Artificial emotion model for human-machine communication by using harmony theory

This paper deals with an artificial emotion model for `Active Human Interface´ that generates emotions and facial expressions from desire state by neural network application. The harmony theory, a type of Boltzmann machine, is employed in this paper, and for this network system, the authors show a method of learning six basic emotions (joy, anger, sadness, fear, disgust and surprise). They also formulate schemata connecting a desire state and facial expression consisting of three facial components (eye, eyebrow and mouth). Some simulation results show the successful emotion generation, demonstrating the effectiveness of the learning