Human-Human Interactional Synchrony Analysis Based on Body Sensor Networks

Abstract

Human-human interactions are widespread in many fields, but rarely investigated by using body sensor networks (BSNs). Due to the fact that distributed sensors can only provide local information, this paper proposes to analyse the interactional synchrony from both local and overall perspectives. The proposed framework analyses synchrony mainly including windowed cross-correlation, peak picking method and power average operator. Experiments conducted in this paper demonstrate that the proposed framework may not only distinguish between synchrony and asynchrony conditions, but also reveal different situations corresponding to different body parts.