1 paper · 1 filter
Qifan Wang, Qihang Fang, Boao Shi +4
Modeling human-object interaction (HOI) in real environments requires jointly representing whole-body motion, manipulated objects, and scene geometry, yet existing datasets trade o…