Accent classification in human speech biometrics for native and non-native english speakers

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Accent classification provides a biometric path to high resolution speech recognition. This preliminary study explores various methods of human accent recognition through classification of locale. Classical, ensemble, timeseries and deep learning techniques are all explored and compared. A set of diphthong vowel sounds are recorded from participants from the United Kingdom and Mexico, and then formed into a large static dataset of statistical descriptions by way of their Mel-frequency Cepstral Coefficients (MFCC) at a sample window length of 0.02 seconds. Using both flat and timeseries data, various machine learning models are trained and compared to the scientific standard Hidden Markov Model (HMM). Results through 10 fold cross validation show that a vote of average probabilities between a Random Forest and Long Short-term Memory Neural Network result in a classification accuracy of 94.74%, outperforming the speech classification standard Hidden Markov Model by a 5% increase in accuracy.
LanguageEnglish
Title of host publicationProceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019
PublisherACM
Pages554-560
Number of pages7
ISBN (Electronic)9781450362320
ISBN (Print)978-1-4503-6232-0
DOIs
Publication statusPublished - 5 Jun 2019
Eventthe 12th ACM International Conference - Rhodes, Greece
Duration: 5 Jun 20197 Jun 2019

Conference

Conferencethe 12th ACM International Conference
Period5/06/197/06/19

Fingerprint

Biometrics
Hidden Markov models
Speech recognition
Learning systems
Acoustic waves
Neural networks

Keywords

  • Accent Recognition
  • Biometrics
  • Computational Linguistics
  • Machine Learning
  • Speech Recognition
  • Voice Assistants

Cite this

Bird, J. J., Wanner, E., Ekárt, A., & Faria, D. R. (2019). Accent classification in human speech biometrics for native and non-native english speakers. In Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019 (pp. 554-560). ACM. https://doi.org/10.1145/3316782.3322780
Bird, Jordan J. ; Wanner, Elizabeth ; Ekárt, Anikó ; Faria, Diego R. / Accent classification in human speech biometrics for native and non-native english speakers. Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019. ACM, 2019. pp. 554-560
@inproceedings{23e56284b08744cfa425e8e639d96819,
title = "Accent classification in human speech biometrics for native and non-native english speakers",
abstract = "Accent classification provides a biometric path to high resolution speech recognition. This preliminary study explores various methods of human accent recognition through classification of locale. Classical, ensemble, timeseries and deep learning techniques are all explored and compared. A set of diphthong vowel sounds are recorded from participants from the United Kingdom and Mexico, and then formed into a large static dataset of statistical descriptions by way of their Mel-frequency Cepstral Coefficients (MFCC) at a sample window length of 0.02 seconds. Using both flat and timeseries data, various machine learning models are trained and compared to the scientific standard Hidden Markov Model (HMM). Results through 10 fold cross validation show that a vote of average probabilities between a Random Forest and Long Short-term Memory Neural Network result in a classification accuracy of 94.74{\%}, outperforming the speech classification standard Hidden Markov Model by a 5{\%} increase in accuracy.",
keywords = "Accent Recognition, Biometrics, Computational Linguistics, Machine Learning, Speech Recognition, Voice Assistants",
author = "Bird, {Jordan J.} and Elizabeth Wanner and Anik{\'o} Ek{\'a}rt and Faria, {Diego R.}",
year = "2019",
month = "6",
day = "5",
doi = "10.1145/3316782.3322780",
language = "English",
isbn = "978-1-4503-6232-0",
pages = "554--560",
booktitle = "Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019",
publisher = "ACM",
address = "United States",

}

Bird, JJ, Wanner, E, Ekárt, A & Faria, DR 2019, Accent classification in human speech biometrics for native and non-native english speakers. in Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019. ACM, pp. 554-560, the 12th ACM International Conference, 5/06/19. https://doi.org/10.1145/3316782.3322780

Accent classification in human speech biometrics for native and non-native english speakers. / Bird, Jordan J.; Wanner, Elizabeth; Ekárt, Anikó; Faria, Diego R.

Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019. ACM, 2019. p. 554-560.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Accent classification in human speech biometrics for native and non-native english speakers

AU - Bird, Jordan J.

AU - Wanner, Elizabeth

AU - Ekárt, Anikó

AU - Faria, Diego R.

PY - 2019/6/5

Y1 - 2019/6/5

N2 - Accent classification provides a biometric path to high resolution speech recognition. This preliminary study explores various methods of human accent recognition through classification of locale. Classical, ensemble, timeseries and deep learning techniques are all explored and compared. A set of diphthong vowel sounds are recorded from participants from the United Kingdom and Mexico, and then formed into a large static dataset of statistical descriptions by way of their Mel-frequency Cepstral Coefficients (MFCC) at a sample window length of 0.02 seconds. Using both flat and timeseries data, various machine learning models are trained and compared to the scientific standard Hidden Markov Model (HMM). Results through 10 fold cross validation show that a vote of average probabilities between a Random Forest and Long Short-term Memory Neural Network result in a classification accuracy of 94.74%, outperforming the speech classification standard Hidden Markov Model by a 5% increase in accuracy.

AB - Accent classification provides a biometric path to high resolution speech recognition. This preliminary study explores various methods of human accent recognition through classification of locale. Classical, ensemble, timeseries and deep learning techniques are all explored and compared. A set of diphthong vowel sounds are recorded from participants from the United Kingdom and Mexico, and then formed into a large static dataset of statistical descriptions by way of their Mel-frequency Cepstral Coefficients (MFCC) at a sample window length of 0.02 seconds. Using both flat and timeseries data, various machine learning models are trained and compared to the scientific standard Hidden Markov Model (HMM). Results through 10 fold cross validation show that a vote of average probabilities between a Random Forest and Long Short-term Memory Neural Network result in a classification accuracy of 94.74%, outperforming the speech classification standard Hidden Markov Model by a 5% increase in accuracy.

KW - Accent Recognition

KW - Biometrics

KW - Computational Linguistics

KW - Machine Learning

KW - Speech Recognition

KW - Voice Assistants

UR - http://dl.acm.org/citation.cfm?doid=3316782.3322780

UR - http://www.scopus.com/inward/record.url?scp=85069152198&partnerID=8YFLogxK

U2 - 10.1145/3316782.3322780

DO - 10.1145/3316782.3322780

M3 - Conference contribution

SN - 978-1-4503-6232-0

SP - 554

EP - 560

BT - Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019

PB - ACM

ER -

Bird JJ, Wanner E, Ekárt A, Faria DR. Accent classification in human speech biometrics for native and non-native english speakers. In Proceedings of the 12th ACM International Conference on PErvasive Technologies Related to Assistive Environments, PETRA 2019. ACM. 2019. p. 554-560 https://doi.org/10.1145/3316782.3322780