3 papers
cs.LG2026
PE-OPSD: Internalizing Prompt Enhancement into Flow-matching Models via On-Policy Self-Distillation
Mingfeng Lin, Chengfei Cai, Lin Xu +3
Text-to-image users often provide concise and underspecified prompts, whereas generative models benefit from detailed textual conditions for reliable instruction following. Existin…
cs.LG2026
DreOPD: Degraded-Reference Extrapolative On-Policy Distillation for Flow-matching Models
Mingfeng Lin, Chengfei Cai, Lin Xu +2
Flow-matching models are now a mainstream method to image generation, but its adaptation to diverse downstream scenarios typically relies on post-training, which may cause conflict…
cs.CV2026
FREPix: Frequency-Heterogeneous Flow Matching for Pixel-Space Image Generation
Mingfeng Lin, Jiakun Chen, Liang Han +1
Pixel-space diffusion has re-emerged as a promising alternative to latent-space generation because it avoids the representation bottleneck introduced by VAEs. Yet most existing met…