Natural, Affect Aware interfaces: Gesture and Body Expressivity Aspects

Recently, a growing number of researchers focus their work on a fledgling field related to the identification, recording, interpreting, processing and simulation of emotion and affective states. Affective computing, along with the wide range of related application areas, is expanded increasingly due to the emerging demand for natural, intelligent, adaptive and personalized multimodal interaction context. This paper focuses on the gesture and bodily expressivity analysis, investigating and reporting the entire spectrum of related aspects such as issues involved in affectively enhanced corpora (stimuli and input streams), respective expressivity formalization approaches as well as machine learning and pattern recognition aspects (feature extraction, multimodal fusion and classification). Finally, the synthesis counterpart of affectively enhanced behavior is examined and future directions of the research area are identified.