A cognitive robotic model of grasping.

Abstract

In this paper we present a cognitive robotic model of object manipulation (i.e. grasping) based on psychologically plausible embodied cognition principles. Specifically, the robotic simulation model is inspired by recent theories of embodied cognition, in which vision, action and semantic systems are linked together in a dynamic and mutually interactive manner. The robotic agent is based on a simulation model of the iCub humanoid robot. It uses a connectionist control system trained with experimental data on object manipulation. Simulation analyses show that the robot is capable to reproduce phenomena observed in human experiments, such as the Stimulus-Response Compatibility effect. 1.