Evaluation of biometric recognition in the COVID-19 period

Abstract

The new coronavirus is causing a global health crisis. Many biometric authentication methods have been affected in this epidemic. In this study, we first give a comprehensive evaluation of the COVID-19's impact on various biometric authentication methods and provide an alternative solution based on periocular recognition. The experiments on the Ethnic-ocular dataset show that the Siamese structure could provide a better performance for periocular recognition