Lessons learned from the application of biometric-tests on a real case involving identity verification of a missing child

Abstract

The topic of identity verification and age progression received considerable attention among researchers in the computer vision research community leading to the development of a variety of promising techniques that have been developed and tested using dedicated datasets. However, a key question that arises is whether biometric-based techniques could be used in real identity verification of age separated faces with the same effectiveness as other identity verification tests or at least as a soft filtering approach before applying more reliable but intrusive tests such as DNA tests. Towards this end we apply biometric-based techniques in an attempt to verify the identity of a missing child, using real data. The aim of the paper is not to provide answers to the aforementioned question but to report the problems faced in similar scenarios and propose future research directions that aim to make biometric-based techniques more applicable to real-life scenarios.