Xingtai Gui

I am a second-year Ph.D. student at the State Key Laboratory of Internet of Things for Smart City, University of Macau, supervised by Prof. Jianbing Shen.
I received both my B.Eng. (2016-2020) and M.Eng. (2020-2023) degrees from the University of Electronic Science and Technology of China (UESTC). Before starting my Ph.D., I worked as a Researcher at MEGVII from 2023 to 2025.

Research Interests

• End-to-end autonomous driving
• World model for physical agent

Recent Publications

First author publications

1. Bridging Scene Generation and Planning: Driving with World Model via Unifying Vision and Motion Representation (Arxiv)
   Authors: Xingtai Gui, Meijie Zhang, Tianyi Yan, et al.
   Links: [Paper] [GitHub]

2. TrajDiff: End-to-end Autonomous Driving without Perception Annotation (Arxiv)
   Authors: Xingtai Gui, Jianbo Zhao, Wencheng Han, et al.
   Links: [Paper] [GitHub]

3. FipTR: A Simple yet Effective Transformer Framework for Future Instance Prediction in Autonomous Driving (2024, ECCV)
   Authors: Xingtai Gui, Tengteng Huang, Haonan Shao, et al.
   Links: [Paper] [GitHub]

Other publications

1. SurfelOcc: Self-supervised Occupancy Prediction via 2D Surfel Splatting (2026, CVPR Findings)
   Authors: Jikai Wang, Xingtai Gui, et al.
   Link: [Paper]
2. AD-R1: Closed-Loop Reinforcement Learning for End-to-End Autonomous Driving with Impartial World Models (2026, CVPR Findings)
   Authors: Tianyi Yan, Tao Tang, Xingtai Gui, et al.
   Link: [Paper]

Experiences 💼

• [2026 – Present] Research Intern, Zhuoyu Technology (领航者计划)
• [2025 – 2026] Research Intern, AFARI (Autonomous Driving)