Presentation Attack Detection: A Systematic Literature Review

Abstract

Identity authentication is the process of verifying one’s identity. Many identity authentication methods have been developed, from the conventional username-password systems to the recent electroencephalography-based authentication. Among them, biometric authentication shows particular importance due to its convenience and wide application in real-world scenarios. Face recognition is one of the most widely used biometric authentication methods, but simultaneously it receives various attacks. To overcome attacks, face presentation attack detection has been intensively studied in the last two decades regarding diverse domains of datasets, evaluation methods, and attack types. In this systematic literature review, we identify and categorise the state-of-the-art approaches in each domain to cover the challenges and solutions in a single place. We provide comparisons of representative methods on widely used datasets, discuss their pros and cons, and hope our insights can inspire future works.