A sensor pairing and fusion system for a multi-user environment

This paper proposes a system for sensor pairing and fusion in an interactive multi user environment. Using the system, we integrated mobile accelerometer and fixed position optical tracking methods, and implemented two active music listening ap plications based on the movement interaction from one or more users. We found that an acceleration domain similarity index between the two tracking methods is able to pair the raw interac tion streams in near real time, and that concurrent sampling of the streams allows for easy sensor fusion. Although these al gorithms still require further refinement, we believe that combin ing accurate position with accurate acceleration data is beneficial for novel interactive applications.