Empathic Interaction Using the Computational Emotion Model

摘要

This paper describes the empathy oriented human-robot interaction model. It is projected to design the model capable of different empathic responses (parallel and reactive) during the course of interaction with the user, depending upon the personality and mood factors of the robot. The proposed model encompasses three main stages i.e., Perception, empathic appraisal and empathic expression. Perception refers to capturing user's emotion state via facial expression recognition. Empathic appraisal is based on the computational emotional model for generating its internal emotions, mood state and empathic responses. The internal emotions are defined using psychological studies and generated on 2D (pleasure-arousal) scaling model, whereas, fuzzy logic is used to calculate the intensity of the each emotion. A virtual facial expression simulator is applied for expression of resultant empathic emotions. Preliminary experimental results show that the proposed model is capable of exhibiting different empathic responses with respect to the personality and mood factors.