A review on robot learning and controlling: imitation learning and human-computer interaction

Abstract

Robots have been increasingly and significantly more powerful and intelligent over the last decade, and moving towards more service oriented roles. Imitation learning and human-robot interaction play important roles in effectively improving robot's intelligence and ability to co-work with human being. This paper reviews the current state of the art in robot learning and controlling based on imitation learning and human-robot interaction. Recent research, classification method and new developments in this area are reviewed. Specifically the definition of action-based imitation, and the control strategy for imitation learning are addressed. Human-robot physical interaction and multi-modal interaction are emphasized.