stefanopini/simple-HRNet

Turn off some keypoints

shinbehavior opened this issue · 0 comments

Hi, this is excellent work!

  1. Can we disable some key point detections? For example, I don't want to detect legs, only hands and head.

  2. And, can I train the model on the voxel-pose dataset? Like this https://campar.in.tum.de/Chair/MultiHumanPose.

  3. If I want to label my videos in CVAT, which tool do I need to use to label for retraining, for example, pose_hrnet_w32_256x192.pth ?