pyskl icon indicating copy to clipboard operation
pyskl copied to clipboard

A toolbox for skeleton-based action recognition.

Results 96 pyskl issues
Sort by recently updated
recently updated
newest added

For the kinetics-skeleton dataset mentioned in the paper, which version of mmdetection is the faster rcnn used?

Thinks for your great work, and I want to know how soon will the Volleyball skeleton dataset and model be relased? thanks a lot.

Thanks for your significant job! Here I try to train the RGB-only model using the FineGym dataset under mmaction2 frameworks and have some questions. In addition, I utilize ResNet3dSlowOnly backbone...

replied

When the {num_gpus} is 1, 'bash tools/xxx.sh' works. But when I set {num_gpus} as 2 or 3, it shows error: RuntimeError: Address already in us ....... subprocess.CalledProcessError: Command '[...............,'--local_rank=0']' returned...

I noticed the python script(tools/data/custom_2d_skeleton.py) for creating 2D custom data, so how to create 3D dataset? or what script tools should we use?

Hello everyone Currently, I'm experiments on PoseC3D and trying to create some inference services to check results from skeleton keypoints. ```python3 inference_pipeline = [ dict(type='GeneratePoseTarget', with_kp=False, with_limb=True, double=False, left_kp=left_kp, right_kp=right_kp),...

i would like to visualize the training and plots with wandb..i replaced TextloggerHook with wandb but still i can't visualize training,.. log_config = dict(interval=20, hooks=[dict(type='TextLoggerHook')]).. Any help? Thanks in advance

@kennymckormick Hello, thank you very much for your great work. Now I want to use torchinfo.summary to print the network structure of posec3d, but I don't know how much input_size...

Hello, how to use POSEC3D to detect long video behavior?

I have installed ffmpeg and libx264, but still get the error: Unknown encoder 'libx264'