Variational regularized two-dimensional nonnegative matrix factorization with the flexible β-Divergence for single channel source separation

Kaiwen Yu, W. L. Woo, S. S. Dlay

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

1 Citation (Scopus)

Abstract

This paper presents an algorithm for nonnegative matrix factorization 2D (NMF-2D) with the flexible β-Divergence. The β-Divergence is a group of cost functions parameterized by a single parameter β. The Least Squares divergence, Kullback-Leibler divergence and the Itakura-Saito divergence are special cases (β=2,1,0).This paper presents a more complete and holistic algorithm which uses a flexible range of β, instead of being limited to the special cases. We describe a maximization minimization (MM) algorithm lead to multiplicative updates. The proposed factorization decomposes an information-bearing matrix into two-dimensional convolution of factor matrices that represent the spectral dictionary and temporal codes with enhanced performance. The method is demonstrated on the separation of audio mixtures recorded from a single channel. The method also enables a generalized criterion for variable sparseness to be imposed onto the solution. Experimental tests and comparisons with other factorization methods have been conducted to verify the efficacy of the proposed method.

Original languageEnglish
Title of host publication2nd IET International Conference on Intelligent Signal Processing 2015 (ISP)
PublisherIEEE
ISBN (Electronic)978-1-78561-137-7
ISBN (Print)978-1-78561-136-0
DOIs
Publication statusPublished - 17 Nov 2016
Externally publishedYes
Event2nd IET International Conference on Intelligent Signal Processing 2015, ISP 2015 - London, United Kingdom
Duration: 1 Dec 20152 Dec 2015

Conference

Conference2nd IET International Conference on Intelligent Signal Processing 2015, ISP 2015
CountryUnited Kingdom
CityLondon
Period1/12/152/12/15

Fingerprint Dive into the research topics of 'Variational regularized two-dimensional nonnegative matrix factorization with the flexible β-Divergence for single channel source separation'. Together they form a unique fingerprint.

Cite this