Introduction to Reinforcement Learning
Introduction to Reinforcement Learning
| Hi-index | 0.00 |
In human interaction with a robot, estimation of the other's intention is thought of as an indispensable factor for achievement of a precise self action. But estimation of the other's intention is heavy loaded information processing, and we don't think humans are always doing it. So, in this paper, we propose a light loaded computational algorithm that achieves human-robot interaction without intention estimation in the self agent. In the method, the self agent assumes the other agent to estimate intention, and searches for an action that is easy to be interpreted by the other agent. We evaluated the effectiveness of the proposed model by computer simulation on a hunter task. This method should be positioned as one of the possible variations of intention-based interaction.