Adversarial Attacks on Face Recognition

Face recognition is becoming a prevailing authentication solution in numerous biometric applications thanks to the rapid development of deep neural networks (DNNs) [18, 37, 39]. Empowered by the excellent performance of DNNs, face recognition models are widely deployed in various safety-critical scenarios ranging from finance/payment to automated surveillance systems. Despite its booming development, recent research in adversarial machine learning has revealed that face recognition models based on DNNs are highly vulnerable to adversarial examples [14, 40], which are maliciously generated to mislead a target model.