Anticipatory visual perception as a bio-inspired mechanism underlying robot locomotion

Abstract

Anticipation of sensory consequences of actions is critical for the predictive control of movement that explains most of our sensory-motor behaviors. Plenty of neuroscientific studies in humans suggest evidence of anticipatory mechanisms based on internal models. Several robotic implementations of predictive behaviors have been inspired on those biological mechanisms in order to achieve adaptive agents. This paper provides an overview of such neuroscientific and robotic evidences; a high-level architecture of sensory-motor coordination based on anticipatory visual perception and internal models is then introduced; and finally, the paper concludes by discussing the relevance of the proposed architecture within the context of current research in humanoid robotics.