My research lies at the intersection of Video Understanding, World Models, and Agentic Multimodal AI. Concretely, I study how world models can serve as internal simulators for multimodal agents, enabling them to mentally roll out future visual trajectories, infer hidden dynamics, and evaluate potential actions before committing to a response or decision. My research spans video understanding and reasoning, predictive video generation, world model learning, and agentic multimodal systems. My long-term vision is to make imagination a first-class component of multimodal intelligence, enabling agents that are not only more capable, but also more faithful, grounded, and compute-efficient—reliable enough to be trusted in long-horizon real-world decision making under uncertainty.

Here are some of my research works.

• Journals: ACM TIST, IEEE TCSS
• Conferences: ACL Rolling Review, ICML, NeurIPS, AAAI, ICLR

I genuinely believe that meaningful progress in academia stems from open dialogue and thoughtful debate. If you have any questions about my research–or if you’ve previously contacted me through GitHub issues and haven’t received a response–please feel free to reach out via email. I’m always happy to chat, collaborate, or offer assistance where I can.

Throughout my academic journey, I’ve been fortunate to receive support and inspiration from many generous people. I’m always eager to give back to the community and engage with others passionate about learning and discovery.

Preferred Email:

• EnjunDu.cs@gmail.com

Optional Email:

• edu719@connect.hkust-gz.edu.cn, enjundu@tencent.com

Please avoid contacting me at enjun_du@bit.edu.cn, as I will no longer have access to this address.