Automatic Accent and Gender Recognition of Regional UK Speakers

Chrisina Jayne, Victor Chang, Jozeene Bailey, Qianwen Ariel Xu

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

22 Downloads (Pure)

Abstract

With the ubiquity of voice assistants across the UK and the world, speech recognition of the regional accents across the British Isles has proven challenging due to varying pronunciations. This paper proposes an automated recognition of the geographical origin and gender of a voice sample based on the six regional dialects of the United Kingdom. Twenty six features are extracted from 17,877 voice samples and then used to design, implement and evaluate machine learning classifiers based on Artificial Neural Networks (ANNs), Support Vector Machine (SVM), Random Forest (RF) and k-nearest neighbors (k-NN) algorithms. The results suggest that the proposed approach could be applicable for areas such as e-commerce and the service industry, and it provides a contribution to NLP audio research.

Original languageEnglish
Title of host publicationEngineering Applications of Neural Networks - 23rd International Conference, EAAAI/EANN 2022, Proceedings
EditorsLazaros Iliadis, Chrisina Jayne, Anastasios Tefas, Elias Pimenidis
PublisherSpringer Science and Business Media Deutschland GmbH
Pages67-80
Number of pages14
ISBN (Print)9783031082221
DOIs
Publication statusPublished - 2022
Event23rd International Conference on Engineering Applications of Neural Networks, EANN 2022 - Chersonisos, Greece
Duration: 17 Jun 202220 Jun 2022

Publication series

NameCommunications in Computer and Information Science
Volume1600 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference23rd International Conference on Engineering Applications of Neural Networks, EANN 2022
Country/TerritoryGreece
CityChersonisos
Period17/06/2220/06/22

Bibliographical note

Publisher Copyright:
© 2022, Springer Nature Switzerland AG.

Fingerprint

Dive into the research topics of 'Automatic Accent and Gender Recognition of Regional UK Speakers'. Together they form a unique fingerprint.

Cite this