Human–Computer Communication Using Recognition and Synthesis of Facial Expression

摘要

To develop a complex computer system such as a robot that can communicate smoothly with humans, it is necessary to equip the system with a function for both understanding human emotions and expressing emotional signals. From both perspectives, facial expression is a promising research area. In our research, we have explored both aspects of facial expression using infrared-ray images and visible-ray images and have developed a personified agent for expressing emotional signals to humans. A human-computer-human communication system using facial expression analysis and a music recommendation system using a personified agent with facial expression synthesis are reported in this paper.