RNN-for-Human-Activity-Recognition-using-2D-Pose-Input icon indicating copy to clipboard operation
RNN-for-Human-Activity-Recognition-using-2D-Pose-Input copied to clipboard

Published 20 hours ago • verified publisher icon stuarteiffert

json_to_text.py [update]

Open VicenteQueiroz opened this issue 7 years ago • 0 comments
trafficstars

Hey just wanted to tell you that openPOSE updated their json output format (https://github.com/CMU-Perceptual-Computing-Lab/openpose/blob/master/doc/output.md) because they added a new 3-D keypoint reconstruction module.

So now, in Line 64: pose_keypoints = data["people"][0]["pose_keypoints"] Should be: pose_keypoints = data["people"][0]["pose_keypoints_2d"]

VicenteQueiroz avatar Mar 23 '18 10:03 VicenteQueiroz