Yunhong Lu

Reward-Forcing. [CVPR 2026 Highlight] Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation

InPO. [CVPR 2025 Highlight] InPO: Inversion Preference Optimization with Reparametrized DDIM for Efficient Diffusion Model Alignment

SmPO. [ICML 2025] Smoothed Preference Optimization via ReNoise Inversion for Aligning Diffusion Models with Varied Human Preferences

SmPO-project-page. SmPO-Diffusion (ICML 2025) Project Page

InPO-project-page. DDIM-InPO (CVPR 2025 Highlight) Project Page