On the Use of Log-Likelihood Ratio Based Model-Specific Score Normalisation in Biometric Authentication

It has been shown that the authentication performance of a biometric system is dependent on the models/templates specific to a user. As a result, some users may be more easily recognised or impersonated than others. The authors propose model-specific (or user-specific) likelihood based score normalisation procedure that can reduce this dependency. While in its original form, such an approach is not feasible due to the paucity of data, especially of the genuine users, they stabilise the estimates of local model parameters with help of the user-independent (hence global) parameters.