Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signals

Bone-conducted (BC) speech signals are inherently challenging to analyze due to their wide frequency range, which leads to ill-conditioning in numerical analysis and linear prediction (LP) techniques. This ill-conditioning is primarily caused by the expansion of eigenvalues, which complicates the st...

Full description

Saved in:
Bibliographic Details
Main Authors: Kanwar Muhammad Afaq, Ammar Amjad, Li-Chia Tai, Hsien-Tsung Chang
Format: Article
Language:English
Published: PeerJ Inc. 2025-05-01
Series:PeerJ Computer Science
Subjects:
Online Access:https://peerj.com/articles/cs-2906.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849328520780578816
author Kanwar Muhammad Afaq
Ammar Amjad
Li-Chia Tai
Hsien-Tsung Chang
author_facet Kanwar Muhammad Afaq
Ammar Amjad
Li-Chia Tai
Hsien-Tsung Chang
author_sort Kanwar Muhammad Afaq
collection DOAJ
description Bone-conducted (BC) speech signals are inherently challenging to analyze due to their wide frequency range, which leads to ill-conditioning in numerical analysis and linear prediction (LP) techniques. This ill-conditioning is primarily caused by the expansion of eigenvalues, which complicates the stability and accuracy of traditional methods. To address this issue, we propose a novel regularized spectral reduction (RSR) method, built upon the regularized least squares (RLS) framework. The RSR method compresses the frequency range of BC speech signals, effectively reducing eigenvalue spread and enhancing the robustness of LP analysis. Key to the RSR approach is a regularization parameter, fine-tuned iteratively to achieve optimal performance. Experimental results demonstrate that RSR significantly outperforms existing techniques in eigenvalue compression, resulting in more accurate LP analysis for both synthetic and real BC speech datasets. These improvements hold promise for applications in hearing aids, voice recognition systems, and speaker identification in noisy environments, where reliable BC speech analysis is critical.
format Article
id doaj-art-af3bdef6694d49f9adeb13b2cd385e63
institution Kabale University
issn 2376-5992
language English
publishDate 2025-05-01
publisher PeerJ Inc.
record_format Article
series PeerJ Computer Science
spelling doaj-art-af3bdef6694d49f9adeb13b2cd385e632025-08-20T03:47:33ZengPeerJ Inc.PeerJ Computer Science2376-59922025-05-0111e290610.7717/peerj-cs.2906Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signalsKanwar Muhammad Afaq0Ammar Amjad1Li-Chia Tai2Hsien-Tsung Chang3Department of Computer Science and Information Engineering, Chang Gung University, Taoyuan, TaiwanDepartment of Electrical and Computer Engineering, National Yang Ming Chiao Tung University, Hsinchu, TaiwanDepartment of Electrical and Computer Engineering, National Yang Ming Chiao Tung University, Hsinchu, TaiwanDepartment of Computer Science and Information Engineering, Chang Gung University, Taoyuan, TaiwanBone-conducted (BC) speech signals are inherently challenging to analyze due to their wide frequency range, which leads to ill-conditioning in numerical analysis and linear prediction (LP) techniques. This ill-conditioning is primarily caused by the expansion of eigenvalues, which complicates the stability and accuracy of traditional methods. To address this issue, we propose a novel regularized spectral reduction (RSR) method, built upon the regularized least squares (RLS) framework. The RSR method compresses the frequency range of BC speech signals, effectively reducing eigenvalue spread and enhancing the robustness of LP analysis. Key to the RSR approach is a regularization parameter, fine-tuned iteratively to achieve optimal performance. Experimental results demonstrate that RSR significantly outperforms existing techniques in eigenvalue compression, resulting in more accurate LP analysis for both synthetic and real BC speech datasets. These improvements hold promise for applications in hearing aids, voice recognition systems, and speaker identification in noisy environments, where reliable BC speech analysis is critical.https://peerj.com/articles/cs-2906.pdfSpectral compressionRegularization methodIll-conditioning improvementSpeech signal analysisBone-conducted voice signals
spellingShingle Kanwar Muhammad Afaq
Ammar Amjad
Li-Chia Tai
Hsien-Tsung Chang
Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signals
PeerJ Computer Science
Spectral compression
Regularization method
Ill-conditioning improvement
Speech signal analysis
Bone-conducted voice signals
title Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signals
title_full Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signals
title_fullStr Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signals
title_full_unstemmed Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signals
title_short Adaptive regularized spectral reduction for stabilizing ill-conditioned bone-conducted speech signals
title_sort adaptive regularized spectral reduction for stabilizing ill conditioned bone conducted speech signals
topic Spectral compression
Regularization method
Ill-conditioning improvement
Speech signal analysis
Bone-conducted voice signals
url https://peerj.com/articles/cs-2906.pdf
work_keys_str_mv AT kanwarmuhammadafaq adaptiveregularizedspectralreductionforstabilizingillconditionedboneconductedspeechsignals
AT ammaramjad adaptiveregularizedspectralreductionforstabilizingillconditionedboneconductedspeechsignals
AT lichiatai adaptiveregularizedspectralreductionforstabilizingillconditionedboneconductedspeechsignals
AT hsientsungchang adaptiveregularizedspectralreductionforstabilizingillconditionedboneconductedspeechsignals