Social Signal Processing: Understanding Nonverbal Communication in Social Interactions

This paper provides a short overview of Social Signal Processing, the domain aimed at bridging the social intelligence gap between people and machines. The focus of Social Signal Processing is on nonverbal behavioral cues that human sciences (psychology, anthropology, sociology, etc.) have identified as conveying social signals, i.e. relational attitudes towards others and social situations. The rationale is that such cues are the physical, machine detectable and synthesizable evidence of phenomena nonotherwise accessible to computers such as empathy, roles, dominance, personality, (dis-)agreement, interest, etc. After providing a brief state-of-the-art of the domain, the paper outlines its future perspectives and some of its most promising applications.