Multi-person Conversation Robot using Multi-modal Interface

摘要

This paper describes a robot who converses with multiple users at the same time, using his limited resource. Conversation with multiple users includes many new problems, which are not cared in conventional one-to-one conversation: such as perception of message exchange (recognize who is speaking and to whom he is speaking), expression problem and planning problem. We solved these problems by utilizing multi-modal interface: face recognition, face direction recognition, sound source estimation, speech recognition and gestural output. The systematic combination of these functions realized human-friendly multi-person conversation system.