-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Description
Will look at either OpenPose or MediaPipe for this. I'll see what I decide as I go.
Regardless, this will be the goal:
Pose Estimation
Apply Pose Estimation Models: Use state-of-the-art pose estimation frameworks (like OpenPose, YOLOv8-Pose, ViTPose, or TensorFlow’s MoveNet) to detect and track body keypoints for each grappler in every frame. This converts video into structured pose data.
Metadata
Metadata
Assignees
Labels
No labels