Detecting morphology of task structures using ART model for teleoperating mobile robot via virtual reality space

摘要

This paper presents a method for detecting the invariant task structures that are commonly shared by the two different actors having different perception-acting capabilities. We call such a property of having common structures as morphology of task structures, and we present how this property functions well for the design of a mobile robot teleoperation system via a VR-based interface. For this purpose, we introduce a neural network learning model of Carpenter and Grossberg's Adaptive Resonance Theory (ART).