Multi-Pitch Estimation and Tracking Using Bayesian Inference in Block Sparsity

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceeding

Abstract

In this paper, we consider the problem of multi-pitch estimation and tracking of an unknown number of harmonic audio sources. The regularized least-squares is a solution for simultaneous sparse source selection and parameter estimation. Exploiting block sparsity, the method allows for reliable tracking of the found sources, without posing detailed a priori assumptions of the number of harmonics for each source. The method incorporates a Bayesian prior and assigns data-dependent regularization coefficients to efficiently incorporate both earlier and future data blocks in the tracking of estimates. In comparison with fix regularization coefficients, the simulation results, using both real and synthetic audio signals, confirm the performance of the proposed method.
Close

Details

In this paper, we consider the problem of multi-pitch estimation and tracking of an unknown number of harmonic audio sources. The regularized least-squares is a solution for simultaneous sparse source selection and parameter estimation. Exploiting block sparsity, the method allows for reliable tracking of the found sources, without posing detailed a priori assumptions of the number of harmonics for each source. The method incorporates a Bayesian prior and assigns data-dependent regularization coefficients to efficiently incorporate both earlier and future data blocks in the tracking of estimates. In comparison with fix regularization coefficients, the simulation results, using both real and synthetic audio signals, confirm the performance of the proposed method.
Original languageEnglish
Title of host publication2015 Proceedings of the 23rd European Signal Processing Conference (EUSIPCO 2015)
PublisherIEEE
Publication date2015
Pages16-20
ISBN (Print)978-0-9928626-3-3
DOI
StatePublished - 2015
Publication categoryResearch
Peer-reviewedYes
EventEuropean Signal Processing Conference (EUSIPCO) - Nice, France
Duration: 31 Aug 20154 Sep 2015

Conference

ConferenceEuropean Signal Processing Conference (EUSIPCO)
LandFrance
ByNice
Periode31/08/201504/09/2015
SeriesProceedings of the European Signal Processing Conference (EUSIPCO)
ISSN2076-1465

    Research areas

  • Multi-pitch estimation, tracking, harmonic signal, regularized least-squares, sparsity

Download statistics

No data available
ID: 212823313