Variable Span Filters for Speech Enhancement

Research output: Contribution to journalConference article in JournalResearchpeer-review

168 Downloads (Pure)

Abstract

In this work, we consider enhancement of multichannel speech recordings. Linear filtering and subspace approaches have been considered previously for solving the problem. The current linear filtering methods, although many variants exist, have limited control of noise reduction and speech distortion. Subspace approaches, on the other hand, can potentially yield better control by filtering in the eigen-domain, but traditionally these approaches have not been optimized explicitly for traditional noise reduction and signal distortion measures. Herein, we combine these approaches by deriving optimal filters using a joint diagonalization as a basis. This gives excellent control over the performance, as we can optimize for noise reduction or signal distortion performance. Results from real data experiments show that the proposed variable span filters can achieve better performance than existing filters. In terms of output SNR, the gain was more than 8~dB, and more than 0.1 in mean opinion score in the conducted experiments.
Original languageEnglish
JournalI E E E International Conference on Acoustics, Speech and Signal Processing. Proceedings
Number of pages5
ISSN1520-6149
DOIs
Publication statusPublished - Mar 2016
EventThe 41st IEEE International Conference on Acoustics, Speech and Signal Processing - Shanghai, China
Duration: 20 Mar 201625 Mar 2016
http://www.icassp2016.org/

Conference

ConferenceThe 41st IEEE International Conference on Acoustics, Speech and Signal Processing
CountryChina
CityShanghai
Period20/03/201625/03/2016
Internet address

Fingerprint

Speech enhancement
Noise abatement
Signal distortion
Experiments

Keywords

  • speech enhancement
  • joint diagonlization
  • optimal filtering
  • multichannel enhancement
  • tradeoff filter

Cite this

@inproceedings{c0713686d024421687f962f03b6da195,
title = "Variable Span Filters for Speech Enhancement",
abstract = "In this work, we consider enhancement of multichannel speech recordings. Linear filtering and subspace approaches have been considered previously for solving the problem. The current linear filtering methods, although many variants exist, have limited control of noise reduction and speech distortion. Subspace approaches, on the other hand, can potentially yield better control by filtering in the eigen-domain, but traditionally these approaches have not been optimized explicitly for traditional noise reduction and signal distortion measures. Herein, we combine these approaches by deriving optimal filters using a joint diagonalization as a basis. This gives excellent control over the performance, as we can optimize for noise reduction or signal distortion performance. Results from real data experiments show that the proposed variable span filters can achieve better performance than existing filters. In terms of output SNR, the gain was more than 8~dB, and more than 0.1 in mean opinion score in the conducted experiments.",
keywords = "speech enhancement, joint diagonalization, optimal filtering, multichannel enhancement, tradeoff filter, speech enhancement, joint diagonlization, optimal filtering, multichannel enhancement, tradeoff filter",
author = "Jensen, {Jesper Rindom} and Jacob Benesty and Christensen, {Mads Gr{\ae}sb{\o}ll}",
year = "2016",
month = "3",
doi = "10.1109/ICASSP.2016.7472930",
language = "English",
journal = "I E E E International Conference on Acoustics, Speech and Signal Processing. Proceedings",
issn = "1520-6149",
publisher = "IEEE Signal Processing Society",

}

Variable Span Filters for Speech Enhancement. / Jensen, Jesper Rindom; Benesty, Jacob; Christensen, Mads Græsbøll.

In: I E E E International Conference on Acoustics, Speech and Signal Processing. Proceedings, 03.2016.

Research output: Contribution to journalConference article in JournalResearchpeer-review

TY - GEN

T1 - Variable Span Filters for Speech Enhancement

AU - Jensen, Jesper Rindom

AU - Benesty, Jacob

AU - Christensen, Mads Græsbøll

PY - 2016/3

Y1 - 2016/3

N2 - In this work, we consider enhancement of multichannel speech recordings. Linear filtering and subspace approaches have been considered previously for solving the problem. The current linear filtering methods, although many variants exist, have limited control of noise reduction and speech distortion. Subspace approaches, on the other hand, can potentially yield better control by filtering in the eigen-domain, but traditionally these approaches have not been optimized explicitly for traditional noise reduction and signal distortion measures. Herein, we combine these approaches by deriving optimal filters using a joint diagonalization as a basis. This gives excellent control over the performance, as we can optimize for noise reduction or signal distortion performance. Results from real data experiments show that the proposed variable span filters can achieve better performance than existing filters. In terms of output SNR, the gain was more than 8~dB, and more than 0.1 in mean opinion score in the conducted experiments.

AB - In this work, we consider enhancement of multichannel speech recordings. Linear filtering and subspace approaches have been considered previously for solving the problem. The current linear filtering methods, although many variants exist, have limited control of noise reduction and speech distortion. Subspace approaches, on the other hand, can potentially yield better control by filtering in the eigen-domain, but traditionally these approaches have not been optimized explicitly for traditional noise reduction and signal distortion measures. Herein, we combine these approaches by deriving optimal filters using a joint diagonalization as a basis. This gives excellent control over the performance, as we can optimize for noise reduction or signal distortion performance. Results from real data experiments show that the proposed variable span filters can achieve better performance than existing filters. In terms of output SNR, the gain was more than 8~dB, and more than 0.1 in mean opinion score in the conducted experiments.

KW - speech enhancement

KW - joint diagonalization

KW - optimal filtering

KW - multichannel enhancement

KW - tradeoff filter

KW - speech enhancement

KW - joint diagonlization

KW - optimal filtering

KW - multichannel enhancement

KW - tradeoff filter

U2 - 10.1109/ICASSP.2016.7472930

DO - 10.1109/ICASSP.2016.7472930

M3 - Conference article in Journal

JO - I E E E International Conference on Acoustics, Speech and Signal Processing. Proceedings

JF - I E E E International Conference on Acoustics, Speech and Signal Processing. Proceedings

SN - 1520-6149

ER -