On the security of face recognition terminals at modern airports

Abstract

The inability of airports to absorb the increasing numbers of passengers year after year, makes the use of e-gates and face recognition terminals a promising solution to enhance the quality of service and security. Thus, over the past few years, many airports around the world have adopted facial recognition technology to facilitate verification and identification of passengers, for example for surveillance systems, and recently for face recognition terminals. We believe that there are still many challenges and questions about face recognition in the context of these huge security applications, concerning performance, security and privacy. We believe that these concerns must be deeply discussed. In this paper, we discuss the potential vulnerabilities of facial recognition, then we present and discuss the result of an adopted attack against security applications that compare the live face of a passenger with his/her stored face image in passport or visa. Our experimental results illustrate the ability of the presented trick attack to be a serious threat for such modern security applications.