SEDA: A tunable Q-factor wavelet-based noise reduction algorithm for multi-talker babble

Roozbeh Soleymani, Ivan Selesnick, David M. Landsberger

Research output: Contribution to journalArticle

Abstract

We introduce a new wavelet-based algorithm to enhance the quality of speech corrupted by multi-talker babble noise. The algorithm comprises three stages: The first stage classifies short frames of the noisy speech as speech-dominated or noise-dominated. We design this classifier specifically for multi-talker babble noise. The second stage performs preliminary de-nosing of noisy speech frames using oversampled wavelet transforms and parallel group thresholding. The final stage performs further denoising by attenuating residual high frequency components in the signal produced by the second stage. A significant improvement in intelligibility and quality was observed in evaluation tests of the algorithm with cochlear implant users.

Original languageEnglish (US)
Pages (from-to)102-115
Number of pages14
JournalSpeech Communication
Volume96
DOIs
StatePublished - Feb 1 2018

Fingerprint

Noise Reduction
Noise abatement
Wavelets
Cochlear implants
test evaluation
Implant
Thresholding
Denoising
Wavelet transforms
Wavelet Transform
Classifiers
Classify
Classifier
Speech
Talkers
Evaluation
Group

ASJC Scopus subject areas

  • Software
  • Modeling and Simulation
  • Communication
  • Language and Linguistics
  • Linguistics and Language
  • Computer Vision and Pattern Recognition
  • Computer Science Applications

Cite this

SEDA : A tunable Q-factor wavelet-based noise reduction algorithm for multi-talker babble. / Soleymani, Roozbeh; Selesnick, Ivan; Landsberger, David M.

In: Speech Communication, Vol. 96, 01.02.2018, p. 102-115.

Research output: Contribution to journalArticle

@article{7a8686d7d9cb4233bb7bbfa6efcfbc07,
title = "SEDA: A tunable Q-factor wavelet-based noise reduction algorithm for multi-talker babble",
abstract = "We introduce a new wavelet-based algorithm to enhance the quality of speech corrupted by multi-talker babble noise. The algorithm comprises three stages: The first stage classifies short frames of the noisy speech as speech-dominated or noise-dominated. We design this classifier specifically for multi-talker babble noise. The second stage performs preliminary de-nosing of noisy speech frames using oversampled wavelet transforms and parallel group thresholding. The final stage performs further denoising by attenuating residual high frequency components in the signal produced by the second stage. A significant improvement in intelligibility and quality was observed in evaluation tests of the algorithm with cochlear implant users.",
author = "Roozbeh Soleymani and Ivan Selesnick and Landsberger, {David M.}",
year = "2018",
month = "2",
day = "1",
doi = "10.1016/j.specom.2017.11.004",
language = "English (US)",
volume = "96",
pages = "102--115",
journal = "Speech Communication",
issn = "0167-6393",
publisher = "Elsevier",

}

TY - JOUR

T1 - SEDA

T2 - A tunable Q-factor wavelet-based noise reduction algorithm for multi-talker babble

AU - Soleymani, Roozbeh

AU - Selesnick, Ivan

AU - Landsberger, David M.

PY - 2018/2/1

Y1 - 2018/2/1

N2 - We introduce a new wavelet-based algorithm to enhance the quality of speech corrupted by multi-talker babble noise. The algorithm comprises three stages: The first stage classifies short frames of the noisy speech as speech-dominated or noise-dominated. We design this classifier specifically for multi-talker babble noise. The second stage performs preliminary de-nosing of noisy speech frames using oversampled wavelet transforms and parallel group thresholding. The final stage performs further denoising by attenuating residual high frequency components in the signal produced by the second stage. A significant improvement in intelligibility and quality was observed in evaluation tests of the algorithm with cochlear implant users.

AB - We introduce a new wavelet-based algorithm to enhance the quality of speech corrupted by multi-talker babble noise. The algorithm comprises three stages: The first stage classifies short frames of the noisy speech as speech-dominated or noise-dominated. We design this classifier specifically for multi-talker babble noise. The second stage performs preliminary de-nosing of noisy speech frames using oversampled wavelet transforms and parallel group thresholding. The final stage performs further denoising by attenuating residual high frequency components in the signal produced by the second stage. A significant improvement in intelligibility and quality was observed in evaluation tests of the algorithm with cochlear implant users.

UR - http://www.scopus.com/inward/record.url?scp=85037544634&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85037544634&partnerID=8YFLogxK

U2 - 10.1016/j.specom.2017.11.004

DO - 10.1016/j.specom.2017.11.004

M3 - Article

VL - 96

SP - 102

EP - 115

JO - Speech Communication

JF - Speech Communication

SN - 0167-6393

ER -