1 paper
Jia Li, Wenjie Zhao, Ziru Huang +2
Unlike traditional visual segmentation, audio-visual segmentation (AVS) requires the model not only to identify and segment objects but also to determine whether they are sound sou…