Affordance learning for robotic grasping

Abstract

This paper describes a prototype robot grasping system that uses human grasping synergies and a self-organizing map to learn object affordances. The bio-inspired design of the system is presented as well as some of the results from affordance learning.