Path Planning by Spatial Concept-Based Probabilistic Inference from Human Speech Instructions

Abstract

The aim of this study is to enable a mobile robot to perform navigation tasks by probabilistic inference using spatial concepts on a probabilistic model. Specifically, path planning is performed to the target state of a spatial concept estimated through human speech instructions such as ``Go to the kitchen''. In the experiment, places instructed by the speech command of the user showed high probability values, and the trajectory toward the target place was correctly estimated.