Proposta de Construção de um Banco de Dados de Amostras de Fala para Uso Forense em um Arcabouço Bayesiano

Résumé

Há pouco mais de dez anos, um novo paradigma para algumas disciplinas forenses foi pela primeira vez descrito. Nesse novo modelo, o arcabouço Bayesiano para avaliação da evidência foi proposto como um modelo teórico comum para a interpretação da evidência. Nos exames de Comparação Forense de Locutor, os esforços para a adoção dessa abordagem já haviam sido iniciados bem antes do novo paradigma proposto, seja na metodologia automática, seja na metodologia combinada clássica. No Brasil, onde se tem optado pela adoção de um modelo binário de decisão e pela adoção de uma escala clássica de probabilidades para expressão dos resultados, os peritos têm, no arcabouço bayesiano, uma boa alternativa. A adoção desse arcabouço requer, porém, a obtenção de uma base de dados suficientemente representativa. O objetivo do presente artigo é fazer uma revisão do arcabouço teórico bayesiano nas metodologias utilizadas em Comparação Forense de Locutor, bem como propor um esboço de uma base de dados de fala forense a ser utilizado no Estado de São Paulo

Références

  1. M.J. Saks, J.J. Koehler. The coming paradigm shift in forensic identification science. Science 309, 892-895, 2005.
  2. T.S. Kuhn. The structure of scientific revolutions. University of Chicago Press, 2012.
  3. G.S. Morrison. Distinguishing between forensic science and forensic pseudoscience: Testing of validity and reliability, and approaches to forensic voice comparison. Science & Justice 54(3), 245-256, 2014.
  4. C. Aitken et al. Expressing evaluative opinions: a position statement. Science & justice 51(1), 1-2, 2011.
  5. C.A.B. Pereira et al. Genuine Bayesian multiallelic significance test for the Hardy-Weinberg equilibrium law. Genet. Mol. Res. 5, 619-631, 2006.
  6. A.N. Kolmogorov. Grundbegriffe der Wahrscheinlichkeitsrechnung, Berlin, 1933. English translation, Chelsea, New York, 1950.
  7. C.A.B. Pereira, J.M. Stern. Evidence and credibility: full Bayesian significance test for precise hypotheses. Entropy 1(4), 99-110, 1999.
  8. C.G.G. Aitken, F. Taroni. Statistics and the evaluation of evidence for forensic scientists. Chichester: Wiley, 2004.
  9. D. Lucy. Introduction to statistics for forensic scientists. John Wiley & Sons, 2013.
  10. D. van Lancker, J. Kreiman, K. Emmorey. Familiar voice recognition: Patterns and parameters. Part I: Recognition of backward voices. Journal of phonetics 13(1), 19-38, 1985.
  11. A.D. Yarmeyet al. Commonsense beliefs and the identification of familiar voices. Applied Cognitive Psychology 15(3), 283-299, 2001.
  12. R.M. Warren. Auditory perception: A new synthesis. Elsevier, 2013.
  13. K.N. Stevens. Sources of Inter - and Intra- speaker Variability in the Acoustics Properties of Speech Sounds. In: A. Rigault (Ed.). Actes Du SeptièmeCongrès International Des Sciences Phonétiques, Tenu À L'Université de Montréal Et À L'Université McGill, 22-28 Août 1971. Mouton, 1972.
  14. C.D. Romero. La identificación de locutores en el ámbito forense. Tese de Doutorado. Universidad Complutense de Madrid, 2001.
  15. P. ROSE. Forensic speaker identification. CRC Press, 2003.
  16. B. Robertson, G.A. Vignaux. Interpreting evidence: evaluating forensic science in the courtroom. 1995.
  17. C. Champod, I.W. Evett. Commentary on APA Broeders (1999) ‘Some observations on the use of probability scales in forensic identification’, Forensic Linguistics 6 (2): 228–41. International Journal of Speech Language and the Law 7(2), 239-243, 2007.
  18. E. Gold, P. French. International practices in forensic speaker comparison. International Journal of Speech Language and the Law 18, 2, 2011.
  19. A. Eriksson. Aural/Acoustic vs. Automatic Methods in Forensic Phonetic Case Work. In: Forensic Speaker Recognition. Springer New York, 41-69, 2012.
