Skip to Main content Skip to Navigation
Conference papers

A tempering approach for Itakura-Saito non-negative matrix factorization. With application to music transcription

Abstract : In this paper we are interested in non-negative matrix factorization (NMF) with the Itakura-Saito (IS) divergence. Previous work has demonstrated the relevance of this cost function for the decomposition of audio power spectrograms. This is in particular due to its scale invariance, which makes it more robust to the wide dynamics of audio, a property which is not shared by other popular costs such as the Euclidean distance or the generalized Kulback-Leibler (KL) divergence. However, while the latter two cost functions are convex, the IS divergence is not, which makes it more prone to convergence to irrelevant local minima, as observed empirically. Thus, the aim of this paper is to propose a tempering scheme that favors convergence of IS-NMF to global minima. Our algorithm is based on NMF with the beta-divergence, where the shape parameter beta acts as a temperature parameter. Results on both synthetical and music data (in a transcription context) show the relevance of our approach.
Complete list of metadata

Cited literature [9 references]  Display  Hide  Download

https://hal.inria.fr/hal-00945283
Contributor : Roland Badeau <>
Submitted on : Tuesday, March 25, 2014 - 8:41:59 AM
Last modification on : Thursday, January 7, 2021 - 8:18:13 PM
Long-term archiving on: : Wednesday, June 25, 2014 - 10:43:29 AM

File

Bertin-ICASSP2009.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00945283, version 1

Collections

Citation

Nancy Bertin, Cédric Févotte, Roland Badeau. A tempering approach for Itakura-Saito non-negative matrix factorization. With application to music transcription. Proc. of IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2009, Taipei, Taiwan. pp.1545--1548. ⟨hal-00945283⟩

Share

Metrics

Record views

448

Files downloads

656