Assessing the effects of accent-mismatched reference population databases on the performance of an automatic speaker recognition system. (27th August 2020)

• Record Type: Journal Article
• Authors: Watt, Dominic; Harrison, Philip; Hughes, Vincent; French, Peter; Llamas, Carmen; Braun, Almut; Robertson, Duncan
• Is Part Of: International journal of speech language and the law. Volume 27:Number 1(2020)
• Page Start: 1
• Page End: 34
• Publication Date: 2020-08-27
• Subjects: forensic phonetics -- automatic speaker recognition -- forensic speaker comparison -- speech technology; Forensic linguistics -- Periodicals; 363.2505
• DOI: 10.1558/ijsll.41466
• Languages: English
• ISSNs: 1748-8885

Abstract

Abstract : Automatic Speaker Recognition (ASR) systems are designed to provide the user with statistics relating to the similarity of two or more speech samples and to the typicality of those shared features in the wider population. When an ASR system is used as part of a forensic investigation, the user must decide what counts as the appropriate 'wider population' and select a reference database accordingly. While it has generally been held that the voices populating the reference database should be similar in accent to that of the samples under consideration, the degree to which the accents should correspond has until now not been investigated empirically. We report in this article on a study in which the composition of the reference database was systematically varied in terms of accent, using corpora of samples of Standard Southern British English and of three subvarieties spoken in North-East England (Newcastle, Sunderland, Middlesbrough).