The Emotion Sign: Human Motion Analysis Classifying Specific Emotion

Abstract

We examine the relationship between human motion and emotions. With recent improvement of sensing technologies, although precise human motion can be measured, the amount of data grows enormously. In this paper, we propose a new analysis method which can describe large amount of data rationally. This method can be used to classify human motions associated with specific emotions. Our approach to motion data analysis is to apply higher order singular value decomposition (HOSVD) directly to motion data. HOSVD can generate a compact vector which specifies each emotion common among people. Experimentally, we obtained motion capture data for “gait” and “standing” actions related to six basic emotions. Human gait motion was also created with an animator. For these motion data, our analysis showed that our method can classify the human motions specific to each emotion.