  20. . L. Rabiner, B.H. Juang. Fundamentals of speech recognition. 1993.
  21. J.P. Campbell Jr. Speaker recognition: A tutorial. Proceedings of the IEEE 85(9), 1437-1462, 1997.
  22. D. Meuwly. Voiceanalysis. Encyclopaedia of Forensic Sciences 3, 1413-1421, 2000.
  23. D. Meuwly. Forensic individualization from biometric data. Science & Justice 46(4), 205-213, 2006.
  24. F. Bimbot et al. A tutorial on text-independent speaker verification. EURASIP Journal on Applied Signal Processing, 430-451, 2004.
  25. C. Champod, D. Meuwly. The inference of identity in forensic speaker recognition. Speech Communication, 31(2), 193-203, 2000.
  26. D.A. Reynolds, T.F. Quatieri, R.B. Dunn. Speaker verification using adapted Gaussian mixture models. Digital signal processing 10(1), 19-41, 2000.
  27. F. Bimbot et al. A tutorial on text-independent speaker verification. EURASIP Journal on Applied Signal Processing, 430-451, 2004.
  28. D. Reynolds. An overview of automatic speaker recognition. In: Proceedings of the International Conference on Acoustics, Speech and Signal Processing (ICASSP)(S. 4072-4075). 2002.
  29. J. Gonzalez-Rodriguez et al. Robust estimation, interpretation and assessment of likelihood ratios in forensic speaker recognition. Computer Speech & Language 20(2), 331-355, 2006.
  30. N. Brümmeret al. Fusion of heterogeneous speaker recognition systems in the STBU submission for the NIST speaker recognition evaluation 2006.Audio, Speech, and Language Processing, IEEE Transactions on 15(7), 2072-2084, 2007.
  31. D.A. van Leeuwen, N. Brümmer. An introduction to application-independent evaluation of speaker recognition systems. In: Speaker classification I. Springer Berlin Heidelberg, 330-353, 2007.
  32. G.R. Doddington. The NIST speaker recognition evaluation–overview, methodology, systems, results, perspective. Speech Communication 31(2), 225-254, 2000.
  33. N.J.G. Pearce et al. A compilation of new and published major and trace element data for NIST SRM 610 and NIST SRM 612 glass reference materials. Geostandards Newsletter, 21(1), 115-144, 1997.
  34. D.E. Sturim, W.M. Campbell, D.A. Reynolds. Classification methods for speaker recognition. In: Speaker Classification I. Springer Berlin Heidelberg, 278-297, 2007.
  35. D. Meuwly, A. Drygajlo. Forensic speaker recognition based on a Bayesian framework and Gaussian mixture modeling (GMM). In: 2001: A Speaker Odyssey-The Speaker Recognition Workshop. 2001.
  36. A. Drygajlo, D. Meuwly, A. Alexander. Statistical methods and Bayesian interpretation of evidence in forensic automatic speaker recognition. In: INTERSPEECH, 2003.
  37. J.P. Campbell et al. Forensic speaker recognition. Institute of Electrical and Electronics Engineers, 2009.
  38. F. Botti, A. Alexander, A. Drygajlo. An interpretation framework for the evaluation of evidence in forensic automatic speaker recognition with limited suspect data. In: ODYSSEY04-The Speaker and Language Recognition Workshop. 2004.
  39. A. Alexander. Forensic automatic speaker recognition using Bayesian interpretation and statistical compensation for mismatched conditions. Tese de Doutorado. Institut de traitement des signaux. Section de Génie Électriqueet Électronique, Indian Institute of Technology, Madras, 2005.
  40. GOLD, Erica Ashley. Calculating likelihood ratios for forensic speaker comparisons using phonetic and linguistic parameters. Tese de Doutorado. The University of York, 2014.
  41. P. Rose. Long-and short-term within-speaker differences in the formants of Australian hello. Journal of the international Phonetic Association 29(1), 1-31, 1999.
  42. T. Alderman. The Bernard data set as a reference distribution for Bayesian likelihood ratio-based forensic speaker identification using formants. In: Proceedings of the 10th Australian International Conference on Speech Science and Technology, 510-515, 2004.
