Release MMSegmentation v1.1.0

.circleci/test.yml

@@ -61,10 +61,11 @@ jobs:
           command: |
             pip install git+https://github.com/open-mmlab/mmengine.git@main
             pip install -U openmim
-            mim install 'mmcv>=2.0.0rc4'
-            pip install mmcls==1.0.0rc6
-            pip install git+https://github.com/open-mmlab/mmdetection.git@main
+            mim install mmcv>=2.0.0
+            pip install mmpretrain>=1.0.0rc7
+            pip install mmdet>=3.0.0
             pip install -r requirements/tests.txt -r requirements/optional.txt
+            python -m pip install albumentations>=0.3.2 --no-binary qudida,albumentations
       - run:
           name: Build and install
           command: |
@@ -96,21 +97,21 @@ jobs:
           name: Clone Repos
           command: |
             git clone -b main --depth 1 https://github.com/open-mmlab/mmengine.git /home/circleci/mmengine
-            git clone -b dev-3.x --depth 1 https://github.com/open-mmlab/mmdetection.git /home/circleci/mmdetection
       - run:
           name: Build Docker image
           command: |
             docker build .circleci/docker -t mmseg:gpu --build-arg PYTORCH=<< parameters.torch >> --build-arg CUDA=<< parameters.cuda >> --build-arg CUDNN=<< parameters.cudnn >>
-            docker run --gpus all -t -d -v /home/circleci/project:/mmseg -v /home/circleci/mmengine:/mmengine -v /home/circleci/mmclassification:/mmclassification -v /home/circleci/mmdetection:/mmdetection -w /mmseg --name mmseg mmseg:gpu
+            docker run --gpus all -t -d -v /home/circleci/project:/mmseg -v /home/circleci/mmengine:/mmengine -v /home/circleci/mmpretrain:/mmpretrain -v /home/circleci/mmdetection:/mmdetection -w /mmseg --name mmseg mmseg:gpu
       - run:
           name: Install mmseg dependencies
           command: |
             docker exec mmseg pip install -e /mmengine
             docker exec mmseg pip install -U openmim
-            docker exec mmseg mim install 'mmcv>=2.0.0rc4'
-            docker exec mmseg pip install mmcls==1.0.0rc6
-            docker exec mmseg pip install -e /mmdetection
+            docker exec mmseg mim install mmcv>=2.0.0
+            docker exec mmseg pip install mmpretrain>=1.0.0rc7
+            docker exec mmseg mim install mmdet>=3.0.0
             docker exec mmseg pip install -r requirements/tests.txt -r requirements/optional.txt
+            docker exec mmseg python -m pip install albumentations>=0.3.2 --no-binary qudida,albumentations
       - run:
           name: Build and install
           command: |
@@ -144,8 +145,8 @@ workflows:
                 - main
       - build_cpu:
           name: minimum_version_cpu
-          torch: 1.6.0
-          torchvision: 0.7.0
+          torch: 1.8.1
+          torchvision: 0.9.1
           python: "3.7"
           requires:
             - lint
@@ -176,10 +177,10 @@ workflows:
     jobs:
       - build_cuda:
           name: minimum_version_gpu
-          torch: 1.6.0
+          torch: 1.8.1
           # Use double quotation mark to explicitly specify its type
           # as string instead of number
-          cuda: "10.1"
+          cuda: "10.2"
           filters:
             branches:
               only:

