Recognising human emotions from body movement and gesture dynamics

Abstract

We present an approach for the recognition of acted emotional states based on the analysis of body movement and gesture expressivity. According to research showing that distinct emotions are often associated with different qualities of body movement, we use non-propositional movement qualities (e.g. amplitude, speed and fluidity of movement) to infer emotions, rather than trying to recognise different gesture shapes expressing specific emotions. We propose a method for the analysis of emotional behaviour based on both direct classification of time series and a model that provides indicators describing the dynamics of expressive motion cues. Finally we show and interpret the recognition rates for both proposals using different classification algorithms. © Springer-Verlag Berlin Heidelberg 2007.