Learning to recognize affective body postures

Robots are assuming an increasingly important role in our society. They now become pets and help support children healing. In other words, they are now trying to entertain an active and affective communication with human agents. However, up to now, such systems have primarily relied on the human agents' ability to empathize with the system. Changes in the behavior of the system could therefore result in changes of mood or behavior in the human partner. But current systems do not seem to react to users, or only in clearly pre-defined ways. In that sense, current systems miss the bi-directionality typical to human social interaction. Social interaction is characterized by a multi-channel communication, in which each actor captures and reacts to signals by the other actor. To this aim, a computer or a robot has to be able to capture and interpret signals sent by the human partner in order to achieve social interaction. One of the most important channels of communication is physical interaction. The body is used to interpret the affective state of an interlocutor. This paper describes experiments we carried out to study the importance of body language in affective communication. The results of the experiments led us to develop a system that can incrementally learn to recognize affective states from body postures.