Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

7 Citationer (Scopus)
50 Downloads (Pure)

Resumé

Advances in speech signal analysis facilitate the development of techniques for remote biomedical voice assessment. However, the performance of these techniques is affected by noise and distortion in signals. In this paper, we focus on the vowel /a/ as the most widely-used voice signal for pathological voice assessments and investigate the impact of four major types of distortion that are commonly present during recording or transmission in voice analysis, namely: background noise, reverberation, clipping and compression, on Mel-frequency cepstral coefficients (MFCCs) – the most widely-used features in biomedical voice analysis. Then, we propose a new distortion classification approach to detect the most dominant distortion in such voice signals. The proposed method involves MFCCs as frame-level features and a support vector machine as classifier to detect the presence and type of distortion in frames of a given voice signal. Experimental results obtained from the healthy and Parkinson’s voices show the effectiveness of the proposed approach in distortion detection and classification.
OriginalsprogEngelsk
TitelProc. INTERSPEECH 2017
Antal sider5
Publikationsdatoaug. 2017
Sider289-293
DOI
StatusUdgivet - aug. 2017
BegivenhedInterspeech 2017 - Stockholm, Sverige
Varighed: 20 aug. 201724 aug. 2017
http://www.interspeech2017.org/

Konference

KonferenceInterspeech 2017
LandSverige
ByStockholm
Periode20/08/201724/08/2017
Internetadresse
NavnProceedings of the International Conference on Spoken Language Processing
ISSN1990-9772

Fingerprint

Processing
Reverberation
Signal analysis
Acoustic noise
Support vector machines
Classifiers

Citer dette

Poorjam, A. H., Jensen, J. R., Little, M. A., & Christensen, M. G. (2017). Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis. I Proc. INTERSPEECH 2017 (s. 289-293). Proceedings of the International Conference on Spoken Language Processing https://doi.org/10.21437/Interspeech.2017-378
Poorjam, Amir Hossein ; Jensen, Jesper Rindom ; Little, Max A ; Christensen, Mads Græsbøll. / Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis. Proc. INTERSPEECH 2017. 2017. s. 289-293 (Proceedings of the International Conference on Spoken Language Processing).
@inproceedings{9688fdcf1a34431e8a91af14ac891a6e,
title = "Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis",
abstract = "Advances in speech signal analysis facilitate the development of techniques for remote biomedical voice assessment. However, the performance of these techniques is affected by noise and distortion in signals. In this paper, we focus on the vowel /a/ as the most widely-used voice signal for pathological voice assessments and investigate the impact of four major types of distortion that are commonly present during recording or transmission in voice analysis, namely: background noise, reverberation, clipping and compression, on Mel-frequency cepstral coefficients (MFCCs) – the most widely-used features in biomedical voice analysis. Then, we propose a new distortion classification approach to detect the most dominant distortion in such voice signals. The proposed method involves MFCCs as frame-level features and a support vector machine as classifier to detect the presence and type of distortion in frames of a given voice signal. Experimental results obtained from the healthy and Parkinson’s voices show the effectiveness of the proposed approach in distortion detection and classification.",
keywords = "distortion analysis, MFCC, remote biomedical voice assessment, Support vector machine (SVM)",
author = "Poorjam, {Amir Hossein} and Jensen, {Jesper Rindom} and Little, {Max A} and Christensen, {Mads Gr{\ae}sb{\o}ll}",
year = "2017",
month = "8",
doi = "10.21437/Interspeech.2017-378",
language = "English",
pages = "289--293",
booktitle = "Proc. INTERSPEECH 2017",

}

Poorjam, AH, Jensen, JR, Little, MA & Christensen, MG 2017, Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis. i Proc. INTERSPEECH 2017. Proceedings of the International Conference on Spoken Language Processing, s. 289-293, Stockholm, Sverige, 20/08/2017. https://doi.org/10.21437/Interspeech.2017-378

Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis. / Poorjam, Amir Hossein; Jensen, Jesper Rindom; Little, Max A; Christensen, Mads Græsbøll.

Proc. INTERSPEECH 2017. 2017. s. 289-293 (Proceedings of the International Conference on Spoken Language Processing).

Publikation: Bidrag til bog/antologi/rapport/konference proceedingKonferenceartikel i proceedingForskningpeer review

TY - GEN

T1 - Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis

AU - Poorjam, Amir Hossein

AU - Jensen, Jesper Rindom

AU - Little, Max A

AU - Christensen, Mads Græsbøll

PY - 2017/8

Y1 - 2017/8

N2 - Advances in speech signal analysis facilitate the development of techniques for remote biomedical voice assessment. However, the performance of these techniques is affected by noise and distortion in signals. In this paper, we focus on the vowel /a/ as the most widely-used voice signal for pathological voice assessments and investigate the impact of four major types of distortion that are commonly present during recording or transmission in voice analysis, namely: background noise, reverberation, clipping and compression, on Mel-frequency cepstral coefficients (MFCCs) – the most widely-used features in biomedical voice analysis. Then, we propose a new distortion classification approach to detect the most dominant distortion in such voice signals. The proposed method involves MFCCs as frame-level features and a support vector machine as classifier to detect the presence and type of distortion in frames of a given voice signal. Experimental results obtained from the healthy and Parkinson’s voices show the effectiveness of the proposed approach in distortion detection and classification.

AB - Advances in speech signal analysis facilitate the development of techniques for remote biomedical voice assessment. However, the performance of these techniques is affected by noise and distortion in signals. In this paper, we focus on the vowel /a/ as the most widely-used voice signal for pathological voice assessments and investigate the impact of four major types of distortion that are commonly present during recording or transmission in voice analysis, namely: background noise, reverberation, clipping and compression, on Mel-frequency cepstral coefficients (MFCCs) – the most widely-used features in biomedical voice analysis. Then, we propose a new distortion classification approach to detect the most dominant distortion in such voice signals. The proposed method involves MFCCs as frame-level features and a support vector machine as classifier to detect the presence and type of distortion in frames of a given voice signal. Experimental results obtained from the healthy and Parkinson’s voices show the effectiveness of the proposed approach in distortion detection and classification.

KW - distortion analysis

KW - MFCC

KW - remote biomedical voice assessment

KW - Support vector machine (SVM)

U2 - 10.21437/Interspeech.2017-378

DO - 10.21437/Interspeech.2017-378

M3 - Article in proceeding

SP - 289

EP - 293

BT - Proc. INTERSPEECH 2017

ER -

Poorjam AH, Jensen JR, Little MA, Christensen MG. Dominant distortion classification for pre-processing of vowels in remote biomedical voice analysis. I Proc. INTERSPEECH 2017. 2017. s. 289-293. (Proceedings of the International Conference on Spoken Language Processing). https://doi.org/10.21437/Interspeech.2017-378