Feature enhancement of reverberant speech by distribution matching and non-negative matrix factorization

Loading...
Thumbnail Image
Journal Title
Journal ISSN
Volume Title
A1 Alkuperäisartikkeli tieteellisessä aikakauslehdessä
This publication is imported from Aalto University research portal.
View publication in the Research portal
View/Open full text file from the Research portal
Date
2015
Department
Major/Subject
Mcode
Degree programme
Language
en
Pages
1-14
Series
EURASIP JOURNAL ON ADVANCES IN SIGNAL PROCESSING, Volume 2015
Abstract
This paper describes a novel two-stage dereverberation feature enhancement method for noise-robust automatic speech recognition. In the first stage, an estimate of the dereverberated speech is generated by matching the distribution of the observed reverberant speech to that of clean speech, in a decorrelated transformation domain that has a long temporal context in order to address the effects of reverberation. The second stage uses this dereverberated signal as an initial estimate within a non-negative matrix factorization framework, which jointly estimates a sparse representation of the clean speech signal and an estimate of the convolutional distortion. The proposed feature enhancement method, when used in conjunction with automatic speech recognizer back-end processing, is shown to improve the recognition performance compared to three other state-of-the-art techniques.
Description
Keywords
Speech dereverberation, Feature enhancement, Non-negative matrix factorization, Distribution matching
Other note
Citation
Keronen , S , Kallasjoki , H , Palomaki , K J , Brown , G J & Gemmeke , J F 2015 , ' Feature enhancement of reverberant speech by distribution matching and non-negative matrix factorization ' , Eurasip Journal on Advances in Signal Processing , vol. 2015 , 76 , pp. 1-14 . https://doi.org/10.1186/s13634-015-0259-1