CLIP-Motion: Learning Reward Functions for Robotic Actions Using Consecutive Observations

Open in new window