Implement human-robot interaction via robot's emotion model

Abstract

Nowadays, many robots are in service for the people, such as pepper, Paro, HSR and so on. The robots are not only as partners but also provide services to daily life. To be a good partner, a robot needs to have a better model of feedback, such that it can interact more naturally. Moreover, in order to design a robot that can interact just like human, we need to design the emotion model for the robot. In our research, first our system extracts face and computes the feature from the facial image. Then our system uses SVM to classify the facial expression into different emotion states. At the same time, the robot also updates its own emotion states. Then the robot's emotion will change to interact with human according to the facial expression recognized. Therefore, the robot can show different feedback actions that expressing the emotion of robot with some funny postures.