search

wmcnally / kapao

KAPAO is an efficient single-stage human pose estimation model that detects keypoints and poses as objects and fuses the detections to predict human poses.
GNU General Public License v3.0
753 stars 103 forks source link

Using frame-to-frame pose difference tracking for video demo #52

Closed austinulfers closed 2 years ago

austinulfers commented 2 years ago

How can I implement the frame-to-frame pose difference tracking for the video demo where we want to track an unknown amount of people? The squash demo has this implemented for finding the two active players, however, how could this be incorporated so that we can track all players in the video (i.e. the flash mob video)?

austinulfers commented 2 years ago

Closing as I've recently created this