|
Chuhan Wang Hi there! 😄 I'm Chuhan (you can also call me Holly), a Ph.D. student at the University of Notre Dame, fortunate to be advised by Prof. Meng Jiang and affiliated with the DM2 Lab. Before that, I was a master's student at UC San Diego, where I worked closely with Prof. Jingbo Shang. My research focuses on multimodal reasoning in physical AI, exploring how AI systems can perceive, reason about, and interact with the physical world through the integration of vision, language, and action. Feel free to reach out for any form of communication or collaboration! Email / CV / Google Scholar / Github / LinkedIn |
|
News
|
Selected Publications (* indicates equal contribution) |
|
Importance Sampling for Multi-Negative Multimodal Direct Preference Optimization
Xintong Li*, Chuhan Wang*, Junda Wu, Rohan Surana, Tong Yu, Julian McAuley, Jingbo Shang ICLR, 2026 arXiv MISP-DPO incorporates multiple semantically diverse negative images into multimodal DPO via a Plackett-Luce objective, using a sparse autoencoder to select informative negatives and importance sampling for efficient training. |
|
SceneAlign: Aligning Multimodal Reasoning to Scene Graphs in Complex Visual Scenes
Chuhan Wang*, Xintong Li*, Jennifer Yuntong Zhang, Junda Wu, Chengkai Huang, Lina Yao, Julian McAuley, Jingbo Shang ACL, 2026 arXiv SceneAlign uses scene graphs to perform controllable structural interventions, building hard negative rationales for DPO that steer multimodal models toward fine-grained, structure-faithful reasoning. |
|
Customized Training and Application of Monocular Depth Ranging for Assisted Driving
Jingwen Li, Chuhan Wang, Pengwei Zhang, Ruipeng Gao, Dan Tao ICCE-TW, 2023 IEEE A monocular depth ranging method combining a deep learning algorithm with a Jetson Nano-based sensor for assisted driving. |
Education
|
Honors & Awards
|