How we use automatic speaker comparison in forensic practice
DOI:
https://doi.org/10.1558/ijsll.23955Keywords:
automatic speaker recognition, automatic speaker comparison, forensic speaker comparison, forensic speaker recognition, court communication, forensic casework, explanatory noteAbstract
Automatic methods are rapidly gaining ground in forensic speaker comparison, next to the existing auditory-acoustic methodology, performed by human experts with an academic background in phonetics. In this article we set out the steps that were taken before we could introduce the automatic method and start combining the two methods (software and human) in casework. We further provide a comprehensive explanation of the automatic method (originally written for readers of forensic reports) in the appendix. We discuss the legal reception of the combined approach, based on a court ruling in an appeal case in which the reliability of the speaker comparison was challenged by the defence. We also address the important issue of how conflicting results from the two methods may be dealt with in practice.
References
Cambier-Langeveld, T., van Rossum, M. and Vermeulen, J. (2014) Whose voice is that? Challenges in forensic phonetics. In J. Caspers, Y. Chen, W. Heeren, J. Pacilly, N. O. Schiller and E. van Zanten (eds) Above and Beyond the Segments. Experiment Linguistics and Phonetics 14–27. Amsterdam: John Benjamins Publishing Co. https://doi.org/10.1075/z.189.02cam
Gold, E. and French, P. (2019) International practices in forensic speaker comparisons: Second survey. International Journal of Speech, Language and the Law 26(1): 1–20. https://doi.org/10.1558/ijsll.38028
Kelly, F., Forth, O., Kent, S., Gerlach, L., Alexander, A. (2019) Deep neural network based forensic automatic speaker recognition in VOCALISE using x-vectors. Audio Engineering Society (AES) Forensics Conference 2019, Porto, Portugal. http://www.aes.org/e-lib/browse.cfm?elib=20477
Kelly, F. and Hansen, J. (2021) Analysis and calibration of Lombard effect and whisper for speaker recognition. IEEE/ACM Transactions on Audio, Speech, and Language Processing 29: 927–942. https://doi.org/10.1109/TASLP.2021.3053388
Klug, K., Jessen, M., Solewicz, Y. and Wagner, I. (2021) Collection and analysis of multi-condition audio recordings for forensic automatic speaker recognition. Studi AISV 8(3): 57–76. https://doi.org/10.17469/O2108AISV000003
van Leeuwen, D. and Brümmer, N. (2013) The distribution of calibrated likelihood-ratios in speaker recognition. Interspeech 2013. https://cls.ru.nl/staff/dvleeuwen/btfs-2013/vanleeuwen-btfs2013.pdf
Morrison, G. S. (2013) Distinguishing between forensic science and forensic pseudoscience: Testing of validity and reliability, and approaches to forensic voice comparison. Science and Justice 2013. https://doi.org/10.1016/j.scijus.2013.07.004
Morrison, G. S. and Enzinger, E. (2016) Multi-laboratory evaluation of forensic voice comparison systems under conditions reflecting those of a real forensic case (forensic_eval_01): Introduction. Speech Communication 85: 119–126. http://dx.doi.org/10.1016/j.specom.2016.07.006
Morrison, G. S., Enzinger, E., Hughes, V., Jessen, M., Meuwly, D., Neumann, C., Planting, S., Thompson, W. C., Van der Vloed, D., Ypma, R. J. F., Zhang, C., Anonymous, A. and Anonymous, B. (2021) Consensus on validation of forensic voice comparison. Science and Justice 61: 229–309. https://doi.org/10.1016/j.scijus.2021.02.002
Nash, J. (2019) The effect of acoustic variability on automatic speaker recognition systems. Unpublished PhD dissertation, University of York.
rechtspraak.nl (2019) ECLI:NL:GHDHA:2019:830. https://uitspraken.rechtspraak.nl/inziendocument?id=ECLI:NL:GHDHA:2019:830
Swofford, H. and Champod, C. (2021) Implementation of algorithms in pattern and impression evidence: A responsible and practical roadmap. Forensic Science International: Synergy 3: 100142. https://doi.org/10.1016/j.fsisyn.2021.100142
van der Vloed, D., Bouten, J. and van Leeuwen, D. A. (2014) NFI-FRITS: A forensic speaker recognition database and some first experiments. Proceedings of Odyssey 2014: The Speaker and Language Recognition Workshop 6–13. Joensuu, Finland: Odyssey. https://doi.org/10.21437/Odyssey.2014-2
van der Vloed, D., Kelly, F. and Alexander, A. (2020) Exploring the effects of device variability on forensic speaker comparison using VOCALISE and NFI-FRIDA, a forensically realistic database. Proceedings of Odyssey 2020: The Speaker and Language Recognition Workshop. Tokyo: Odyssey. https://doi.org/10.21437/Odyssey.2020-57
Wagner, I. (2019) Examples of casework in forensic speaker comparison, Proceedings of the 19th International Congress of Phonetic Sciences, Melbourne, Australia 2019.
