Publication
Our paper, titled Learning Joint Policies for Human-Robot Dialog and Co-Navigation, was published in the proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2023.
This research addresses the challenging problem of enabling robots to simultaneously engage in natural dialog with humans while navigating collaboratively toward shared goals.
Research Overview
Traditional robotic systems treat dialog and navigation as separate modules. Our work presents a unified approach that learns joint policies for both dialog management and navigation control, enabling more natural and effective human-robot collaboration.
Key Contributions
- Unified framework for dialog and navigation
- Joint policy learning approach
- Experimental validation in simulated and real-world scenarios
- Improved task completion rates and user satisfaction
Applications
- Service robots in public spaces
- Collaborative delivery systems
- Assistive robotics for navigation support
- Interactive tour guide robots
Authors: Yohei Hayamizu, Zhou Yu, and Shiqi Zhang
Venue: IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) 2023