In the context of computer vision, most of the traditional action recognition techniques assign
a single label to a video after analyzing the whole video. We believe that understanding of …