Multi-Aspect Approach to Robotic Anticipation for Supporting Human's Multiple-Intelligence in Natural Scene

摘要

A multi-aspect approach is applied to one-step anticipation for tightening human-robot interaction in naturally complex scene. Via symbolic massage exchange, human's linguistic intelligence identifies possible solutions by robotic intelligence. In reference to signal space as operable representation of the scene, human's body-kinesthetic intelligence and robotic control systems are co-oriented. As a common basis of these interaction aspects, in this paper, visible class of 'anticipants' is introduced on multi-viewpoint imagery. Despite not-yet-structured representation, the anticipants focalizes robotic intelligence to the scope of human's subsequent performances.