Robust 3D Mapping and Adaptive Navigation
We study robust 3D mapping, geometry-aware learning, and adaptive navigation in dynamic settings.
Video-Language Grounding for Open-World Agents
We study video-language grounding and open-world recognition for long-term autonomous agents.
Foundation Models for Robotic Manipulation
We study foundation models for robotic manipulation, task abstraction, and reusable control primitives.
Uncertainty Estimation and Robust Deployment
We study uncertainty estimation, risk-sensitive inference, and robust deployment under distribution shift.
Interactive 3D World Models for Embodied Training
We study interactive 3D world models for simulation, prediction, and scalable embodied training pipelines.
