
)
```
2. You may concatenate two `split`.
```python
@ -89,6 +96,7 @@ There 2 ways to concatenate the dataset.
pipeline=train_pipeline
)
```
3. You may concatenate two `ann_dir` and `split` simultaneously.
```python
@ -100,6 +108,7 @@ There 2 ways to concatenate the dataset.
pipeline=train_pipeline
)
```
In this case, `ann_dir_1` and `ann_dir_2` are corresponding to `split_1.txt` and `split_2.txt`.
2. In case the dataset you want to concatenate is different, you can concatenate the dataset configs like the following.
@ -120,7 +129,6 @@ There 2 ways to concatenate the dataset.
)
```
A more complex example that repeats `Dataset_A` and `Dataset_B` by N and M times, respectively, and then concatenates the repeated datasets is as the following.
```python

10
docs/tutorials/training_tricks.md
View File

@ -7,29 +7,36 @@ MMSegmentation support following training tricks out of box.
In semantic segmentation, some methods make the LR of heads larger than backbone to achieve better performance or faster convergence.
In MMSegmentation, you may add following lines to config to make the LR of heads 10 times of backbone.
```python
optimizer=dict(
paramwise_cfg = dict(
custom_keys={
'head': dict(lr_mult=10.)}))
```
With this modification, the LR of any parameter group with `'head'` in name will be multiplied by 10.
You may refer to [MMCV doc](https://mmcv.readthedocs.io/en/latest/api.html#mmcv.runner.DefaultOptimizerConstructor) for further details.
## Online Hard Example Mining (OHEM)
We implement pixel sampler [here](https://github.com/open-mmlab/mmsegmentation/tree/master/mmseg/core/seg/sampler) for training sampling.
Here is an example config of training PSPNet with OHEM enabled.
```python
_base_ = './pspnet_r50-d8_512x1024_40k_cityscapes.py'
model=dict(
decode_head=dict(
sampler=dict(type='OHEMPixelSampler', thresh=0.7, min_kept=100000)) )
```
In this way, only pixels with confidence score under 0.7 are used to train. And we keep at least 100000 pixels during training. If `thresh` is not specified, pixels of top ``min_kept`` loss will be selected.
## Class Balanced Loss
For dataset that is not balanced in classes distribution, you may change the loss weight of each class.
Here is an example for cityscapes dataset.
```python
_base_ = './pspnet_r50-d8_512x1024_40k_cityscapes.py'
model=dict(
@ -41,4 +48,5 @@ model=dict(
1.0489, 0.8786, 1.0023, 0.9539, 0.9843, 1.1116, 0.9037,
1.0865, 1.0955, 1.0865, 1.1529, 1.0507])))
```
`class_weight` will be passed into `CrossEntropyLoss` as `weight` argument. Please refer to [PyTorch Doc ](https://pytorch.org/docs/stable/nn.html?highlight=crossentropy#torch.nn.CrossEntropyLoss) for details.
`class_weight` will be passed into `CrossEntropyLoss` as `weight` argument. Please refer to [PyTorch Doc](https://pytorch.org/docs/stable/nn.html?highlight=crossentropy#torch.nn.CrossEntropyLoss) for details.