Sangjae Bae, Ph.D.
Lead Scientist, Honda Research Institute, USA
My research integrates model predictive control with machine learning — including reinforcement learning, inverse RL, and neural networks — to enable safe and cooperative driving in complex traffic scenarios. Recently, I have been exploring vision-language models (VLMs), vision-language-action models (VLAs), and end-to-end learning approaches for mobility systems. Prior to joining HRI, I received the Ph.D. degree in Systems Engineering at UC Berkeley, advised by Prof. Scott Moura, and B.S. in Electronics Engineering at Kyung Hee University, South Korea.
Research
My work sits at the intersection of control, optimization, and machine learning, with a focus on building decision-making systems that are safe, efficient, and capable of reasoning about the behavior of others. I am particularly interested in problems where classical model-based approaches and modern data-driven methods can complement each other.
- Model predictive control integrated with learning (RL, inverse RL, neural networks)
- Vision-language models (VLMs) and vision-language-action models (VLAs) for driving
- End-to-end learning for mobility systems
- Risk-aware and cooperation-aware driving in dense traffic
Selected Publications
H. Kang, E. Sachdeva, P. Gupta, S. Bae, K. Lee, "GFlowVLM: Enhancing Multi-step Reasoning in Vision-Language Models with Generative Flow Networks", CVPR, 2025. [paper] [project]
P. Gupta, D. Isele, E. Sachdeva, P.-H. Huang, B. Dariush, K. Lee, S. Bae, "Generalized Mission Planning for Heterogeneous Multi-Robot Teams via LLM-Constructed Hierarchical Trees", ICRA, 2025. [paper]
H. Hu, D. Isele, S. Bae, J. F. Fisac, "Active Uncertainty Reduction for Safe and Efficient Interaction Planning: A Shielding-Aware Dual Control Approach", IJRR, 2024. [paper] [project] [code]
S. Bae, D. Isele, A. Nakhaei, P. Xu, A. M. Anon, C. Choi, K. Fujimura, S. Moura, "Lane-Change in Dense Traffic with Model Predictive Control and Neural Networks", IEEE TCST, 2022.