Alpha-Stable Low-Rank Plus Residual Decomposition for Speech Enhancement
Umut Şimşekli, Halil Erdoğan, Simon Leglaive, Antoine Liutkus, Roland Badeau, Gaël Richard · 2018
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 α-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.