1 paper · 1 filter
Zhiyuan Gao, Jiageng Mao, Hong-Xing Yu +5
A longstanding goal in computer vision is to model motions from videos, while the representations behind motions, i.e. the invisible physical interactions that cause objects to def…