Output details
11 - Computer Science and Informatics
Teesside University
Evaluating multimodal affective fusion using physiological signals
<20> Multimodal aspects of affective signals have received little attention outside of improving recognition performance. In particular multimodal affective fusion is relatively unexplored. This paper demonstrates the effectiveness of an affective multimodal fusion method based around the PAD dimensional model of emotion. It is one of few systems to leverage the integrative power of a continuous dimensional model of emotion and to successfully evaluate temporal properties of affective fusion across interactive sessions. The described system was demonstrated at the ICT 2008 exhibition of EC-funded projects, attended by 4,500 delegates, and was featured on the Euronews coverage of the event.