| 摘要 |
PROBLEM TO BE SOLVED: To obtain output more than that required by a user by correcting/ learning an emotion generation algorithm and/or an operation deciding algorithm with the reaction of the user based on the operating result of an object to be controlled on which artificial emotion is reflected as evaluation reference. SOLUTION: A behavior deciding means 40 judges how the artificial emotion of a robot 1 is recognized from the reaction of the user, which is caused by the behavior of the robot itself, based on information obtained from an intention/emotion recognition part 31. The operation deciding algorithm in the behavior deciding means 40 is evaluated based on a difference between the artificial emotion of the robot, which the user recognizes, and the actual artificial emotion of the robot 1. Thus, the robot 1 corrects and learns the intention/recognition algorithm, the emotion generation algorithm and the operation deciding algorithm based on the reaction of the user owing to the motion of the robot itself. Then, self artificial emotion and emotion expression are established while being communicated with the user. |