.dev_scripts/benchmark_full_models.txt

@@ -0,0 +1,56 @@
+ann/ann_r50-d8_4xb2-40k_cityscapes-512x1024.py
+apcnet/apcnet_r50-d8_4xb4-80k_ade20k-512x512.py
+beit/beit-large_upernet_8xb1-amp-160k_ade20k-640x640.py
+bisenetv1/bisenetv1_r18-d32_4xb4-160k_coco-stuff164k-512x512.py
+bisenetv2/bisenetv2_fcn_4xb4-ohem-160k_cityscapes-1024x1024.py
+ccnet/ccnet_r50-d8_4xb2-40k_cityscapes-512x1024.py
+cgnet/cgnet_fcn_4xb4-60k_cityscapes-680x680.py
+convnext/convnext-tiny_upernet_8xb2-amp-160k_ade20k-512x512.py
+danet/danet_r50-d8_4xb2-40k_cityscapes-512x1024.py
+deeplabv3/deeplabv3_r50-d8_4xb2-40k_cityscapes-512x1024.py
+deeplabv3plus/deeplabv3plus_r101-d8_4xb2-40k_cityscapes-769x769.py
+dmnet/dmnet_r50-d8_4xb2-40k_cityscapes-512x1024.py
+dnlnet/dnl_r50-d8_4xb2-40k_cityscapes-512x1024.py
+dpt/dpt_vit-b16_8xb2-160k_ade20k-512x512.py
+emanet/eemanet_r50-d8_4xb2-80k_cityscapes-512x1024.py
+encnet/encnet_r50-d8_4xb4-80k_ade20k-512x512.py
+erfnet/erfnet_fcn_4xb4-160k_cityscapes-512x1024.py
+fastfcn/fastfcn_r50-d32_jpu_aspp_4xb2-80k_cityscapes-512x1024.py
+fastscnn/fast_scnn_8xb4-160k_cityscapes-512x1024.py
+fcn/fcn_r50-d8_4xb2-40k_cityscapes-512x1024.py
+gcnet/gcnet_r50-d8_4xb2-40k_cityscapes-769x769.py
+hrnet/fcn_hr18s_4xb4-80k_ade20k-512x512.py
+icnet/icnet_r50-d8_4xb2-80k_cityscapes-832x832.py
+isanet/isanet_r50-d8_4xb2-80k_cityscapes-512x1024.py
+knet/knet-s3_swin-l_upernet_8xb2-adamw-80k_ade20k-640x640.py
+mae/mae-base_upernet_8xb2-amp-160k_ade20k-512x512.py
+mask2former/mask2former_r50_8xb2-90k_cityscapes-512x1024.py
+mask2former/mask2former_swin-t_8xb2-90k_cityscapes-512x1024.py
+mask2former/mask2former_swin-s_8xb2-160k_ade20k-512x512.py
+mask2former/mask2former_swin-l-in22k-384x384-pre_8xb2-160k_ade20k-640x640.py
+maskformer/maskformer_r50-d32_8xb2-160k_ade20k-512x512.py
+maskformer/maskformer_r101-d32_8xb2-160k_ade20k-512x512.py
+maskformer/maskformer_swin-t_upernet_8xb2-160k_ade20k-512x512.py
+maskformer/maskformer_swin-s_upernet_8xb2-160k_ade20k-512x512.py
+mobilenet_v2/mobilenet-v2-d8_pspnet_4xb2-80k_cityscapes-512x1024.py
+mobilenet_v3/mobilenet-v3-d8-scratch_lraspp_4xb4-320k_cityscapes-512x1024.py
+nonlocal_net/nonlocal_r50-d8_4xb2-40k_cityscapes-512x1024.py
+ocrnet/ocrnet_hr18_4xb4-80k_ade20k-512x512.py
+pidnet/pidnet-l_2xb6-120k_1024x1024-cityscapes.py
+point_rend/pointrend_r50_4xb2-80k_cityscapes-512x1024.py
+poolformer/fpn_poolformer_m48_8xb4-40k_ade20k-512x512.py
+psanet/psanet_r101-d8_4xb4-80k_ade20k-512x512.py
+pspnet/pspnet_r50-d8-rsb_4xb2-adamw-80k_cityscapes-512x1024.py
+resnest/resnest_s101-d8_deeplabv3_4xb4-160k_ade20k-512x512.py
+segformer/segformer_mit-b5_8xb1-160k_cityscapes-1024x1024.py
+segformer/segformer_mit-b5_8xb2-160k_ade20k-512x512.py
+segmenter/segmenter_vit-t_mask_8xb1-160k_ade20k-512x512.py
+segnext/segnext_mscan-t_1xb16-adamw-160k_ade20k-512x512.py
+sem_fpn/fpn_r101_4xb2-80k_cityscapes-512x1024.py
+setr/setr_vit-l-mla_8xb1-160k_ade20k-512x512.py
+stdc/stdc1_in1k-pre_4xb12-80k_cityscapes-512x1024.py
+swin/swin-small-patch4-window7-in1k-pre_upernet_8xb2-160k_ade20k-512x512.py
+twins/twins_pcpvt-s_uperhead_8xb4-160k_ade20k-512x512.py
+unet/unet-s5-d16_fcn_4xb4-160k_cityscapes-512x1024.py
+upernet/upernet_r50_4xb2-40k_cityscapes-769x769.py
+vit/vit_deit-s16_upernet_8xb2-80k_ade20k-512x512.py

