作者位次 #2/12SIGIR 2026Oral· Information Retrieval (Full)
ProEchoMem: Enhancing Long Video Understanding via Multi-Trace Probe-Echo Memory
Derong Xu · Yanxin Chen · an Jiaotong University) · an Jiaotong University) · Chao Zhang · Maolin WANG · Yiqi Wang · Jipeng Qiang · Xuetao Wei · Hongzhi Yin · an Jiaotong University) · an Jiaotong University)
摘要
Large vision-language models (LVLMs) have shown significant progress in video understanding, but they struggle to scale to long videos due to limited context windows. Existing methods reduce input dimensionality via frame sampling and feature compression, yet discard details and incur high comput...