1 paper
Michael Li, Alison Ding
Recent advances in generative artificial intelligence have enabled the synthesis of complex human motion with unprecedented fidelity. However, current text-to-motion systems rely s…