OpenMMLab Pose Estimation Toolbox and Benchmark.
Highlights
New Features
Improvements
Bug Fixes
Highlights
New Features
Improvements
Bug Fixes
min_keypoints
in tracking apis (#1398) @pallgeuertorch.meshgrid
(#1402) @pallgeuermmpose.datasets.backbones.utils
(#1405) @ly015Highlights
New Features
Improvements
TopDownPoseTrack18VideoDataset
(#1335) @liqikai9Bug Fixes
ignore_to_much
is True (#1341) @daixinghomebody3d_two_stage_video_demo.py
(#1295) @mucozcanBreaking Changes
v0.26.0 | v0.25.0 | |
---|---|---|
Dataset (e.g. TopDownCOCODataset) |
|
|
Pipeline Config (e.g. HRNet+COCO) |
|
|
Advantage |
|
- |
BC Breaking | The method _xywh2cs of dataset base classes (e.g. Kpt2dSviewRgbImgTopDownDataset) will be deprecated in the future. Custom datasets will need modifications to move the bbox format conversion to pipelines. |
- |
This release is meant to fix the compatibility with the latest mmcv v1.5.0
Highlights
Smoother
module for temporal smoothing of the pose estimation with configurable filters (#1127) @ailingzengzzz, @ly015New Features
Smoother
module for temporal smoothing of the pose estimation with configurable filters (#1127) @ailingzengzzz, @ly015Improvements
Bug Fixes
channel_shuffle
that may cause exporting failure (#1242) @PINTO0309custom_hook
can not be set in configs (#1236) @bladromeHighlights
New Features
Improvements
Bug Fixes
Highlights
New Features
Improvements
Add multi-processing configurations to speed up distributed training and testing (#1146) @ly015
Add default runtime config (#1145)
Upgrade isort in pre-commit hook (#1179) @liqikai9
Update README and documents (#1171, #1167, #1153, #1149, #1148, #1147, #1140) @jin-s13, @wusize, @TommyZihao, @ly015
Bug Fixes
Highlights
New Features
Improvements
Bug Fixes
sigmas
in bottom-up image demo (#1107, #1101) @chenxinfeng4, @liqikai9args.show==False
in demo scripts (#1104) @bladromeHighlights
New Features
Improvements
Bug Fixes
Highlights
New Features
Bug Fixes
Improvements