1 paper
Hongbo Kang, Tianyi Zhou, Qingyang Yang +4
Recovering scene-consistent 4D crowd motion from monocular video in large-scale scenes remains challenging due to severe depth ambiguity and complex scene geometry. Existing monocu…