Tianyi Yan
IOTSC, University of Macau. Macau / Beijing..
tianyi.yan123@gmail.com
I am a Ph.D. student at the State Key Laboratory of Internet of Things for Smart City (SKL-IOTSC), University of Macau, advised by Prof. Jianbing Shen and Prof. Chengzhong Xu. Previously, I obtained my Master’s degree from the Institute of Automation, Chinese Academy of Sciences (CASIA), working with Prof. Jinqiao Wang and Prof. Ming Tang, and my B.Eng. degree from BUPT, working with Prof. Chuan Shi. I have also gained industrial experience at Li Auto, Xiaomi EV as a research intern.
My research lies at the intersection of Generative World Models, End-to-End Autonomous Driving, and Embodied AI. Currently, my primary focus is on enabling agents to achieve infinite driving and self-play within physically reliable, imaginative world models.
news
| Sep 19, 2025 | One paper Rlgf: Reinforcement learning with geometric feedback for autonomous driving video generation is accepted by NeurIPS 2025! See you in San Diego! |
|---|---|
| Jul 06, 2025 | One paper is accepted by ACMMM 2025! |
| Jun 26, 2025 | One paper is accepted by ICCV 2025! |
| Feb 27, 2025 | One paper Drivingsphere: Building a high-fidelity 4d world for closed-loop simulation is accepted by CVPR 2025! It is about Generative Closed-loop Simulation. See you in Nashvile! |
| Dec 10, 2024 | One paper OLiDM: Object-aware LiDAR Diffusion Models for Autonomous Driving is accepted by AAAI 2025! It is about LiDAR Generation. |
latest posts
| Mar 26, 2025 | a post with plotly.js |
|---|---|
| Dec 04, 2024 | a post with image galleries |
| May 14, 2024 | Google Gemini updates: Flash 1.5, Gemma 2 and Project Astra |
selected publications
- AD-R1: Closed-Loop Reinforcement Learning for End-to-End Autonomous Driving with Impartial World ModelsarXiv preprint arXiv:2511.20325, 2025
- 3d and 4d world modeling: A surveyarXiv preprint arXiv:2509.07996, 2025
- RLGF: Reinforcement Learning with Geometric Feedback for Autonomous Driving Video GenerationarXiv preprint arXiv:2509.16500, 2025
- Drivingsphere: Building a high-fidelity 4d world for closed-loop simulationIn Proceedings of the Computer Vision and Pattern Recognition Conference, 2025
- OLiDM: Object-aware LiDAR Diffusion Models for Autonomous DrivingIn Proceedings of the AAAI Conference on Artificial Intelligence, 2025