CVPR2024
Benchmarking Audio Visual Segmentation for Long-Untrimmed Videos
Chen Liu, Peike Patrick Li, Qingtao Yu, Hongwei Sheng, Dadong Wang, Lincheng Li, Xin Yu
Abstract
Segment Results Stage1：Mask Generation for the First Frame Remaining Frames XMem Stage2: Tracking prompt Stage3: Bad Results Correction Prototype to Points # Frame t SAM … #Frame m #Frame n #Frame 2 #Frame T … … Stage 1&2