  43. P. Rose. Forensic speaker discrimination with Australian English vowel acoustics. ICPhS XVI Saarbrucken 6, 10, 2007.
  44. C. Zhang, G.S. Morrison, P. Rose. Forensic speaker recognition in Chinese: a multivariate likelihood ratio discrimination on/i/and/y/. In: Interspeech, 1937-1940, 2008.
  45. E. Enzinger. Characterizing formant tracks in Viennese diphthongs for forensic speaker comparison. In: Audio Engineering Society Conference: 39th International Conference: Audio Forensics: Practices and Challenges. Audio Engineering Society, 2010.
  46. Y. Kinoshita et al. Within speaker variation in diphthongal dynamics: What can we compare. In: Proceedings of the 11th Australasian International Conference on Speech Science & Technology, Auckland, New Zealand. Australia: Australasian Speech Science & Technology Association, Canberra, 112-117, 2006.
  47. G.S. Morrison. Likelihood-ratio forensic voice comparison using parametric representations of the formant trajectories of diphthongs).The Journal of the Acoustical Society of America 125(4), 2387-2397, 2009.
  48. T. Becker, M. Jessen, C. Grigoras. Forensic speaker verification using formant features and Gaussian mixture models. In: Interspeech, 1505-1508, 2008.
  49. C.M. Kavanagh. New consonantal acoustic parameters for forensic speaker comparison. Tese de Doutorado. University of York, 2012.
  50. Y. Kinoshita. Does Lindley's LR estimation formula work for speech data? Investigation using long-term F0.International Journal of Speech Language and the Law 12(2), 235-254, 2007.
  51. Y. Kinoshita, S. Ishihara, P. Rose. Exploring the discriminatory potential of F0 distribution parameters in traditional forensic speaker recognition. International Journal of Speech Language and the Law 16(1), 91-111, 2009.
  52. Y. Kinoshita. Testing realistic forensic speaker identification in Japanese: A likelihood ratio based approach using formants. Tese de Doutorado. Australian National University, 2001.
  53. A. Machado. Uso de técnicas acústicas para verificação de locutor em simulação experimental. Língua, Literatura e Ensino. ISSN 1981-6871, 6, 2012.
  54. A.C. Constantini, P.A. Barbosa. Prosodic characteristics of different varieties of Brazilian Portuguese. Rev. Bras. Crimin. 4(3), 44-53, 2015.
  55. G.S. Morrison. A comparison of procedures for the calculation of forensic likelihood ratios from acoustic–phonetic data: Multivariate kernel density (MVKD) versus Gaussian mixture model–universal background model (GMM–UBM). Speech Communication 53(2), 242-256, 2011.
  56. G.S. Morrison. Tutorial on logistic-regression calibration and fusion: converting a score to a likelihood ratio. Australian Journal of Forensic Science 45(2), 173-197, 2013.
  57. V. Hughes, P. Foulkes. Effects of variation on the computation of numerical likelihood ratios for forensic voice comparison. In: International Association of Forensic Phonetics and Acoustics conference, Universidad International Menedez Pelayo, Santander, 5th-8th August. 2012.
  58. P. Rose. The effect of correlation on strength of evidence estimates in Forensic Voice Comparison: uni-and multivariate Likelihood Ratio-based discrimination with Australian English vowel acoustics. International Journal of Biometrics 2(4), 316-329, 2010.
  59. G.S. Morrison. Tutorial on logistic-regression calibration and fusion: converting a score to a likelihood ratio. Australian Journal of Forensic Science 45(2), 173-197, 2013.
  60. F. Nolan et al. A forensic phonetic study of' dynamic sources of variability in speech: the DyViS project. In: Proceedings of the 11th Australasian International Conference on Speech Science and Technology, 13-18, 2006.
  61. G. de Jong, K. McDougal, F. Nolan. Sound change and speaker identity: an acoustic study. In: Speaker Classification II. Springer Berlin Heidelberg, 130-141, 2007.
  62. C. Zhang, G.S. Morrison, T. Thiruvaran. Forensic voice comparison using Chinese/iau. In: Proceedings of the 17th International Congress of Phonetic Sciences, 2280-2283, 2011.