.dev_scripts/benchmark_options.py

@@ -0,0 +1,10 @@
+third_part_libs = [
+    'pip install mmengine',
+    'pip install mmcv>=2.0.0',
+    'pip install mmcls==1.0.0rc6',
+    'pip install mmdet==3.0.0',
+    'pip install -r requirements.txt',
+    'pip install timm',
+]
+
+default_floating_range = 0.5

.dev_scripts/benchmark_train_models.txt

@@ -0,0 +1,25 @@
+bisenetv1/bisenetv1_r18-d32_4xb4-160k_coco-stuff164k-512x512.py
+bisenetv2/bisenetv2_fcn_4xb4-ohem-160k_cityscapes-1024x1024.py
+deeplabv3/deeplabv3_r50-d8_4xb2-40k_cityscapes-512x1024.py
+deeplabv3/deeplabv3_r101-d8_4xb4-160k_ade20k-512x512.py
+deeplabv3plus/deeplabv3plus_r101-d8_4xb2-40k_cityscapes-769x769.py
+deeplabv3plus/deeplabv3plus_r50-d8_4xb4-80k_ade20k-512x512.py
+hrnet/fcn_hr18s_4xb4-80k_ade20k-512x512.py
+hrnet/fcn_hr18_4xb2-160k_cityscapes-512x1024.py
+knet/knet-s3_swin-l_upernet_8xb2-adamw-80k_ade20k-512x512.py
+knet/knet-s3_r50-d8_pspnet_8xb2-adamw-80k_ade20k-512x512.py
+mae/mae-base_upernet_8xb2-amp-160k_ade20k-512x512.py
+mask2former/mask2former_swin-l-in22k-384x384-pre_8xb2-160k_ade20k-640x640.py
+maskformer/maskformer_swin-s_upernet_8xb2-160k_ade20k-512x512.py
+mobilenet_v2/mobilenet-v2-d8_deeplabv3_4xb2-80k_cityscapes-512x1024.py
+ocrnet/ocrnet_hr48_4xb4-80k_ade20k-512x512.py
+pidnet/pidnet-m_2xb6-120k_1024x1024-cityscapes.py
+pspnet/pspnet_r50-d8-rsb_4xb2-adamw-80k_cityscapes-512x1024.py
+pspnet/pspnet_r101-d8_4xb4-80k_ade20k-512x512.py
+segformer/segformer_mit-b5_8xb2-160k_ade20k-512x512.py
+segmenter/segmenter_vit-t_mask_8xb1-160k_ade20k-512x512.py
+segnext/segnext_mscan-t_1xb16-adamw-160k_ade20k-512x512.py
+swin/swin-base-patch4-window12-in22k-384x384-pre_upernet_8xb2-160k_ade20k-512x512.py
+twins/twins_pcpvt-l_uperhead_8xb2-160k_ade20k-512x512.py
+unet/unet-s5-d16_fcn_4xb4-160k_cityscapes-512x1024.py
+upernet/upernet_r101_4xb2-40k_cityscapes-512x1024.py

