Featured Research

The ultimate goal of our research is to build trustworthy, interactive, and human-centered autonomous agents that can perceive, understand, and reason about the physical world; safely interact and collaborate with humans and other agents, and clearly explain their behaviors to build trust with humans so that they can benefit society in daily lives. To achieve this goal, we have been pursuing interdisciplinary research and unifying the techniques and tools from robotics, machine learning, reinforcement learning, explainable AI, control theory, optimization, and computer vision.

Explainable Relational Reasoning and Multi-Agent Interaction Modeling (Social & Physical)

We investigate relational reasoning and interaction modeling in the context of the trajectory prediction task, which aims to generate accurate, diverse future trajectory hypotheses or state sequences based on historical observations. Our research introduced the first unified relational reasoning toolbox that systematically infers the underlying relations/interactions between entities at different scales (e.g., pairwise, group-wise) and different abstraction levels (e.g., multiplex) by learning dynamic latent interaction graphs and hypergraphs from observable states (e.g., positions) in an unsupervised manner. The learned latent graphs are explainable and generalizable, significantly improving the performance of downstream tasks, including prediction, sequential decision making, and control. We also proposed a physics-guided relational learning approach for physical dynamics modeling.
Related Publications:
6. Multi-Agent Dynamic Relational Reasoning for Social Robot Navigation, submitted to IEEE Transactions on Robotics (T-RO), under review.

Interaction-Aware Decision Making and Model-Based Control

How Will Self-Driving Cars Be Insured in the Future?