Advertisement

Determining the Likelihood Ratio From Perceptual Attributes of Voice

  • Jorge A. Gurlekian
    Correspondence
    Address correspondence and reprint requests to Jorge A. Gurlekian, Laboratorio de Investigaciones Sensoriales (LIS) INIGEM, Science and Justice Program, National Council of Scientific and Technological Research, Hospital de Clinicas. Ciudad de Buenos Aires, Argentina.
    Affiliations
    Laboratorio de Investigaciones Sensoriales (LIS) INIGEM, Science and Justice Program, National Council of Scientific and Technological Research, Hospital de Clinicas, Ciudad de Buenos Aires, Argentina
    Search for articles by this author
  • Stefania Suligoy
    Affiliations
    Acoustic Forensic Division, Direction of Criminalistic and Forensic Studies of the Argentine National Gendarmerie, Ciudad de Buenos Aires, Argentina
    Search for articles by this author
  • Pedro Univaso
    Affiliations
    Argentine Catholic University (UCA), BlackVox Company, Ciudad de Buenos Aires, Argentina
    Search for articles by this author
  • Humberto Torres
    Affiliations
    Laboratorio de Investigaciones Sensoriales (LIS) INIGEM, National Council of Scientific and Technological Research CONICET, Biomedical Institute, Faculty of Engineering, University of Buenos Aires, Ciudad de Buenos Aires, Argentina
    Search for articles by this author
  • Evangelina Masessa
    Affiliations
    Acoustic Forensic Division, Direction of Criminalistic and Forensic Studies of the Argentine National Gendarmerie, Ciudad de Buenos Aires, Argentina
    Search for articles by this author
  • Nancy Molina
    Affiliations
    Faculty of Medicine, University of Buenos Aires, Ciudad de Buenos Aires, Argentina
    Search for articles by this author

      Summary

      Purpose

      To provide voice experts with a method for determining the likelihood ratio (LR) from the perceptual evaluation of distinctive voice attribute scores. The proposed method aims to obtain the similarity and typicality judgments made by forensic voice experts (FVEs) during the comparison of attributes in voice pairs.

      Method

      It is based on the scoring method for LR calculation. In the first stage, 17 perceptual attributes grouped into six vocalic categories are specified. A novel graphical interface is used to obtain discriminative responses both globally and for each attribute from ten pairs of test sentences produced by the same and different speakers. The FVEs should discriminate whether the attributes are similar or different in each pair and should indicate the degree to which the attributes are present. In addition, for six specific attributes, the FVEs must decide whether the attribute is typical or atypical in the reference population. In the second stage, the mean score obtained in the first stage is converted to LR using probability density functions of listeners' responses to 1680 same/different speaker pairs discriminated for female and male speakers.

      Results

      The responses of the FVEs to the test pairs show the discriminatory power of the attributes, the incidence of the typicality factor on the final score and the performance of each FVE. With the application of the probability density functions obtained for the responses to pairs of the same or different origin taken from the reference population, the final scores are converted into LRs that are compared with the true conditions of each pair.

      Conclusions and future work

      The application of the developed system allows the global and discriminated evaluation of the perceptual attributes with high agreement in the comparison of pairs of voices. Obtaining the LRs allows associating the perceptual evaluation method with the automatic methods that are used nowadays. The responses of the FVEs taken as a reference, will allow training and evaluating the performance of young FVEs.

      Key Words

      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Journal of Voice
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      REFERENCES

        • Gold E
        • French P.
        An international investigation of forensic speaker comparison practices.
        in: Dans les actes de Proceedings of the 17th International Congress of Phonetic Sciences, Hong Kong, China2011: 1254-1257
        • Hollien H.
        About forensic phonetics.
        Linguistica. 2012; 52: 27-53https://doi.org/10.4312/linguistica.52.1.27-53
        • Nolan F.
        Speaker identification evidence: its forms, limitations, and roles.
        in: Dans les actes de Proceedings of the Conference “Law and Language: Prospect and Retrospect”. Levi, Finnish Lapland2001: 1-19
        • Rose P
        • Morrison G.
        A response to the UK position statement on forensic speaker comparison.
        Int J Speech Lang Law. 2009; 16: 139
      1. Daubert. Daubert v. Merrell Dow pharmaceuticals, Inc. 113 S Ct 2786. 1993.

