Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection

Mpho Mashika; Dustin van der Haar

doi:10.1007/978-3-031-35748-0_18

Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection

Mpho Mashika
, Dustin van der Haar

Academy of Computer Science and Software Engineering

University of Johannesburg

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

5 Citations (Scopus)

Abstract

Asthma, pneumonia, chronic obstructive pulmonary disease (COPD), and most recently, the covid-19 illness all include cough as one of its most noticeable symptoms. This paper proposes a method that uses audio signals to detect cough events. To train the models, we used data obtained from the ESC-50 dataset. We built models based on different features selected from Mel Frequency Cepstral Coefficients (MFCC), Zero Crossing Rate (ZCR), and Energy. The classification algorithms are K-NN, Support Vector Machine (SVM), and Multilayer Perceptron (MLP). The best model used the MFFC features and the SVM classification algorithm. The best model realised an accuracy of 92.20%, a precision of 92.86%, a recall of 91.55%, and an F1-score of 92.20%.

Original language	English
Title of host publication	Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings
Editors	Vincent G. Duffy
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	250-259
Number of pages	10
ISBN (Print)	9783031357473
DOIs	https://doi.org/10.1007/978-3-031-35748-0_18
Publication status	Published - 2023
Event	14th International Conference Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management, DHM 2023, held as part of the 25th International Conference on Human-Computer Interaction, HCII 2023 - Copenhagen, Denmark Duration: 23 Jul 2023 → 28 Jul 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14029 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	14th International Conference Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management, DHM 2023, held as part of the 25th International Conference on Human-Computer Interaction, HCII 2023
Country/Territory	Denmark
City	Copenhagen
Period	23/07/23 → 28/07/23

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Keywords

Cough Detection
mel-frequency cepstral coefficient
support vector machine

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-031-35748-0_18

Cite this

Mashika, M., & van der Haar, D. (2023). Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection. In V. G. Duffy (Ed.), Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings (pp. 250-259). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14029 LNCS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-35748-0_18

Mashika, Mpho ; van der Haar, Dustin. / Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection. Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings. editor / Vincent G. Duffy. Springer Science and Business Media Deutschland GmbH, 2023. pp. 250-259 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{fed09ded8827430b86587f0c0333e2c7,

title = "Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection",

abstract = "Asthma, pneumonia, chronic obstructive pulmonary disease (COPD), and most recently, the covid-19 illness all include cough as one of its most noticeable symptoms. This paper proposes a method that uses audio signals to detect cough events. To train the models, we used data obtained from the ESC-50 dataset. We built models based on different features selected from Mel Frequency Cepstral Coefficients (MFCC), Zero Crossing Rate (ZCR), and Energy. The classification algorithms are K-NN, Support Vector Machine (SVM), and Multilayer Perceptron (MLP). The best model used the MFFC features and the SVM classification algorithm. The best model realised an accuracy of 92.20\%, a precision of 92.86\%, a recall of 91.55\%, and an F1-score of 92.20\%.",

keywords = "Cough Detection, mel-frequency cepstral coefficient, support vector machine",

author = "Mpho Mashika and \{van der Haar\}, Dustin",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.; 14th International Conference Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management, DHM 2023, held as part of the 25th International Conference on Human-Computer Interaction, HCII 2023 ; Conference date: 23-07-2023 Through 28-07-2023",

year = "2023",

doi = "10.1007/978-3-031-35748-0\_18",

language = "English",

isbn = "9783031357473",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "250--259",

editor = "Duffy, \{Vincent G.\}",

booktitle = "Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings",

address = "Germany",

}

Mashika, M & van der Haar, D 2023, Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection. in VG Duffy (ed.), Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14029 LNCS, Springer Science and Business Media Deutschland GmbH, pp. 250-259, 14th International Conference Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management, DHM 2023, held as part of the 25th International Conference on Human-Computer Interaction, HCII 2023, Copenhagen, Denmark, 23/07/23. https://doi.org/10.1007/978-3-031-35748-0_18

Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection. / Mashika, Mpho; van der Haar, Dustin.
Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings. ed. / Vincent G. Duffy. Springer Science and Business Media Deutschland GmbH, 2023. p. 250-259 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14029 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection

AU - Mashika, Mpho

AU - van der Haar, Dustin

PY - 2023

Y1 - 2023

N2 - Asthma, pneumonia, chronic obstructive pulmonary disease (COPD), and most recently, the covid-19 illness all include cough as one of its most noticeable symptoms. This paper proposes a method that uses audio signals to detect cough events. To train the models, we used data obtained from the ESC-50 dataset. We built models based on different features selected from Mel Frequency Cepstral Coefficients (MFCC), Zero Crossing Rate (ZCR), and Energy. The classification algorithms are K-NN, Support Vector Machine (SVM), and Multilayer Perceptron (MLP). The best model used the MFFC features and the SVM classification algorithm. The best model realised an accuracy of 92.20%, a precision of 92.86%, a recall of 91.55%, and an F1-score of 92.20%.

AB - Asthma, pneumonia, chronic obstructive pulmonary disease (COPD), and most recently, the covid-19 illness all include cough as one of its most noticeable symptoms. This paper proposes a method that uses audio signals to detect cough events. To train the models, we used data obtained from the ESC-50 dataset. We built models based on different features selected from Mel Frequency Cepstral Coefficients (MFCC), Zero Crossing Rate (ZCR), and Energy. The classification algorithms are K-NN, Support Vector Machine (SVM), and Multilayer Perceptron (MLP). The best model used the MFFC features and the SVM classification algorithm. The best model realised an accuracy of 92.20%, a precision of 92.86%, a recall of 91.55%, and an F1-score of 92.20%.

KW - Cough Detection

KW - mel-frequency cepstral coefficient

KW - support vector machine

UR - https://www.scopus.com/pages/publications/85169470932

U2 - 10.1007/978-3-031-35748-0_18

DO - 10.1007/978-3-031-35748-0_18

M3 - Conference contribution

AN - SCOPUS:85169470932

SN - 9783031357473

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 250

EP - 259

BT - Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings

A2 - Duffy, Vincent G.

PB - Springer Science and Business Media Deutschland GmbH

T2 - 14th International Conference Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management, DHM 2023, held as part of the 25th International Conference on Human-Computer Interaction, HCII 2023

Y2 - 23 July 2023 through 28 July 2023

ER -

Mashika M, van der Haar D. Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection. In Duffy VG, editor, Digital Human Modeling and Applications in Health, Safety, Ergonomics and Risk Management - 14th International Conference, DHM 2023, Held as Part of the 25th HCI International Conference, HCII 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2023. p. 250-259. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-031-35748-0_18

Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection

Abstract

Publication series

Conference

UN SDGs

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this