On the Use of Automatic Speaker Verification Systems in Forensic Casework
Johan Koolwaaij, Lou Boves · 1999
In forensic applications of speaker recognition it is necessary to be able to specify a confidence level for a decision that two sets of recordings have been produced by the same speaker (or by different speakers). Forensic phoneticians are sometimes incriminated because they find it impossible to provide 'hard' estimates of the confidence level of an expert opinion. In this paper it is investigated to what extent the problem can be solved by deploying automatic speaker verification algorithms, to work alone or to support the work of forensic phoneticians. It is shown that for several reasons hard estimates of the confidence of an opinion in a specific case cannot be provided by automatic speaker verification either. 1. Introduction Automatic Speaker Verification (SV) and Forensic Casework have long been considered as essentially unrelated disciplines, because the former was seen as a one alternative forced choice problem, whereas the latter used to be presented as a an open set iden...