Automatic Alignment of Music Performances with Scores Aimed at Educational Applications

Abstract

We present a system for automatic real time alignment of an acoustic music performance with a digital representation of its score, a problem which is usually defined score following. The alignment is based on an application of hidden Markov models. A model is automatically built from a music score, while decoding is used to compute the most probable location of the performance along the score model. The effectiveness of the proposed approach has been tested with a collection of recordings of orchestral music. Even if the typical application of a score following system is automatic accompaniment, in this paper we propose a set of novel applications that are targeted also to non musicians, for educational use.