TY - GEN
T1 - Mel Frequency Cepstral Coefficients and Support Vector Machines for Cough Detection
AU - Mashika, Mpho
AU - van der Haar, Dustin
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
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 - http://www.scopus.com/inward/record.url?scp=85169470932&partnerID=8YFLogxK
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 -
