How Many Do We Need? Exploration of the Population Size Effect on the Performance of Forensic Speaker Classification

Shunichi Ishihara, Yuko Kinoshita

    Research output: A Conference proceeding or a Chapter in BookConference contributionpeer-review

    18 Citations (Scopus)

    Abstract

    This paper investigates how change in population size affects the reliability of the likelihood ratio (LR)-based forensic speaker classification. Using features of the long term F0 distribution, we performed LR-based speaker classification and examined its performance with population sizes from 10 to 120. The results revealed that LRs could be heavily influenced by the population data. We discovered that the reliability of LR-based evaluation of the evidence was heavily compromised if the population data was limited to a small number of speakers.
    Original languageEnglish
    Title of host publicationProceedings of the 9th Annual conference of International Speech Communication Association (Interspeech 2008)
    EditorsM Wagner
    Place of PublicationChina
    PublisherIEEE, Institute of Electrical and Electronics Engineers
    Pages1941-1944
    Number of pages4
    Volume1
    ISBN (Print)9780769531748
    Publication statusPublished - 2008
    EventInterspeech 2008 - Brisbane, Australia
    Duration: 22 Sept 200826 Sept 2008

    Conference

    ConferenceInterspeech 2008
    Country/TerritoryAustralia
    CityBrisbane
    Period22/09/0826/09/08

    Fingerprint

    Dive into the research topics of 'How Many Do We Need? Exploration of the Population Size Effect on the Performance of Forensic Speaker Classification'. Together they form a unique fingerprint.

    Cite this