I am developing GenAI models at HKUST. My research interests cover content generation and cross-modality representation learning, especially for image, video and audio synthesis.
Publications
Foundation Cures Personalization: Recovering Facial Personalized Models' Prompt Consistency
Arxiv 2024 / Paper
MDT-A2G: Exploring Masked Diffusion Transformers for Co-Speech Gesture Generation
ACM MM 2024 / Paper
Learning Unified Reference Representation for Unsupervised Multi-class Anomaly Detection
ECCV 2024 / Paper
Honors and Awards