I am Xinting Hu. Currently, I am a postdoctoral researcher at CVMI Lab, HKU.

My objective is to develop visual intelligence that can perceive, interpret, and simulate the visual world, with emphasis on: 1) visual learning systems that remain efficient, robust, and continually adaptable; 2) generative modeling with strong multi-view visual consistency and interactive controllability; and 3) multimodal intelligence that integrates diverse signals to improve reasoning under real-world settings. Besides, I am interested in leveraging AI to advance scientific and medical research.

Previously, I completed my Ph.D. at MReal Lab, Nanyang Technological University, and conducted postdoctoral research at D2, Max Planck Institute for Informatics. Before that, I got a B.Eng. degree from USTC. My biography is available here.

News

• [Nov. 2025] I will host a Workshop about Next Generation VideGen, NeurIPS 2025. Check this!
• [Oct. 2025] I am selected as a Top Reviewer for NeurIPS 2025.
• [Sep. 2025] I will serve as an Area Chair of ICLR 2025.
• [Sep. 2025] Our benchmark paper for Reasoning-based Image Editing is accepted by NeurIPS 2025 DB track.
• [Aug. 2025] Our paper for Training-Free 3D Asset Editing is accepted by ICCV 2025.
• [Jun. 2025] I am selected as an Outstanding Reviewer for CVPR 2025.
• [Jun. 2025] Our AI4Medicine paper about Zero-shot HLA-peptide Binding Prediction is accepted by Cell Reports.
• [May. 2025] Our survey paper for Personalized Generation In Large Model Era is accepted by ACL 2025.
• [Mar. 2025] Our paper about Training-Free HOI Image Generation is accepted by CVPR 2025.
• [Mar. 2025] Our paper about Visual Prompt for Video Temporal Grounding is accepted by CVPR 2025.
• [Feb. 2025] I served as an Area Chair of ACL ARR 2025 February.