Hear to See: Discerning Stateful Listening for Audio-Visual Instance Segmentation

作者：Leiye Liu, Miao Zhang, Jiahong Jiang, Jingjing Li, Jialong Zhong, Kai Peng, Tingwei Liu, Wei Ji, Yongri Piao, Huchuan Lu · 单位：Dalian University of Technology, Carnegie Mellon University, Yale University, Dalian University of Technology Dalian Liaoning China, Carnegie Mellon University Pittsburgh Pennsylvania USA, Yale University New Haven Connecticut USA · 会议/期刊：arXiv preprint · 方向：cs.MM · 发布日期：2026-08-05