Skip to content

Releases: open-mmlab/mmsegmentation

MMSegmentation v0.30.0 Release

11 Jan 10:22
ed83982
Compare
Choose a tag to compare

v0.30.0 (01/11/2023)

New Features

  • Support Delving into High-Quality Synthetic Face Occlusion Segmentation Datasets (#2194)

Bug Fixes

  • Fix incorrect test_cfg setting in UNet base configs (#2347)
  • Fix KNet IterativeDecodeHead bug in master branch (#2333)
  • Fix deadlock issue related with MMSegWandbHook (#2398)

Enhancement

  • Update CI and pre-commit checking (#2309,#2331)
  • Add Projects/ folder, and the first example project in 0.x (#2457)
  • Fix the deprecation of np.float and CI configuration problems (#2451)

Documentation

  • Add high quality synthetic face occlusion dataset link to readme (#2453)
  • Fix the docstring error in the PascalContextDataset59 class (#2450)

Contributors

MMSegmentation v1.0.0rc3 Released

31 Dec 11:29
Compare
Choose a tag to compare
Pre-release

What's new

Highlights

  • Support test time augmentation (#2184)
  • Add 'Projects/' folder and the first example project (#2412)

Features

  • Add Biomedical 3D array random crop transform (#2378)

Documentation

  • Add Chinese version of config tutorial (#2371)
  • Add Chinese version of train & test tutorial (#2355)
  • Add Chinese version of overview ((#2397)))
  • Add Chinese version of get_started (#2417)
  • Add datasets in Chinese (#2387)
  • Add dataflow document (#2403)
  • Add pspnet model structure graph (#2437)
  • Update some content of engine Chinese documentation (#2341)
  • Update TTA to migration documentation (#2335)

Bug fix

  • Remove dependency mmdet when do not use MaskFormerHead and MMDET_Mask2FormerHead (#2448)

Enhancement

  • Add torch1.13 checking in CI (#2402)
  • Fix pytorch version for merge stage test (#2449)

New Contributors

MMSegmentation v1.0.0rc2 Released

06 Dec 09:28
8a611e1
Compare
Choose a tag to compare
Pre-release

What's new

Highlights

  • Support MaskFormer (#2215)
  • Support Mask2Former (#2255)

Features

  • Add ResizeShortestEdge transform (#2339)
  • Support padding in data pre-processor for model testing(#2290)
  • Fix the problem of post-processing not removing padding (#2367)

Bug fix

  • Fix links in README (#2024)
  • Fix swin load state_dict (#2304)
  • Fix typo of BaseSegDataset docstring (#2322)
  • Fix the bug in the visualization step (#2326)
  • Fix ignore class id from -1 to 255 in BaseSegDataset (#2332)
  • Fix KNet IterativeDecodeHead bug (#2334)
  • Add input argument for datasets (#2379)
  • Fix typo in warning on binary classification (#2382)

Enhancement

  • Fix ci for 1.x (#2011, #2019)
  • Fix lint and pre-commit hook (#2308)
  • Add data string in .gitignore file in dev-1.x branch (#2336)
  • Make scipy as a default dependency in runtime (#2362)
  • Delete mmcls in runtime.txt (#2368)

Documentation

  • Update configuration documentation (#2048)
  • Update inference documentation (#2052)
  • Update the documentation for model training and testing (#2061)
  • Update get started documentation (#2148)
  • Update transforms documentation (#2088)
  • Add MMEval projects like in README (#2259)
  • Translate the visualization documentation (#2298)

New Contributors

MMSegmentation v0.29.1 Release

03 Nov 08:31
7b09967
Compare
Choose a tag to compare

v0.29.1 (11/3/2022)

New Features

  • Add model ensemble tools (#2218)

Bug Fixes

  • Use SyncBN in MobileNetV2 (#2207)

Documentation

  • Update FAQ doc about binary segmentation and ReduceZeroLabel (#2206)
  • Fix typos (#2249)
  • Fix model results (#2190, #2114)

Contributors

MMSegmentation v1.0.0rc1 Released

02 Nov 09:49
eafb5da
Compare
Choose a tag to compare
Pre-release

Changelog

v1.0.0rc1 (2/11/2022)

Highlights

  • Support PoolFormer (#2191)
  • Add Decathlon dataset (#2227)

Features

  • Add BioMedical data loading (#2176)
  • Add LIP dataset (#2251)
  • GenerateEdge data transform (#2210)

Bug fix

  • Fix segmenter-vit-s_fcn config (#2037)
  • Fix binary segmentation (#2101)
  • Fix MMSegmentation colab demo (#2089)
  • Fix ResizeToMultiple transform (#2185)
  • Use SyncBN in mobilenet_v2 (#2198)
  • Fix typo in installation (#2175)
  • Fix typo in visualization.md (#2116)

Enhancement

  • Add mim extras_requires in setup.py (#2012)
  • Fix CI (#2029)
  • Remove ops module (#2063)
  • Add pyupgrade pre-commit hook (#2078)
  • Add out_file in add_datasample of SegLocalVisualizer to directly save image (#2090)
  • Upgrade pre commit hooks (#2154)
  • Ignore test timm in CI when torch<1.7 (#2158)
  • Update requirements (#2186)
  • Fix Windows platform CI (#2202)

Documentation

  • Add Overview documentation (#2042)
  • Add Evaluation documentation (#2077)
  • Add Migration documentation (#2066)
  • Add Structures documentation (#2070)
  • Add Structures ZN documentation (#2129)
  • Add Engine ZN documentation (#2157)
  • Update Prepare datasets and Visualization doc (#2054)
  • Update Models documentation (#2160)
  • Update Add New Modules documentation (#2067)
  • Fix the installation commands in get_started.md (#2174)
  • Add MMYOLO to README.md (#2220)

