A multiple perception model on emotional speech

Jianhua Tao, Aijun Li, Shifeng Pan

Abstract: More and more efforts have been made for the research of emotional speech recently. Although we may, sometimes be able to make a definite perceptual decision on emotion state, emotion is actually a kind of cline in a large vector space. Different emotions can be thought of as zones along an emotional vector. To resolve the ambiguity of emotion perception, the authors make an array of perception experiments. Furthermore, the paper creates a statistical model to simulate emotion perception and finds that there is an underlying consistency in the patterns of responses. The results show important cues on how to do multiple perception of the emotional speech.