Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation Localisation

This paper proposes a comprehensive study on machine listening for localisation of snore sound excitation. Here we investigate the effects of varied frame sizes, and overlap of the analysed audio chunk for extracting low-level descriptors. In addition, we explore the performance of each kind of feat...

Full description

Saved in:
Bibliographic Details
Main Authors: Kun QIAN, Christoph JANOTT, Zixing ZHANG, Jun DENG, Alice BAIRD, Clemens HEISER, Winfried HOHENHORST, Michael HERZOG, Werner HEMMERT, Björn SCHULLER
Format: Article
Language:English
Published: Institute of Fundamental Technological Research Polish Academy of Sciences 2018-07-01
Series:Archives of Acoustics
Subjects:
Online Access:https://acoustics.ippt.pan.pl/index.php/aa/article/view/2155
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849705442472624128
author Kun QIAN
Christoph JANOTT
Zixing ZHANG
Jun DENG
Alice BAIRD
Clemens HEISER
Winfried HOHENHORST
Michael HERZOG
Werner HEMMERT
Björn SCHULLER
author_facet Kun QIAN
Christoph JANOTT
Zixing ZHANG
Jun DENG
Alice BAIRD
Clemens HEISER
Winfried HOHENHORST
Michael HERZOG
Werner HEMMERT
Björn SCHULLER
author_sort Kun QIAN
collection DOAJ
description This paper proposes a comprehensive study on machine listening for localisation of snore sound excitation. Here we investigate the effects of varied frame sizes, and overlap of the analysed audio chunk for extracting low-level descriptors. In addition, we explore the performance of each kind of feature when it is fed into varied classifier models, including support vector machines, $k$-nearest neighbours, linear discriminant analysis, random forests, extreme learning machines, kernel-based extreme learning machines, multilayer perceptrons, and deep neural networks. Experimental results demonstrate that, wavelet packet transform energy can outperform most other features. A deep neural network trained with subband energy ratios reaches the highest performance achieving an unweighted average recall of 72.8% from four types for snoring.
format Article
id doaj-art-0ffbbd0c10f8431e92e235087780a592
institution DOAJ
issn 0137-5075
2300-262X
language English
publishDate 2018-07-01
publisher Institute of Fundamental Technological Research Polish Academy of Sciences
record_format Article
series Archives of Acoustics
spelling doaj-art-0ffbbd0c10f8431e92e235087780a5922025-08-20T03:16:28ZengInstitute of Fundamental Technological Research Polish Academy of SciencesArchives of Acoustics0137-50752300-262X2018-07-0143310.24425/123918Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation LocalisationKun QIAN0Christoph JANOTT1Zixing ZHANG2Jun DENG3Alice BAIRD4Clemens HEISER5Winfried HOHENHORST6Michael HERZOG7Werner HEMMERT8Björn SCHULLER9Technical University of Munich, University of PassauTechnical University of MunichUniversity of PassauaudEERING GmbHUniversity of PassauTechnical University of MunichClinic for ENT Medicine, Head and Neck Surgery, Alfried Krupp Krankenhaus, Essen, GermanyClinic for ENT Medicine, Head and Neck Surgery, Cottbus, GermanyTechnical University of MunichUniversity of Passau, Imperial College London, audEERING GmbHThis paper proposes a comprehensive study on machine listening for localisation of snore sound excitation. Here we investigate the effects of varied frame sizes, and overlap of the analysed audio chunk for extracting low-level descriptors. In addition, we explore the performance of each kind of feature when it is fed into varied classifier models, including support vector machines, $k$-nearest neighbours, linear discriminant analysis, random forests, extreme learning machines, kernel-based extreme learning machines, multilayer perceptrons, and deep neural networks. Experimental results demonstrate that, wavelet packet transform energy can outperform most other features. A deep neural network trained with subband energy ratios reaches the highest performance achieving an unweighted average recall of 72.8% from four types for snoring.https://acoustics.ippt.pan.pl/index.php/aa/article/view/2155snore soundobstructive sleep apneaacoustic featuresmachine learning
spellingShingle Kun QIAN
Christoph JANOTT
Zixing ZHANG
Jun DENG
Alice BAIRD
Clemens HEISER
Winfried HOHENHORST
Michael HERZOG
Werner HEMMERT
Björn SCHULLER
Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation Localisation
Archives of Acoustics
snore sound
obstructive sleep apnea
acoustic features
machine learning
title Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation Localisation
title_full Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation Localisation
title_fullStr Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation Localisation
title_full_unstemmed Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation Localisation
title_short Teaching Machines on Snoring: A Benchmark on Computer Audition for Snore Sound Excitation Localisation
title_sort teaching machines on snoring a benchmark on computer audition for snore sound excitation localisation
topic snore sound
obstructive sleep apnea
acoustic features
machine learning
url https://acoustics.ippt.pan.pl/index.php/aa/article/view/2155
work_keys_str_mv AT kunqian teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT christophjanott teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT zixingzhang teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT jundeng teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT alicebaird teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT clemensheiser teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT winfriedhohenhorst teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT michaelherzog teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT wernerhemmert teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation
AT bjornschuller teachingmachinesonsnoringabenchmarkoncomputerauditionforsnoresoundexcitationlocalisation