Dingbang Huang

Dingbang Huang 「黄定邦」

I'm a senior undergraduate student from Shanghai Jiao Tong University, affiliated with School of Physics and School of Computer Science, ranking 2nd in my major.

I am fortunate to be advised by Prof. Georgios Pavlakos and Prof. Qixing Huang at UT Austin, Prof. Bo Dai at Shanghai AI Lab, Prof. Cewu Lu, Prof. Yonglu Li and Prof. Hesheng Wang at Shanghai Jiao Tong University. I also work closely with Prof. Etienne Vouga and Prof. Volkan Isler at UT Austin.

I am actively seeking a Ph.D. position beginning in Fall 2026 !!

Email / CV / Scholar

Research Interest
In general, my research interests focus on the intersection of 3D Computer Vision and Robotics. More specifically, I am interested in how human interact with the dynamic physical world, including human-object/scene interaction (HOI/HSI), 4D reconstruction, human motion, physics-based character animation, humanoid whole-body manipulation and visual generative modelig.

Goal
My long-term goal is to build intelligent systems that can percieve and understand the dynamic physical world and enable embodied agents like digital humans and humanoid robots to autonomously interact with their surroundings as humans do.

News

Feb, 2026 Our paper on physical HOI from videos is accepted by CVPR 2026! See you in Denver!
Sep, 2025 PSDiffusion accepted by WACV 2026 (Round 1 direct accept, top 6.5%).
Jun, 2025 I'll be joining UT Austin as a research intern.
Jun, 2025 Presenting our Open3DHOI at CVPR 2025.
Feb, 2025 Open3DHOI accepted by CVPR 2025.

Publications

	Recovering Physically Plausible Human-Object Interactions from Monocular Videos Dingbang Huang, Etienne Vouga, Qixing Huang, Georgios Pavlakos CVPR, 2026 (to be released in March) Recovering physically plausible human-object interactions in simulation from monocular videos using reinforcement learning.
	PSDiffusion: Harmonized Multi-Layer Image Generation via Layout and Appearance Alignment Dingbang Huang, Wenbo Li, Yifei Zhao, Xinyu Pan, Chun Wang, Yanhong Zeng, Bo Dai WACV, 2026 (Round 1 direct accept, top 6.5%) project page / paper Simultaneous harmonized multi-layer image generation using diffusion models via layout and appearance alignment.
	Reconstructing In-the-Wild Open-Vocabulary Human-Object Interactions Boran Wen, Dingbang Huang, Zichen Zhang, Jiahong Zhou, Jianbin Deng, Jingyu Gong, Yulong Chen, Lizhuang Ma, Yong-Lu Li CVPR, 2025 project page / paper Reconstructing open-vocabulary human-object interactions from in-the-wild images.
	Efficient and Scalable Monocular Human-Object Interaction Motion Reconstruction Boran Wen, Ye Lu, Keyan Wan, Sirui Wang, Jiahong Zhou, Junxuan Liang, Xinpeng Liu, Bang Xiao, Dingbang Huang, Ruiyang Liu, Yong-Lu Li Arxiv, 2025 project page / paper Efficient 4D human-Object interaction reconstruction from in-the-wild videos.
	Memorize My Movement: Efficient Sensorimotor Navigation with Self-motion-based Spatial Cognition Qiming Liu, Dingbang Huang, Zhe Liu, Hesheng Wang IEEE Transactions on Automation Science and Engineering (T-ASE) paper Efficient memory mechanism for visual navigation using self-motion-based spatial cognition.

Website template from Jon Barron.