        • Morrison GS
        • Enzinger E.
        Score based procedures for the calculation of forensic likelihood ratios – scores should take account of both similarity and typicality.
        Sci Justice. 2018; 58: 47-58https://doi.org/10.1016/j.scijus.2017.06.005
        • Rose P.
        Forensic Speaker Identification.
        Taylor & Francis, London2002
      2. Drygajlo A, Jessen M, Gfroerer S, et al. Methodological Guidelines for Best Practice in Forensic Semiautomatic and Automatic Speaker Recognition. Frankfurt: Verlag für Polizeiwissenschaft. 2016. http://enfsi.eu/wp-.content/uploads/2016/09/guidelines_fasr_and_fsasr_0.pdf

        • Evett IW
        • Weir BS.
        Interpreting DNA Evidence: Statistical Genetics for Forensic Scientists.
        Sinaur Associates. Inc, Sunderland, Massachusetts, 1998
        • Enzinger E.
        PhD Thesis.
        University of New South Wales, School of Electrical Engineering & Telecommunications, Faculty of Engineering, August 2015
      3. National Institute of Standards and Technology. Organizational structure. Accessed 17 January 2022. https://www.nist.gov/forensics/organization-scientific-area-committees-forensic-science

      4. National Institute of Standards and Technology, OSAC. Holistic audio perceptual analysis. 2022. https://www.nist.gov/system/files/documents/2019/11/05/speaker_recognition_process_map_20190930.pdf

        • Hughes V
        • Harrison PT
        • Foulkes P
        • et al.
        Mapping across feature spaces in forensic voice comparison: the contribution of auditory-based voice quality to (semi-) automatic system testing.
        in: Proceedings of Interspeech. Interspeech, Stockholm, Sweden2017 (York)
        • Park SJ
        • Afshan A
        • Kreiman J
        • et al.
        Target and non-target speaker discrimination by humans and machines.
        IEEE-ICASSP. 2019; : 6326-6330https://doi.org/10.1109/ICASSP.2019.8683362
        • González-Rodríguez J
        • Gil J
        • Pérez R
        • et al.
        What are we missing with i-vectors? A perceptual analysis of i-vector-based falsely accepted trials.
        Proceedings Odyssey. 2014; 2014: 33-40
        • Alexander A
        • Dessimoz D
        • Botti F
        • et al.
        Aural and automatic forensic speaker recognition in mismatched conditions.
        Int J Speech Lang Law. 2005; 2 (12): 214-234
        • Eriksson A.
        Aural/acoustic vs. automatic methods in forensic phonetic case work.
        in: Neustein A. Patil H.A. Forensic Speaker Recognition. Law Enforcement and Counter-Terrorism. Springer, Berlin2012: 41-69
        • Gurlekian JA
        • Güemes M
        • Evin Dy
        • et al.
        Normalización del texto “Los sentidos” y su aplicación en la evaluación de habla continua.
        Onomazein. 2021; 67 (Noin press)https://doi.org/10.7764/onomazein 67.06
        • Molina N
        • Masessa E
        • Suligoy S
        • et al.
        Hacia la evaluación perceptual forense. Transformación de puntajes de similitud a relaciones de verosimilitud. Parte 1: hablantes masculinos.
        Gaceta Internacional de Ciencias Forenses GICF. 2022; 42: 12-20
        • Masessa E
        • Molina N
        • Suligoy S
        • et al.
        Hacia la evaluación perceptual forense. Transformación de puntajes de similitud a relaciones de verosimilitud. Parte 2: hablantes femeninos.
        Gaceta Internacional de Ciencias Forenses GICF. 2022; 42: 21-29
        • Gurlekian JA
        • Torres HM
        • Rincón Cediel M.
        A perceptual method to rate dysphonic voices.
        J Voice. 2019; 33-4: 453-464https://doi.org/10.1016/j.jvoice. 2018.01.007
        • Van Leeuwen DA,
        • Brümmer N.
        An Introduction to Application-Independent Evaluation of Speaker Recognition Systems.
        Speaker Classification I: Fundamentals, Features, and Methods. Springer Berlin Heidelberg, Berlin, Heidelberg2007: 330-353
        • Hollien H.
        Forensic Voice Identification.
        Academic Press, 2002