1 paper
Ziyu Ma, Chenhui Gou, Hengcan Shi +4
Most of the existing methods for video understanding primarily focus on videos only lasting tens of seconds, with limited exploration of techniques for handling long videos. The in…