Towards Hands-Free Human-Robot Interaction through Spoken Dialog

摘要

We present an approach to hands-free interaction with autonomous robots through spoken dialog. Our approach is based on passive knowledge rarefication through goal disambiguation, a technique that allows the robot to refine and acquire knowledge through spoken dialog with a human operator. A key assumption underlying our approach is that the operator and the robot share the same set of goals. Another key idea is that language and vision have some memory structures in common. We discuss how our approach achieves four types of human-robot interaction: command, elaboration, introspection, and instruction-based learning. We discuss our experiences with implementing our approach on an autonomous robot.