.dev/generate_benchmark_evaluation_script.py → .dev_scripts/generate_benchmark_evaluation_script.py

@@ -17,7 +17,7 @@ def parse_args():
     parser.add_argument(
         '--out',
         type=str,
-        default='.dev/benchmark_evaluation.sh',
+        default='.dev_scripts/benchmark_evaluation.sh',
         help='path to save model benchmark script')
 
     args = parser.parse_args()

.dev/generate_benchmark_train_script.py → .dev_scripts/generate_benchmark_train_script.py

@@ -19,7 +19,7 @@ def parse_args():
     parser.add_argument(
         '--out',
         type=str,
-        default='.dev/benchmark_train.sh',
+        default='.dev_scripts/benchmark_train.sh',
         help='path to save model benchmark script')
 
     args = parser.parse_args()

.owners.yml

@@ -5,8 +5,3 @@ assign:
     daily-shift-based
   assignees:
     - xiexinch
-    - csatsurnh
-    - xiexinch
-    - csatsurnh
-    - xiexinch
-    - csatsurnh

.pre-commit-config.yaml

@@ -47,7 +47,7 @@ repos:
       - id: update-model-index
         name: update-model-index
         description: Collect model information and update model-index.yml
-        entry: .dev/update_model_index.py
+        entry: .dev_scripts/update_model_index.py
         additional_dependencies: [pyyaml]
         language: python
         require_serial: true

MANIFEST.in

@@ -1,4 +1,5 @@
 include requirements/*.txt
 include mmseg/.mim/model-index.yml
+include mmaction/.mim/dataset-index.yml
 recursive-include mmseg/.mim/configs *.py *.yaml
 recursive-include mmseg/.mim/tools *.py *.sh

README.md

@@ -88,12 +88,12 @@ MMSegmentation v1.x brings remarkable improvements over the 0.x release, offerin
 
 ## What's New
 
-v1.0.0 was released on 04/06/2023.
+v1.1.0 was released on 07/04/2023.
 Please refer to [changelog.md](docs/en/notes/changelog.md) for details and release history.
 
-- Add Mapillary Vistas Datasets support to MMSegmentation Core Package ([#2576](https://github.com/open-mmlab/mmsegmentation/pull/2576))
-- Support PIDNet ([#2609](https://github.com/open-mmlab/mmsegmentation/pull/2609))
-- Support SegNeXt ([#2654](https://github.com/open-mmlab/mmsegmentation/pull/2654))
+- Support 24 medical image datasets in [projects](./projects/medical/).
+- Add GDAL backend and support remote sensing datasets [LEVIR-CD](https://github.com/open-mmlab/mmsegmentation/pull/2903).
+- Support [DDRNet](https://github.com/open-mmlab/mmsegmentation/pull/2855).
 
 ## Installation
 
@@ -110,6 +110,60 @@ A Colab tutorial is also provided. You may preview the notebook [here](demo/MMSe
 
 To migrate from MMSegmentation 0.x, please refer to [migration](docs/en/migration).
 
