Existing statistical methods for estimating the log- likelihood ratio from biometric scores include parametric estimation, kernel density estimation, and recently adopted logistic regression estimation. There has been a growing in- terest to study the repeatability and reproducibility of these methods on biometric datasets after the 2009 National Re- search Council report and the more recent report from the 2016 Presidents Council of Advisors on Science and Tech- nology. For a statistical forensic evaluation method to be repeatable, it needs to generate consistent log-likelihood ratio scores for various sample size ratios between the genuine (mated) and imposter (non-mated) scores computed using the same database. It is a well known fact, that for logistic regression methods, the estimated intercept value depends on the sample size ratio between the two groups. There- fore, when computing log-likelihood ratios using logistic regression estimation, different genuine and impostor sample size ratios could result in different log-likelihood ratio Values. We performed extensive simulations and used different face and fingerprint biometric datasets to investigate the repeatability and reproducibility of the existing log-likelihood ratio estimation methods.
Conference Dates: October 1-4, 2017
Conference Location: Denver, CO
Conference Title: The International Joint Conference on Biometrics (IJCB 2017)
Pub Type: Conferences