Ziyu Guo

6 papers Department of Computer Science and Engineering, The Chinese University of Hong Kong

• Adaptive Classifier-Free Guidance via Dynamic Low-Confidence Masking
• Delving into RL for Image Generation with CoT: A Study on DPO vs. GRPO
• Perceive Anything: Recognize, Explain, Caption, and Segment Anything in Images and Videos
• T2I-R1: Reinforcing Image Generation with Collaborative Semantic-level and Token-level CoT
• UniCTokens: Boosting Personalized Understanding and Generation via Unified Concept Tokens
• What We Miss Matters: Learning from the Overlooked in Point Cloud Transformers