Skip to main content

ETHNOS_APP

Home • Search • Journals • List 0

Aging effects on voice features used in forensic speaker comparison

Bibliographic Data

ID19535776
AuthorsRichard Rhodes (0000-0003-0320-4060, corresponding author)
Year2017
Volume24
Issue2
Pages177-199
Publication date2017-12-20
Open AccessNo
TypeARTICLE
VenueInternational Journal of Speech Language and the Law (JOURNAL)
Journal identifiersISSN: 1748-8885 • E-ISSN: 1748-8893
PublisherUniversity of Toronto Press Inc. (UTPress) (PUBLISHER)
DOI10.1558/ijsll.34096
OpenAlexW2779728617
LanguageEN
Citations received2

This article assesses the impact of long-term non-contemporaneity between recordings on the strength of forensic voice evidence and provides recommendations for casework. Analyses of longitudinal recordings from a British television documentary series illustrate the effects of aging on forensically useful acoustic parameters. Recordings from eight speakers span five seven-year intervals, between ages 21 and 49. The frequency of the first three formants of nine monophthongs and two diphthongs decrease over time. Strength of evidence estimates for non-contemporaneous comparisons are calculated from these data using a likelihood ratio approach. Longer delays predictably result in weaker and fewer correct LRs. The effect of aging on the performance of an ASR system (BATVOX) is tested; performance varies between speakers, but deteriorates considerably in longer delays for all speakers. Findings from this and similar studies should be considered when carrying out formant- or ASR-based comparisons across long delays, and in selecting age-appropriate reference data

Audiology · Diphthong · Formant · Speech recognition · Computer Science · Medicine · Music and Audio Processing · Psychology · Speech and Audio Processing · Speech Recognition and Synthesis

  • Who owns your voice? Linguistic and legal perspectives on the relationship between vocal distinctiveness and the rights of the individual speaker

    Dominic Watt, Peter S Harrison et al.•International Journal of Speech…•2020

  • The impact of mismatched recordings on an automatic-speaker-recognition system and human listeners

    Open Access•Tomáš Nechanský, Tomáš Bořil et al.•AUC PHILOLOGICA•2023

Citations per year0.33
Citation span2020 - 2023 (4)
Highly citedNo

Tools

Open DOI
Ethnos_APP • Open Source Project • MIT License • Frontend v2.0.0 • Privacy and Cookies • API Documentation: api.ethnos.app/docs • API Source Code: GitHub • DOI: 10.5281/zenodo.17049435 • Frontend Source Code: GitHub • DOI: 10.5281/zenodo.17050053 • cruz.rio.br • Expectantes Misericordiae