2 papers
cs.CV2026
Learning from Noisy Preferences: A Semi-Supervised Learning Approach to Direct Preference Optimization
Xinxin Liu, Ming Li, Zonglin Lyu +2
Human visual preferences are inherently multi-dimensional, encompassing aesthetics, detail fidelity, and semantic alignment. However, existing datasets provide only single, holisti…
cs.CV2025
CPO: Condition Preference Optimization for Controllable Image Generation
Zonglin Lyu, Ming Li, Xinxin Liu +1
To enhance controllability in text-to-image generation, ControlNet introduces image-based control signals, while ControlNet++ improves pixel-level cycle consistency between generat…