search

open-mmlab / mmaction2

OpenMMLab's Next Generation Video Understanding Toolbox and Benchmark
https://mmaction2.readthedocs.io
Apache License 2.0
4.16k stars 1.22k forks source link

Is there a config for training PoseC3D in AVA dataset #1663

Open MMz000 opened 2 years ago

MMz000 commented 2 years ago

I want to run poseC3D model in spatio temporal action detection and I notice that there is a checkpoint in demo/demo_video_structuralize.py

parser.add_argument(
    '--skeleton-stdet-checkpoint',
    default=('https://download.openmmlab.com/mmaction/skeleton/posec3d/'
             'posec3d_ava.pth'),
    help='skeleton-based spatio temporal detection checkpoint file/url')

But I can't find the training config in config folder.

kennymckormick commented 2 years ago

That is not available yet.

MMz000 commented 2 years ago

Is that means it will be available in the futrue?

kennymckormick commented 2 years ago

Maybe late this year, but there's no guarantee

hotelbread commented 1 year ago

I also waiting for this!!!