Open ZYX-MLer opened 1 month ago
https://github.com/facebookresearch/InterWild/blob/0c4cd6ca2f5a4220f4630ec9c8a3250c0b612948/data/ReInterHand/ReInterHand.py#L353
What is the difference between 'joint_img' and 'mano_joint_img'? Which one should I use for hand keypoint detection?
thank you.
joint_img is from dataset groundtruth 2.5D keypoints. mano_joint_img is from projecting groundtruth MANO 3D keypoints to the 2.5D space. As in the script, I suggest using both.
joint_img
mano_joint_img
https://github.com/facebookresearch/InterWild/blob/0c4cd6ca2f5a4220f4630ec9c8a3250c0b612948/data/ReInterHand/ReInterHand.py#L353
What is the difference between 'joint_img' and 'mano_joint_img'? Which one should I use for hand keypoint detection?
thank you.