Speckled Tango Dancers: Real-Time Motion Capture of Two-Body Interactions Using On-body Wireless Sensor Networks

This paper describes the application of a wireless network of on-body inertial sensors for the capture of 3-D motion of Tango dancers. Accomplished Tango dancers exhibit both individual flair and good coordination between partners. We have identified features which reflect the performance of both the individual (such as the angle of bend of the dancer's chest and synchronisation of the movements between the chest and the feet), and of the partnership (correlation of the dancers' chest movements). These features have been captured in the live data from on-body sensors and used to characterise the dancers' performance. The aim in the future is to design a dance tutoring tool which will analyse the sensor data and provide immediate feedback for improvement. The system can be extended to the analysis of other activities involving physical movement which have to be captured unobtrusively for real-time analysis and feedback.