||Affective computing is an interdisciplinary research field that aims to study, design and develop artificial systems that can recognize, interpret, and respond to human affect and emotions. Examples of affective technologies include social robots, virtual characters, computer games, intelligent interfaces, and applications range from therapy and rehabilitation to education to entertainment. Social perception abilities are amongst the most important skills necessary for artificial systems to engage with humans users in natural forms of interaction. These include affect sensitivity, that is, the ability to recognize people?s affective expressions and states, understand their social signals, and account for the context in which the interaction takes place. Building software for the automatic recognition of people?s emotions involves using data from a number of sensors to detect body language, facial expressions, tone of voice and physiological indicators such as electro-dermal activity, and applying machine learning algorithms to build models from this data.
This lecture will discuss methods to endow machines with automatic affect recognition abilities, as well as open challenges. A case study will be presented to show how the integration of these abilities can lead to machines capable of adapting to human users in a more socially-aware and personalised way.