1 paper
Xin Li, Keren Fu, Qijun Zhao
Existing video camouflaged object detection (VCOD) methods primarily rely on spatial appearances for motion perception. However, the high foreground-background similarity in VCOD l…