+## Tutorial
+
+<details>
+<summary>Get Started</summary>
+
+- [MMSeg overview](docs/en/overview.md)
+- [MMSeg Installation](docs/en/get_started.md)
+- [FAQ](docs/en/notes/faq.md)
+
+</details>
+
+<details>
+<summary>MMSeg Basic Tutorial</summary>
+
+- [Tutorial 1: Learn about Configs](docs/en/user_guides/1_config.md)
+- [Tutorial 2: Prepare datasets](docs/en/user_guides/2_dataset_prepare.md)
+- [Tutorial 3: Inference with existing models](docs/en/user_guides/3_inference.md)
+- [Tutorial 4: Train and test with existing models](docs/en/user_guides/4_train_test.md)
+- [Tutorial 5: Model deployment](docs/en/user_guides/5_deployment.md)
+- [Useful Tools](docs/en/user_guides/useful_tools.md)
+- [Feature Map Visualization](docs/en/user_guides/visualization_feature_map.md)
+- [Visualization](docs/en/user_guides/visualization.md)
+
+</details>
+
+<details>
+<summary>MMSeg Detail Tutorial</summary>
+
+- [MMSeg Dataset](docs/en/advanced_guides/datasets.md)
+- [MMSeg Models](docs/en/advanced_guides/models.md)
+- [MMSeg Dataset Structures](docs/en/advanced_guides/structures.md)
+- [MMSeg Data Transforms](docs/en/advanced_guides/transforms.md)
+- [MMSeg Dataflow](docs/en/advanced_guides/data_flow.md)
+- [MMSeg Training Engine](docs/en/advanced_guides/engine.md)
+- [MMSeg Evaluation](docs/en/advanced_guides/evaluation.md)
+
+</details>
+
+<details>
+<summary>MMSeg Development Tutorial</summary>
+
+- [Add New Datasets](docs/en/advanced_guides/add_datasets.md)
+- [Add New Metrics](docs/en/advanced_guides/add_metrics.md)
+- [Add New Modules](docs/en/advanced_guides/add_models.md)
+- [Add New Data Transforms](docs/en/advanced_guides/add_transforms.md)
+- [Customize Runtime Settings](docs/en/advanced_guides/customize_runtime.md)
+- [Training Tricks](docs/en/advanced_guides/training_tricks.md)
+- [Contribute code to MMSeg](.github/CONTRIBUTING.md)
+- [Contribute a standard dataset in projects](docs/zh_cn/advanced_guides/contribute_dataset.md)
+- [NPU (HUAWEI Ascend)](docs/en/device/npu.md)
+- [0.x → 1.x migration](docs/en/migration/interface.md)，[0.x → 1.x package](docs/en/migration/package.md)
+
+</details>
+
 ## Benchmark and model zoo
 
 Results and models are available in the [model zoo](docs/en/model_zoo.md).
@@ -174,6 +228,7 @@ Results and models are available in the [model zoo](docs/en/model_zoo.md).
 - [x] [MaskFormer (NeurIPS'2021)](configs/maskformer)
 - [x] [Mask2Former (CVPR'2022)](configs/mask2former)
 - [x] [PIDNet (ArXiv'2022)](configs/pidnet)
+- [x] [DDRNet (T-ITS'2022)](configs/ddrnet)
 
 </details>
 
@@ -197,6 +252,7 @@ Results and models are available in the [model zoo](docs/en/model_zoo.md).
 - [x] [Vaihingen](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/en/user_guides/2_dataset_prepare.md#isprs-vaihingen)
 - [x] [iSAID](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/en/user_guides/2_dataset_prepare.md#isaid)
 - [x] [Mapillary Vistas](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/en/user_guides/2_dataset_prepare.md#mapillary-vistas-datasets)
+- [x] [LEVIR-CD](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/en/user_guides/2_dataset_prepare.md#levir-cd)
 
 </details>
 

README_zh-CN.md

@@ -87,7 +87,7 @@ MMSegmentation v1.x 在 0.x 版本的基础上有了显著的提升，提供了
 
 ## 更新日志
 
-最新版本 v1.0.0 在 2023.04.06 发布。
+最新版本 v1.1.0 在 2023.07.04 发布。
 如果想了解更多版本更新细节和历史信息，请阅读[更新日志](docs/en/notes/changelog.md)。
 
