We conduct cutting-edge research in multimodal learning, visual perception, video understanding, and language semantics, pushing AI systems toward deeper and more reliable semantic intelligence.

Current core directions:

• Multimodal large language models and agentic reasoning.
• Text-to-image and text-to-video generation and editing.
• Efficient and trustworthy large language models.

We welcome self-motivated research interns and PhD applicants with strong mathematical and engineering backgrounds, especially those interested in multimodal AI, LLMs, and generative modeling.

Explore our recent work on the Publications page.