A connectionist model for learning robotic grasps using reinforcement learning

摘要

We present a connectionist model for learning robotic grasps. These grasps are represented as generic grasping functions. Our approach is to learn these functions by experimentation with the environment. Grasping rules are mapped to generic representations that can then be learned by experiments using neural networks. Furthermore, grasping rules acquired in this format can then be used on different objects using different grippers. During experimentation, reinforcement learning is used to minimize the number of failed experiments. Results show that the system is able to learn how to grasp various objects while maintaining a small number of experiments.