2 papers
cs.CV2026
On-Policy Self-Distillation in Diffusion Models
Wei Zhou, Xiongwei Zhu, Lingdong Kong +14
Reinforcement learning can align diffusion models with human preferences and task-specific objectives, but endpoint rewards do not specify how an intermediate denoising prediction…
cs.CV2026
DanceOPD: On-Policy Generative Field Distillation
Wei Zhou, Xiongwei Zhu, Zelin Xu +8
Modern image generation demands a single model that unifies diverse capabilities, including text-to-image (T2I), local editing, and global editing. However, these capabilities are…