Output details
11 - Computer Science and Informatics
University College London
Automatic Recognition of Non-Acted Affective Postures
<20>This is the first paper to propose an automatic emotion recognition system based on body posture in naturalistic game situations. Previous works focused on acted data. This is instrumental for the design of more effective entertainment and serious games where emotions are critical for their success. The work is based on extensive perceptual studies with a thorough statistical analysis to identify emotion-discriminative body features and the creation and testing of the automatic emotion recognition system. It led to an invited 2012- TEDxStMartin talk with 870 views and to be invited to write a chapter in the Handbook of Affective Computing.