We develop machine learning methods for urban transportation and mobility, with a focus on building models that are generative, probabilistic, and scalable. Our work spans deep generative modeling, spatiotemporal forecasting, reinforcement learning for connected vehicles, and large language/vision models for transportation safety.
 DGMinTransportation Tutorial & survey on deep generative models in transportation TR Part C, 2025 |
 BO4Mob Bayesian optimization benchmarks for high-dimensional urban mobility NeurIPS, 2025 |
 TrajFlow Flow-based trajectory generation |
 HighwayVLM Vision-language models for highway safety analysis |
 TrafficNetQA LLM-powered traffic network question answering |
 camera2detector Real-time traffic camera validation with YOLO computer vision |
Full Publication List
See our complete and up-to-date publication list on Google Scholar or the lab website.
| Principal Investigator | |||
|---|---|---|---|
| Dr. Seongjin Choi — Assistant Professor, CEGE, University of Minnesota Twin Cities | |||
| Graduate Students | |||
| Lindong Liu Ph.D. · Causal generative models, Inverse problems |
Christopher Cheong Ph.D. · Traffic forecasting, Large-scale spatiotemporal models |
Sicong Jiang Ph.D. (McGill) · Autonomous driving, Distributional RL |
Pouya Parsa M.S. · Computer vision, 3D reconstruction, LLM/VLM |
| Mithell Kosieradzki M.S. Data Science · Probabilistic trajectory prediction |
Rammesh Adhav Saravanan M.S. Robotics · VLM for autonomous vehicles |
Seunghee Ryu Visiting (Korea Univ.) · Bayesian OD estimation |
Donghoon Kwon Visiting (Korea Univ.) · LLM for traffic simulation |
We also build AI Teaching Assistants (AITA) — LLM-powered course chatbots deployed across multiple UMN engineering courses.
| Repo | Description |
|---|---|
| aita-core | Shared core package for AITA chatbots |
| aita-template | Template for creating new course chatbots |
| SynPopPred | Tutorial: LLM-based synthetic population generation (IPF, CTGAN, DistilGPT-2, Llama 3.1) |
Interested in joining or collaborating? Reach out at chois@umn.edu or visit our lab website.