How are the gt pose_2d files saved by 'prepare_data_humaneva.py' linked to the frame numbers in each '*.avi' video? Overlaying pose_2d on the frames (tried with +offset, -offset, 0 offset) are still not yielding good overlays. Some of the joints are displaced.
How are the gt pose_2d files saved by 'prepare_data_humaneva.py' linked to the frame numbers in each '*.avi' video? Overlaying pose_2d on the frames (tried with +offset, -offset, 0 offset) are still not yielding good overlays. Some of the joints are displaced.