Evaluation of face recognition technologies for identity verification in an eGate based on operational data of an airport

Face recognition technologies play an important part in the field of automated border control (ABC). The demand for the reliable verification of a passenger´s identity is based on efficiency and security aspects of ABC systems which have to be taken into account. Face recognition technologies allow for an automation of the traditional manual inspection by a border guard. This paper presents an evaluation of the face verification performance of commercial face recognition technologies based on operational data from an existing eGate at an airport. In addition to comparisons of the overall performance, more detailed considerations such as the influence of the remaining validity of the passport are made. Moreover, the performance implications of the nationality of the passports, which may be the result of the different quality of the passport photos or increased cooperativeness of the passengers based on previous experiences with eGates, are considered as well.