MASR: Self-Reflective Reasoning through Multimodal Hierarchical Attention Focusing for Agent-based Video Understanding
机构 * Li Auto Inc.(利亚汽车公司) ; University of Chinese Academy of Sciences(中国科学院大学)
大厂专区
机构 * Li Auto Inc.(利亚汽车公司) ; University of Chinese Academy of Sciences(中国科学院大学)
机构 * Li Auto(利亚 Auto)
Comments 37 pages, 9 figures