1 paper · 2 filters
Bin Cao, Sipeng Zheng, Hao Luo +3
Text-to-motion (T2M) generation aims to create realistic human movements from text descriptions, with promising applications in animation and robotics. Despite recent progress, cur…