Abstract

Computing that is sensitive to affective and social phenomena aims to equip devices and interfaces with the means to interpret, understand, and respond to human personality, affect, moods and intentions – similarly to how humans rely on their senses to assess each other’s affective and social behaviour.

This talk will focus on automatic recognition of affect and social signals, and will present an overview of recent research works I have conducted in these fields. Automatic affect recognition will be covered in the context of human-computer interaction (expression and affect recognition of individuals) and co-located human-human interactions (analysis of dyadic interactions for collaborative music making, and multi-person affect recognition). Social signal processing will be explored in the context of automatic prediction of trait impressions in human-virtual character interactions and human-robot interactions.