A Laban-Based Approach to Emotional Motion Rendering for Human-Robot Interaction

Abstract

A motion-rendering system that adds target emotion to basic movements of human form robot (HFR) by modifying the movements was created. Pleasure, anger, sadness or relaxation is considered as target emotion. This method not only keeps the user interested, but it also makes the user perceive the robot’s emotions and form an attachment to the robot more easily. An experiment was conducted using a real HFR to test how well our system adds target emotion to basic movements. The average of the success rates for adding the target emotion to basic motions were over 60%. This suggests that our method succeeded in adding the target emotions to arbitrary movements.