Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition

Direction-of-arrival (DOA) estimation is a fundamental technique in array signal processing due to its wide applications in beamforming, speech enhancement and many other assistive speech processing technologies. In this paper, we devise a novel DOA technique based on randomized singular value decom...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Serkan Tokgoz, Issa M. S. Panahi
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Hearing aid device low SNR non-uniform microphone arrays randomized algorithm real-time implementation singular value decomposition Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/55bae1f996e24073ac53f1f8037ab17c
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:55bae1f996e24073ac53f1f8037ab17c
record_format	dspace
spelling	oai:doaj.org-article:55bae1f996e24073ac53f1f8037ab17c2021-12-03T00:00:36ZRobust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition2169-353610.1109/ACCESS.2021.3130180https://doaj.org/article/55bae1f996e24073ac53f1f8037ab17c2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9624952/https://doaj.org/toc/2169-3536Direction-of-arrival (DOA) estimation is a fundamental technique in array signal processing due to its wide applications in beamforming, speech enhancement and many other assistive speech processing technologies. In this paper, we devise a novel DOA technique based on randomized singular value decomposition (RSVD) to improve the performance of non-uniform non-linear microphone arrays (NUNLA). The accurate and efficient singular value decomposition of large data matrices is computationally challenging, and randomization provides an effective tool for performing matrix approximation, therefore, the developed DOA estimation utilizes a modified dictionary-based RSVD method for localizing single speech sources under low signal-to-noise ratios (SNR). Unlike previous methods developed for uniform linear microphone arrays, the proposed approach with L-shaped three microphone setup has no ‘left-right’ ambiguity. We present the performance of our proposed method in comparison to other techniques. The demonstrated experiments shows at-least 20% performance improvement using simulated data and 25% performance improvement using real data when compared with similar DoA estimation techniques for NUNLA. The proposed method exploits frame-based online time delay of arrival (TDOA) measurements which facilitates the proposed algorithm to run on real-time devices. We also show an efficient real-time implementation of the proposed method on a Pixel 3 Android smartphone using its built-in three microphones for hearing aid applications.Serkan TokgozIssa M. S. PanahiIEEEarticleHearing aid devicelow SNRnon-uniform microphone arraysrandomized algorithmreal-time implementationsingular value decompositionElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 157800-157811 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Hearing aid device low SNR non-uniform microphone arrays randomized algorithm real-time implementation singular value decomposition Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Hearing aid device low SNR non-uniform microphone arrays randomized algorithm real-time implementation singular value decomposition Electrical engineering. Electronics. Nuclear engineering TK1-9971 Serkan Tokgoz Issa M. S. Panahi Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition
description	Direction-of-arrival (DOA) estimation is a fundamental technique in array signal processing due to its wide applications in beamforming, speech enhancement and many other assistive speech processing technologies. In this paper, we devise a novel DOA technique based on randomized singular value decomposition (RSVD) to improve the performance of non-uniform non-linear microphone arrays (NUNLA). The accurate and efficient singular value decomposition of large data matrices is computationally challenging, and randomization provides an effective tool for performing matrix approximation, therefore, the developed DOA estimation utilizes a modified dictionary-based RSVD method for localizing single speech sources under low signal-to-noise ratios (SNR). Unlike previous methods developed for uniform linear microphone arrays, the proposed approach with L-shaped three microphone setup has no ‘left-right’ ambiguity. We present the performance of our proposed method in comparison to other techniques. The demonstrated experiments shows at-least 20% performance improvement using simulated data and 25% performance improvement using real data when compared with similar DoA estimation techniques for NUNLA. The proposed method exploits frame-based online time delay of arrival (TDOA) measurements which facilitates the proposed algorithm to run on real-time devices. We also show an efficient real-time implementation of the proposed method on a Pixel 3 Android smartphone using its built-in three microphones for hearing aid applications.
format	article
author	Serkan Tokgoz Issa M. S. Panahi
author_facet	Serkan Tokgoz Issa M. S. Panahi
author_sort	Serkan Tokgoz
title	Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition
title_short	Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition
title_full	Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition
title_fullStr	Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition
title_full_unstemmed	Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition
title_sort	robust three-microphone speech source localization using randomized singular value decomposition
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/55bae1f996e24073ac53f1f8037ab17c
work_keys_str_mv	AT serkantokgoz robustthreemicrophonespeechsourcelocalizationusingrandomizedsingularvaluedecomposition AT issamspanahi robustthreemicrophonespeechsourcelocalizationusingrandomizedsingularvaluedecomposition
_version_	1718373987901243392

Robust Three-Microphone Speech Source Localization Using Randomized Singular Value Decomposition

Ejemplares similares