Bayesian Modelling of Temporal Structure in Musical Audio

This paper presents a probabilistic model of temporal structure in music which allows joint inference of tempo, meter and rhythmic pattern. The framework of the model naturally quantifies these three musical concepts in terms of hidden state-variables, allowing resolution of otherwise apparent ambiguities in musical structure. At the heart of the system is a probabilistic model of a hypothetical ‘bar-pointer’ which maps an input signal to one cycle of a latent, periodic rhythmical pattern. The system flexibly accommodates different input signals via two observation models: a Poisson points model for use with MIDI onset data and a Gaussian process model for use with raw audio signals. The discrete state-space permits exact computation of posterior probability distributions for the quantities of interest. Results are presented for both observation models, demonstrating the ability of the system to correctly detect changes in rhythmic pattern and meter, whilst tracking tempo.