About Me

I am an Assistant Professor at AI Thrust, Information Hub of Hong Kong University of Science and Technology (Guangzhou). I am also an Affiliated Assistant Professor of the department of Computer Science & Engineering (Clear Water Bay Campus). I am a faculty member of the Deep Vision Lab. Previously, I was a Postdoctoral Associate at Computer Science & Artificial Intelligence Lab of Massachusetts Institute of Technology, where I had the privilege of working with Prof. Dina Katabi. I earned my Ph.D. degree from the Chinese University of Hong Kong, under the mentorship of Prof. Jiaya Jia. I am honored as Distinguished Young Scholars (Overseas). For more information about my research group, please visit EnVision-Research.

Interests
  • Computer Vision
  • Generative Models
  • AI+X
📚 My Research
My research focuses on visual generative models, exploring their fundamental principles with the aim of improving their quality, efficiency, diversity, and controllability. Beyond foundational research, I am dedicated to applying these models to solve real-world challenges in sectors such as autonomous driving, smart manufacturing, and content creation. My overarching goal is to advance the field of generative models by tackling sophisticated real-world challenges, thereby pushing the boundaries of academic research in different disciplines. Please find the collection of our open-source code at .
Recent News
Recent Publications
DisEnvisioner: Disentangled and Enriched Visual Prompt for Customized Image Generation. International Conference on Learning Representations, 2025.
Kiss3DGen: Repurposing Image Diffusion Models for 3D Asset Generation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2025.
Lotus: Diffusion-based visual foundation model for high-quality dense prediction. International Conference on Learning Representations, 2025.
POSTA: A Go-to Framework for Customized Artistic Poster Generation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2025.
RectifiedHR: Enable Efficient High-Resolution Image Generation via Energy Rectification. arXiv preprint arXiv:2503.02537, 2025.
Playground