Face Image Analysis in Mobile Biometric Accessibility Evaluations
Barbara Corsetti, Raúl Sánchez-Reillo, Richard Guest, Marco Santopietro · 2019
Smartphones cameras are widely used for biometric authentication purposes. This enables more and more users experience face recognition in different common scenarios (e.g., unlocking phones, banking, access controls). One of its advantages is that face recognition requires low interaction with the systems (by simply looking at the smartphone's screen). Thus, it may be useful for people affected by mobility concerns. For this reason, researchers recently started to conduct mobile biometric evaluations recruiting accessibility populations. The aim is to analyse all those factors that, depending on the users' capabilities, influence the biometrics recognition process. In this paper we focus our attention on sample quality, analysing the face images collected during a mobile biometric accessibility study. Results obtained enable us to understand how the users' accessibility concerns influence the biometric sample quality and discuss possible solutions for eradicating this inconvenience. This assessment had been conducted following the recommendations of the ISO/IEC TR 29794-5.