Alpha-Stable Low-Rank Plus Residual Decomposition for Speech Enhancement

In this study, we propose a novel probabilistic model for separating clean speech signals from noisy mixtures by decomposing the mixture spectra into a structured speech part and a more flexible residual part. The main novelty in our model is that it uses a family of heavy-tailed distributions, so c...

Full description

Saved in:
Bibliographic Details
Published in2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) pp. 651 - 655
Main Authors Simsekli, Umut, Erdogan, Halil, Leglaive, Simon, Liutkus, Antoine, Badeau, Roland, Richard, Gael
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.04.2018
Subjects
Alpha-stable distributions
Approximation algorithms
Audio source separation
Computational modeling
Gaussian distribution
Monte Carlo Expectation-Maximization
Optimization
Probabilistic logic
Speech enhancement
Time-frequency analysis
Online AccessGet full text

Cover

More Information
Summary:In this study, we propose a novel probabilistic model for separating clean speech signals from noisy mixtures by decomposing the mixture spectra into a structured speech part and a more flexible residual part. The main novelty in our model is that it uses a family of heavy-tailed distributions, so called the \alpha -stable distributions, for modeling the residual signal. We develop an expectation-maximization algorithm for parameter estimation and a Monte Carlo scheme for posterior estimation of the clean speech. Our experiments show that the proposed method outperforms relevant factorization-based algorithms by a significant margin.
ISSN:2379-190X
DOI:10.1109/ICASSP.2018.8461539