Home > Research > Publications & Outputs > A segmentally informed solution to automatic ac...

Associated organisational unit

Electronic data

  • segmental_solution_IJSLL_resub

    Accepted author manuscript, 571 KB, PDF document

    Embargo ends: 8/07/24

    Available under license: CC BY-NC-ND: Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License

Links

Text available via DOI:

View graph of relations

A segmentally informed solution to automatic accent classification and its advantages to forensic applications

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

A segmentally informed solution to automatic accent classification and its advantages to forensic applications. / Brown, Georgina; Franco-Pedroso, Javier; Gonzalez-Rodriguez, Joaquin.

In: International Journal of Speech, Language and the Law, Vol. 28, No. 2, 08.07.2022, p. 201-232.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Brown, G, Franco-Pedroso, J & Gonzalez-Rodriguez, J 2022, 'A segmentally informed solution to automatic accent classification and its advantages to forensic applications', International Journal of Speech, Language and the Law, vol. 28, no. 2, pp. 201-232. https://doi.org/10.1558/ijsll.20446

APA

Brown, G., Franco-Pedroso, J., & Gonzalez-Rodriguez, J. (2022). A segmentally informed solution to automatic accent classification and its advantages to forensic applications. International Journal of Speech, Language and the Law, 28(2), 201-232. https://doi.org/10.1558/ijsll.20446

Vancouver

Brown G, Franco-Pedroso J, Gonzalez-Rodriguez J. A segmentally informed solution to automatic accent classification and its advantages to forensic applications. International Journal of Speech, Language and the Law. 2022 Jul 8;28(2):201-232. doi: 10.1558/ijsll.20446

Author

Brown, Georgina ; Franco-Pedroso, Javier ; Gonzalez-Rodriguez, Joaquin. / A segmentally informed solution to automatic accent classification and its advantages to forensic applications. In: International Journal of Speech, Language and the Law. 2022 ; Vol. 28, No. 2. pp. 201-232.

Bibtex

@article{a7063c77749e455f946264fe1cf8b6c0,
title = "A segmentally informed solution to automatic accent classification and its advantages to forensic applications",
abstract = "Traditionally, work in automatic accent recognition has followed a similar research trajectory to that of language identification, dialect identification and automatic speaker recognition. The same acoustic modelling approaches that have been implemented in speaker recognition (such as GMM-UBM and i-vector-based systems) have also been applied to automatic accent recognition. These approaches form models of speakers{\textquoteright} accents by taking acoustic features from right across the speech signal without knowledge of its phonetic content. Particularly for accent recognition, however, phonetic information is expected to add substantial value to the task. The current work presents an alternative modelling approach to automatic accent recognition, which forms models of speakers{\textquoteright} pronunciation systems using segmental information. This article claims that such an approach to the problem makes for a more explainable method and therefore is a more appropriate method to deploy in settings where it is important to be able to communicate methods, such as forensic applications. We discuss the issue of explainability and show how the system operates on a large 700-speaker dataset of non-native English conversational telephone recordings.",
keywords = "automatic accent recognition, explainable technologies, segmental features, forensic applications",
author = "Georgina Brown and Javier Franco-Pedroso and Joaquin Gonzalez-Rodriguez",
year = "2022",
month = jul,
day = "8",
doi = "10.1558/ijsll.20446",
language = "English",
volume = "28",
pages = "201--232",
journal = "International Journal of Speech, Language and the Law",
issn = "1748-8885",
publisher = "Equinox Publishing Ltd",
number = "2",

}

RIS

TY - JOUR

T1 - A segmentally informed solution to automatic accent classification and its advantages to forensic applications

AU - Brown, Georgina

AU - Franco-Pedroso, Javier

AU - Gonzalez-Rodriguez, Joaquin

PY - 2022/7/8

Y1 - 2022/7/8

N2 - Traditionally, work in automatic accent recognition has followed a similar research trajectory to that of language identification, dialect identification and automatic speaker recognition. The same acoustic modelling approaches that have been implemented in speaker recognition (such as GMM-UBM and i-vector-based systems) have also been applied to automatic accent recognition. These approaches form models of speakers’ accents by taking acoustic features from right across the speech signal without knowledge of its phonetic content. Particularly for accent recognition, however, phonetic information is expected to add substantial value to the task. The current work presents an alternative modelling approach to automatic accent recognition, which forms models of speakers’ pronunciation systems using segmental information. This article claims that such an approach to the problem makes for a more explainable method and therefore is a more appropriate method to deploy in settings where it is important to be able to communicate methods, such as forensic applications. We discuss the issue of explainability and show how the system operates on a large 700-speaker dataset of non-native English conversational telephone recordings.

AB - Traditionally, work in automatic accent recognition has followed a similar research trajectory to that of language identification, dialect identification and automatic speaker recognition. The same acoustic modelling approaches that have been implemented in speaker recognition (such as GMM-UBM and i-vector-based systems) have also been applied to automatic accent recognition. These approaches form models of speakers’ accents by taking acoustic features from right across the speech signal without knowledge of its phonetic content. Particularly for accent recognition, however, phonetic information is expected to add substantial value to the task. The current work presents an alternative modelling approach to automatic accent recognition, which forms models of speakers’ pronunciation systems using segmental information. This article claims that such an approach to the problem makes for a more explainable method and therefore is a more appropriate method to deploy in settings where it is important to be able to communicate methods, such as forensic applications. We discuss the issue of explainability and show how the system operates on a large 700-speaker dataset of non-native English conversational telephone recordings.

KW - automatic accent recognition

KW - explainable technologies

KW - segmental features

KW - forensic applications

U2 - 10.1558/ijsll.20446

DO - 10.1558/ijsll.20446

M3 - Journal article

VL - 28

SP - 201

EP - 232

JO - International Journal of Speech, Language and the Law

JF - International Journal of Speech, Language and the Law

SN - 1748-8885

IS - 2

ER -