Automatic scoring of synchronization from fingers motion capture and music beats

Abstract : Thanks to various technological progresses such as musical rhythm estimation and motion capture systems, the evaluation of synchronization performances between motion and music beats is today possible. In this paper, we propose an innovative playful method to assess synchronization between hand motion and music. In this application, the hand gestures are tracked from live webcam motion capture based on MediaPipe open-source framework. Musical beats are estimated from the Librosa library. The synchronization between finger motion and beats is then computed from the dynamic time warping algorithm. A preliminary study was conducted with two different songs mixed with tempo beats at different intensities. The results are encouraging and show different levels of performance according to the tempo of the songs.