Jinhui Tang

5 papers Nanjing University of Science and Technology

• Aligning Text-to-Image Diffusion Models to Human Preference by Classification
• DISCO: DISCrete nOise for Conditional Control in Text-to-Image Diffusion Models
• OmniGaze: Reward-inspired Generalizable Gaze Estimation in the Wild
• Plenodium: Underwater 3D Scene Reconstruction with Plenoptic Medium Representation
• Vision-centric Token Compression in Large Language Model