Skip to Main Content
Emotion awareness has become one of the key characteristics in human-computer interactions in order to achieve more natural and intelligent communications. As an alternative channel for emotional communication, physiological signals have gradually earned attentions in the field of affective computing. Ever since Dr. Picard and colleagues brought forward the initial concept of physiological signals based emotion recognition, various studies have been reported following the same system structure, which tends to perform worse when dealing with multiple subjects. In this paper, we proposed a novel biometric signature based system that is able to improve the performance of multi-subject emotion recognition. The basic idea is to transform a subject-independent case into several subject-dependent cases by using mixture multivariate t-distributions and Maximum a Posteriori (MAP) rule to remove the influence of individual varieties. The result shows great improvement compared to the traditional subject-independent procedure.