Multichannel HR-NMF for modelling convolutive mixtures of non-stationary signals in the time-frequency domain

Roland Badeau, Mark D. Plumbley

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

Abstract

Several probabilistic models involving latent components have been proposed for modelling time-frequency (TF) representations of audio signals (such as spectrograms), notably in the nonnegative matrix factorization (NMF) literature. Among them, the recent high resolution NMF (HR-NMF) model is able to take both phases and local correlations in each frequency band into account, and its potential has been illustrated in applications such as source separation and audio inpainting. In this paper, HR-NMF is extended to multichannel signals and to convolutive mixtures. A fast variational expectation-maximization (EM) algorithm is proposed to estimate the enhanced model. This algorithm is applied to a stereophonic piano signal, and proves capable of accurately modelling reverberation and restoring missing observations.

Original languageEnglish
Title of host publication2013 IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2013
DOIs
Publication statusPublished - 1 Dec 2013
Externally publishedYes
Event2013 14th IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2013 - New Paltz, NY, United States
Duration: 20 Oct 201323 Oct 2013

Publication series

NameIEEE Workshop on Applications of Signal Processing to Audio and Acoustics

Conference

Conference2013 14th IEEE Workshop on Applications of Signal Processing to Audio and Acoustics, WASPAA 2013
Country/TerritoryUnited States
CityNew Paltz, NY
Period20/10/1323/10/13

Keywords

  • Multichannel signal analysis
  • Non-stationary signal modelling
  • Separation of convolutive mixtures
  • Time-frequency analysis
  • Variational EM algorithm

Fingerprint

Dive into the research topics of 'Multichannel HR-NMF for modelling convolutive mixtures of non-stationary signals in the time-frequency domain'. Together they form a unique fingerprint.

Cite this