New Contributors

MMSegmentation v0.29.0 Release

10 Oct 06:13
356215b
Compare
Choose a tag to compare

Changelog

v0.29.0 (10/10/2022)

New Features

  • Support PoolFormer (CVPR'2022) (#1537)

Enhancement

  • Improve structure and readability for FCNHead (#2142)
  • Support IterableDataset in distributed training (#2151)
  • Upgrade .dev scripts (#2020)
  • Upgrade pre-commit hooks (#2155)

Bug Fixes

  • Fix mmseg.api.inference inference_segmentor (#1849)
  • fix bug about label_map in evaluation part (#2075)
  • Add missing dependencies to torchserve docker file (#2133)
  • Fix ddp unittest (#2060)

Contributors

MMSegmentation Release v0.28.0

08 Sep 08:59
b51670b
Compare
Choose a tag to compare

Changelog

V0.28.0 (9/8/2022)

New Features

  • Support Tversky Loss (#1896)

Bug Fixes

Contributors

MMSegmentation v1.0.0rc0 Released

31 Aug 16:37
87a1d32
Compare
Choose a tag to compare
Pre-release

We are excited to announce the release of MMSegmentation 1.0.0rc0. MMSeg 1.0.0rc0 is the first version of MMSegmentation 1.x, a part of the OpenMMLab 2.0 projects. Built upon the new training engine, MMSeg 1.x unifies the interfaces of dataset, models, evaluation, and visualization with faster training and testing speed.

Highlights

  1. New engines MMSeg 1.x is based on MMEngine, which provides a general and powerful runner that allows more flexible customizations and significantly simplifies the entrypoints of high-level interfaces.

  2. Unified interfaces As a part of the OpenMMLab 2.0 projects, MMSeg 1.x unifies and refactors the interfaces and internal logics of train, testing, datasets, models, evaluation, and visualization. All the OpenMMLab 2.0 projects share the same design in those interfaces and logics to allow the emergence of multi-task/modality algorithms.

  3. Faster speed We optimize the training and inference speed for common models.

  4. New features:

    • Support TverskyLoss function
  5. More documentation and tutorials. We add a bunch of documentation and tutorials to help users get started more smoothly. Read it here.

Breaking Changes

We briefly list the major breaking changes here.
We will update the migration guide to provide complete details and migration instructions.

Training and testing

  • MMSeg 1.x runs on PyTorch>=1.6. We have deprecated the support of PyTorch 1.5 to embrace the mixed precision training and other new features since PyTorch 1.6. Some models can still run on PyTorch 1.5, but the full functionality of MMSeg 1.x is not guaranteed.

  • MMSeg 1.x uses Runner in MMEngine rather than that in MMCV. The new Runner implements and unifies the building logic of dataset, model, evaluation, and visualizer. Therefore, MMSeg 1.x no longer maintains the building logics of those modules in mmseg.train.apis and tools/train.py. Those code have been migrated into MMEngine. Please refer to the migration guide of Runner in MMEngine for more details.

  • The Runner in MMEngine also supports testing and validation. The testing scripts are also simplified, which has similar logic as that in training scripts to build the runner.

  • The execution points of hooks in the new Runner have been enriched to allow more flexible customization. Please refer to the migration guide of Hook in MMEngine for more details.

  • Learning rate and momentum scheduling has been migrated from Hook to Parameter Scheduler in MMEngine. Please refer to the migration guide of Parameter Scheduler in MMEngine for more details.

Configs

Components

  • Dataset
  • Data Transforms
  • Model
  • Evaluation
  • Visualization

Improvements

  • Support mixed precision training of all the models. However, some models may got Nan results due to some numerical issues. We will update the documentation and list their results (accuracy of failure) of mixed precision training.

Bug Fixes

  • Fix several config file errors #1994

New Features

  1. Support data structures and encapsulating seg_logits in data samples, which can be return from models to support more common evaluation metrics.

Ongoing changes

  1. Test-time augmentation: which is supported in MMSeg 0.x is not implemented in this version due to limited time slot. We will support it in the following releases with a new and simplified design.

  2. Inference interfaces: a unified inference interfaces will be supported in the future to ease the use of released models.

  3. Interfaces of useful tools that can be used in notebook: more useful tools that implemented in the tools directory will have their python interfaces so that they can be used through notebook and in downstream libraries.

  4. Documentation: we will add more design docs, tutorials, and migration guidance so that the community can deep dive into our new design, participate the future development, and smoothly migrate downstream libraries to MMSeg 1.x.

v0.27.0

28 Jul 15:11
eeeaff9
Compare
Choose a tag to compare

Changelog

V0.27.0 (7/28/2022)

Enhancement

  • Add Swin-L Transformer models (#1471)
  • Update ERFNet results (#1744)

Bug Fixes

Contributors

MMSegmentation v0.26.0 Release

01 Jul 12:46
17056b6
Compare
Choose a tag to compare

Highlights

  • Update New SegFormer models on ADE20K (1705)
  • Dedicated MMSegWandbHook for MMSegmentation (1603)

New Features

  • Update New SegFormer models on ADE20K (1705)
  • Dedicated MMSegWandbHook for MMSegmentation (1603)
  • Add UPerNet r18 results (1669)

Enhancement

  • Keep dimension of cls_token_weight for easier ONNX deployment (1642)
  • Support infererence with padding (1607)

Bug Fixes

Documentation

  • Fix mdformat version to support python3.6 and remove ruby installation (1672)

New Contributors

Full Changelog: v0.25.0...v0.26.0