 ## 安装
@@ -104,6 +104,60 @@ MMSegmentation v1.x 在 0.x 版本的基础上有了显著的提升，提供了
 
 若需要将 0.x 版本的代码迁移至新版，请参考[迁移文档](docs/zh_cn/migration)。
 
+## 教程文档
+
+<details>
+<summary>开启 MMSeg 之旅</summary>
+
+- [MMSeg 概述](docs/zh_cn/overview.md)
+- [安装和验证](docs/zh_cn/get_started.md)
+- [常见问题解答](docs/zh_cn/notes/faq.md)
+
+</details>
+
+<details>
+<summary>MMSeg 快速入门教程</summary>
+
+- [教程1：了解配置文件](docs/zh_cn/user_guides/1_config.md)
+- [教程2：准备数据集](docs/zh_cn/user_guides/2_dataset_prepare.md)
+- [教程3：使用预训练模型推理](docs/zh_cn/user_guides/3_inference.md)
+- [教程4：使用现有模型进行训练和测试](docs/zh_cn/user_guides/4_train_test.md)
+- [教程5：模型部署](docs/zh_cn/user_guides/5_deployment.md)
+- [常用工具](docs/zh_cn/user_guides/useful_tools.md)
+- [特征图可视化](docs/zh_cn/user_guides/visualization_feature_map.md)
+- [可视化](docs/zh_cn/user_guides/visualization.md)
+
+</details>
+
+<details>
+<summary>MMSeg 细节介绍</summary>
+
+- [MMSeg 数据集介绍](docs/zh_cn/advanced_guides/datasets.md)
+- [MMSeg 模型介绍](docs/zh_cn/advanced_guides/models.md)
+- [MMSeg 数据结构介绍](docs/zh_cn/advanced_guides/structures.md)
+- [MMSeg 数据增强介绍](docs/zh_cn/advanced_guides/transforms.md)
+- [MMSeg 数据流介绍](docs/zh_cn/advanced_guides/data_flow.md)
+- [MMSeg 训练引擎介绍](docs/zh_cn/advanced_guides/engine.md)
+- [MMSeg 模型评测介绍](docs/zh_cn/advanced_guides/evaluation.md)
+
+</details>
+
+<details>
+<summary>MMSeg 开发教程</summary>
+
+- [新增自定义数据集](docs/zh_cn/advanced_guides/add_datasets.md)
+- [新增评测指标](docs/zh_cn/advanced_guides/add_metrics.md)
+- [新增自定义模型](docs/zh_cn/advanced_guides/add_models.md)
+- [新增自定义数据增强](docs/zh_cn/advanced_guides/add_transforms.md)
+- [自定义运行设定](docs/zh_cn/advanced_guides/customize_runtime.md)
+- [训练技巧](docs/zh_cn/advanced_guides/training_tricks.md)
+- [如何给 MMSeg贡献代码](.github/CONTRIBUTING.md)
+- [在 projects 给 MMSeg 贡献一个标准数据集](docs/zh_cn/advanced_guides/contribute_dataset.md)
+- [NPU (华为 昇腾)](docs/zh_cn/device/npu.md)
+- [0.x → 1.x 迁移文档](docs/zh_cn/migration/interface.md)，[0.x → 1.x 库变更文档](docs/zh_cn/migration/package.md)
+
+</details>
+
 ## 基准测试和模型库
 
 测试结果和模型可以在[模型库](docs/zh_cn/model_zoo.md)中找到。
@@ -168,6 +222,7 @@ MMSegmentation v1.x 在 0.x 版本的基础上有了显著的提升，提供了
 - [x] [MaskFormer (NeurIPS'2021)](configs/maskformer)
 - [x] [Mask2Former (CVPR'2022)](configs/mask2former)
 - [x] [PIDNet (ArXiv'2022)](configs/pidnet)
+- [x] [DDRNet (T-ITS'2022)](configs/ddrnet)
 
 </details>
 
@@ -191,6 +246,7 @@ MMSegmentation v1.x 在 0.x 版本的基础上有了显著的提升，提供了
 - [x] [Vaihingen](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/zh_cn/dataset_prepare.md#isprs-vaihingen)
 - [x] [iSAID](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/zh_cn/dataset_prepare.md#isaid)
 - [x] [Mapillary Vistas](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/en/user_guides/2_dataset_prepare.md#mapillary-vistas-datasets)
+- [x] [LEVIR-CD](https://github.com/open-mmlab/mmsegmentation/blob/main/docs/en/user_guides/2_dataset_prepare.md#levir-cd)
 
 </details>