  63. P. Rose. Catching criminals by their voice–combining automatic and traditional methods for optimum performance. In: Forensic Speaker Identification, 2006.
  64. G.S. Morrison et al. Making Demonstrably Valid and Reliable Forensic Voice Comparison a Practical Everyday Reality in Australia-Database Collection Protocol. In: Invited presentation at 13th Australasian International Conference on Speech Science and Technology, 14-16, 2010.
  65. H. Melin. Databases for speaker recognition: Activities in COST250 working group 2. COST 250-Speaker Recognition in Telephony, Final Report 1999, 1999.
  66. C.Y. Wang, P. Rose. Likelihood Ratio-Based Forensic Voice Comparison with Cantonese/i/F-Pattern and Tonal F0. In: Proceedings of the 14th Australasian International Conference on Speech Science and Technology (SST 2012), 209-212, 2012.
  67. G.S. Morrison, P. Rose, C. Zhang.Protocol for the collection of databases of recordings for forensic-voice-comparison research and practice. Australian Journal of Forensic Science 44(2), 155-167, 2012.
  68. Mapa do Encarceramento. Os jovens do Brasil. Secretaria Geral da Presidência da República, Brasília, 2014.
  69. R.Soares. A maioridade penal no Brasil e em outros países. Consultoria Legislativa. Fevereiro, 2006.
  70. D.R.S. de Lorenziet al. Fatores associados à qualidade de vida após menopausa. Rev. Assoc. Med. Bras. 52(5), 312-7, 2006.
  71. A.O. Pedro et al. Age at natural menopause among Brazilian women: results from a population-based survey. Cad. Saúde Pública 19(1), 17-25, 2003.
  72. A.C. Bonaccorsi. Andropausa: insuficiência androgênica parcial do homem idoso. Uma revisão. Arquivos Brasileiros de Endocrinologia & Metabologia 45(2), 123-133, 2001.
  73. P. Rose. Long-and short-term within-speaker differences in the formants of Australian “hello”. Journal of the International Phonetic Association 29(1), 1-31, 1999.
  74. F. McGhee. An experimental study of voice recognition. The Journal of General Psychology 31(1), 53-65, 1944.
  75. F. McGhee. The reliability of the identification of the human voice. The Journal of General Psychology 17(2), 249-271, 1937.
  76. C.B.G. Fouquet. A influência no dialeto nordestino frente ao dialeto paulista. Tese de Doutorado. Universidade de São Paulo, 2013.
  77. L. Oushiro, R.B. Mendes. A pronúncia do (-r) em coda silábica no português paulistano. Revista do GEL, 8(2), 66-95, 2013.
  78. C.M.B. Leite. O/R/EM posição de coda silábica na capital do interior paulista: uma abordagem sociolinguística. Sínteses - ISSN 1981-1314, 15, 2012.
  79. A.A.M.D. Silveira, L.E. Tenani. Elevação vocálica no dileto do interior paulista: contribuições para os estudos de variação fonológica do português do Brasil. Revista de Estudos Linguísticos 36, 1, 2007.
  80. L.M.C. Imaguire. Estudo geolingüístico de alguns municípios do litoral sul paulista: abordagem de aspectos semântico-lexicais. Tese de Doutorado. Faculdade de Filosofia, Letras e Ciências Humanas da Universidade de São Paulo, São Paulo, 2004.
  81. G.S. Morrison, F. Ochoa, T. Thiruvaran. Database selection for forensic voice comparison. In: Proceedings of Odyssey, 62-77, 2012.
  82. A.H. Anderson et al. The HCRC map task corpus. Language and speech 34(4), 351-366, 1991.
  83. J. Elliott. Auditory and F-pattern variations in Australian okay: a forensic investigation. Acoustics Australia 29(1), 37-41, 2001.
Licence Creative Commons

Cette œuvre est sous licence Creative Commons Attribution - Pas d'Utilisation Commerciale - Partage dans les Mêmes Conditions 4.0 International.

(c) Tous droits réservés Revista Brasileira de Criminalística 2016

Partager

Télécharger

PDF (Português (Brasil))

Auteurs)

  • Gerson Albuquerque Silva,

    • Gerson Albuquerque Silva

    Universidade de São Paulo

    Foneticista Forense Instituto de Criminalística de São Paulo