1 paper · 1 filter
Siyi Chen, Shaowei Liu, Yixuan Jia +4
Recent progress has shown promise in distilling multi-step video diffusion models into efficient few-step students. Among them, Distribution Matching Distillation (DMD) and its suc…