Advanced

Speech enhancement with natural sounding residual noise based on connected time-frequency speech presence regions

Sorensen, KV and Andersen, Sören Vang LU (2005) In Eurasip Journal on Applied Signal Processing 2005(18).
Abstract
We propose time-frequency domain methods for noise estimation and speech enhancement. A speech presence detection method is used to find connected time-frequency regions of speech presence. These regions are used by a noise estimation method and both the speech presence decisions and the noise estimate are used in the speech enhancement method. Different attenuation rules are applied to regions with and without speech presence to achieve enhanced speech with natural sounding attenuated background noise. The proposed speech enhancement method has a computational complexity, which makes it feasible for application in hearing aids. An informal listening test shows that the proposed speech enhancement method has significantly higher mean... (More)
We propose time-frequency domain methods for noise estimation and speech enhancement. A speech presence detection method is used to find connected time-frequency regions of speech presence. These regions are used by a noise estimation method and both the speech presence decisions and the noise estimate are used in the speech enhancement method. Different attenuation rules are applied to regions with and without speech presence to achieve enhanced speech with natural sounding attenuated background noise. The proposed speech enhancement method has a computational complexity, which makes it feasible for application in hearing aids. An informal listening test shows that the proposed speech enhancement method has significantly higher mean opinion scores than minimum mean-square error log-spectral amplitude (MMSE-LSA) and decision-directed MMSE-LSA. (Less)
Please use this url to cite or link to this publication:
author
publishing date
type
Contribution to journal
publication status
published
subject
keywords
speech enhancement, noise estimation, minimum statistics, speech, presence detection
in
Eurasip Journal on Applied Signal Processing
volume
2005
issue
18
publisher
Hindawi Publishing Corporation
external identifiers
  • wos:000235836400005
  • scopus:33751294092
ISSN
1110-8657
DOI
10.1155/ASP.2005.2954
language
English
LU publication?
no
id
dcb660a2-34e5-44a9-9d99-4c5512ec2630 (old id 4092530)
date added to LUP
2013-10-17 10:40:00
date last changed
2017-08-27 05:18:53
@article{dcb660a2-34e5-44a9-9d99-4c5512ec2630,
  abstract     = {We propose time-frequency domain methods for noise estimation and speech enhancement. A speech presence detection method is used to find connected time-frequency regions of speech presence. These regions are used by a noise estimation method and both the speech presence decisions and the noise estimate are used in the speech enhancement method. Different attenuation rules are applied to regions with and without speech presence to achieve enhanced speech with natural sounding attenuated background noise. The proposed speech enhancement method has a computational complexity, which makes it feasible for application in hearing aids. An informal listening test shows that the proposed speech enhancement method has significantly higher mean opinion scores than minimum mean-square error log-spectral amplitude (MMSE-LSA) and decision-directed MMSE-LSA.},
  articleno    = {305909},
  author       = {Sorensen, KV and Andersen, Sören Vang},
  issn         = {1110-8657},
  keyword      = {speech enhancement,noise estimation,minimum statistics,speech,presence detection},
  language     = {eng},
  number       = {18},
  publisher    = {Hindawi Publishing Corporation},
  series       = {Eurasip Journal on Applied Signal Processing},
  title        = {Speech enhancement with natural sounding residual noise based on connected time-frequency speech presence regions},
  url          = {http://dx.doi.org/10.1155/ASP.2005.2954},
  volume       = {2005},
  year         = {2005},
}