You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
[](https://github.com/open-mmlab/mmaction2/issues)
26
+
[](https://github.com/open-mmlab/mmaction2/issues)
[](https://github.com/open-mmlab/mmaction2/issues)
37
-
[](https://github.com/open-mmlab/mmaction2/issues)
38
-
39
39
MMAction2 is an open-source toolbox for video understanding based on PyTorch.
40
40
It is a part of the [OpenMMLab](http://openmmlab.org/) project.
41
41
42
-
The master branch works with **PyTorch 1.6+**.
42
+
The 1.x branch works with **PyTorch 1.6+**.
43
43
44
44
<divalign="center">
45
45
<divstyle="float:left;margin-right:10px;">
@@ -48,7 +48,7 @@ The master branch works with **PyTorch 1.6+**.
<pstyle="font-size:1.5vw;">Skeleton-base Action Recognition Results on NTU-RGB+D-120</p>
51
+
<pstyle="font-size:1.5vw;">Skeleton-based Action Recognition Results on NTU-RGB+D-120</p>
52
52
</div>
53
53
</div>
54
54
<divalign="center">
@@ -70,10 +70,9 @@ The master branch works with **PyTorch 1.6+**.
70
70
71
71
## What's New
72
72
73
-
**v1.0.0rc0**was released in 31/8/2022:
73
+
- (2022-10-11) We support **Video Swin Transformer**on Kinetics400 and additionally train a Swin-L model on Kinetics700 to extract video features for downstream tasks.
74
74
75
-
- Unifies interfaces of all components based on [MMEngine](https://github.com/open-mmlab/mmengine).
76
-
- Faster training and testing speed with complete support of mixed precision training.
75
+
**Release**: v1.0.0rc1 was released in 14/10/2022. Please refer to [changelog.md](docs/en/notes/changelog.md) for details and release history.
77
76
78
77
## Installation
79
78
@@ -109,8 +108,7 @@ Please refer to [install.md](https://mmaction2.readthedocs.io/en/1.x/get_started
0 commit comments