Published multiple papers, including 'Glance2Gaze: Efficient Vision-Language Models from Glance Fusion to Gaze Compression' (NeurIPS 2025), 'Process-Supervised Reinforcement Learning for Code Generation' (EMNLP 2025), 'Pi-GPS: Enhancing Geometry Problem Solving by Unleashing the Power of Diagrammatic Information' (ICCV 2025), etc.
Research Experience
Interned at MSRA from 2013 to 2017 and received the Microsoft Research Asia Fellowship award in 2015. Joined MSRA in 2017 and then joined Beijing Normal University in 2024.
Education
B.S. (2012) and Ph.D. (2017) from University of Science and Technology of China (USTC).
Background
Currently an Associate Professor at the School of Artificial Intelligence, Beijing Normal University. Research interests include computer vision, such as content generation, representation learning, and large-scale indexing.