15th Annual Conference of the International Speech Communication Association

September 14-18, 2014

Bayesian Calibration for Forensic Evidence Reporting

Niko Brümmer, Albert Swart

AGNITiO, South Africa

We introduce a Bayesian solution for the problem in forensic speaker recognition, where there may be very little background material for estimating score calibration parameters. We work within the Bayesian paradigm of evidence reporting and develop a principled probabilistic treatment of the problem, which results in a Bayesian likelihood-ratio as the vehicle for reporting weight of evidence. We show in contrast, that reporting a likelihood-ratio distribution does not solve this problem. Our solution is experimentally exercised on a simulated forensic scenario, using NIST SRE'12 scores, which demonstrates a clear advantage for the proposed method compared to the traditional plugin calibration recipe.

Full Paper

Bibliographic reference.  Brümmer, Niko / Swart, Albert (2014): "Bayesian calibration for forensic evidence reporting", In INTERSPEECH-2014, 388-392.