Abstract: In this work, we present a method for RGB-based action recognition using multi-view videos. We present a supervised contrastive learning framework to learn a feature embedding robust to ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback