A Generalized Bayesian Model for Tracking Long Metrical Cycles in Acoustic Music Signals

This is a companion page for the paper

A Generalized Bayesian Model for Tracking Long Metrical Cycles in Acoustic Music Signals

Ajay Srinivasamurthy, Andre Holzapfel, Ali Taylan Cemgil, Xavier Serra

Proceedings of the 41st IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP 2016) (pp. 76–80), 2016, Shanghai, China

Abstract: Most musical phenomena involve repetitive structures that enable listeners to track meter, i.e. the tactus or beat, the longer over-arching measure or bar, and possibly other related layers. Meters with long measure duration, sometimes lasting more than a minute, occur in many music cultures, e.g. from India, Turkey, and Korea. However, current meter tracking algorithms, which were devised for cycles of a few seconds length, cannot process such structures accurately. We present a novel generalization to an existing Bayesian model for meter tracking that overcomes this limitation. The proposed model is evaluated on a set of Indian Hindustani music recordings, and we document significant performance increase over the previous models. The presented model opens the way for computational analysis of performances with long metrical cycles, and has important applications in music studies as well as in commercial applications that involve such musics.

[Full text, BibTeX]

This webpage provides additional resources to the paper. 

Examples of Hindustani music

A few audio examples of Hindustani music can be found here: 

http://compmusic.upf.edu/examples-taal-hindustani

Dataset

The dataset used in the paper is described in detail here: 

http://compmusic.upf.edu/hindustani-rhythm-dataset

Code

A MATLAB implementation of the bar pointer model (BPM) and the section pointer model (SPM) will be made available for research purposes. Please contact us at the email address below if you want access to the code repository.