A Dress Coordination Robot System Which Can Improve User’s Ability by a Dialogue Robot

Abstract

People who are not good at coordination of apparel tend to wear similar clothes many times. Unless they notice this fact, they won’t be able to change. Therefore, We developed a dress recommendation system that can cope with this problem by a conversational robot. A big screen displays life-size clothes and the robot explains the recommendation Dress is divided into three parts, an inner, an outer, and bottoms and recommended as a combination of those three. When it is not user’s favorite combination, user can ask a robot the reason for recommendation. We propose two dialogue strategies, blind spots pointing and chance discovery mechanism, to improve the user’s coordination power. The system adapts to the user through the interaction. Furthermore, since the robot explains the user reasons of the recommendation and blind spot pointing, user can easily carry out the chance discovery. Thus, user’s dress coordination power is improved.