MSc Project Plan Face Recognition Software: Non-zero effort attacks ’ effect on False Acceptance Rate

By the end of the year 2006 the Visa Information System (VIS) is to be implemented, and the members of Schengen can start connect-ing their national VIS systems towards VIS. VIS is the new Europe-wide visa system operating together with national VIS-systems that includes biometric authentication. Biometrics will be used to imple-ment measures to strengthen the security by making it harder for an adversary to perform identity theft. The intention with the system is to prevent identity fraud and ensure the implementation of the Eu-ropean Union’s asylum politics, by exchanging information about the citizens applying for visa [1], [2]. The VIS will consist of a several national VIS’s (operational from 2005). When choosing the products for face recognition in a border control environment, such as VIS, it is important that the data on which the evaluation is based is established by using environment similar to that in a border control environment. One of the criteria applied when deciding which product to use will be its False Acceptance Rate (FAR). Literature confirms that biomet-rics like fingerprints can be easily spoofed using methods like silicon dummy-fingers [3], and it is also possible to spoof face recognition products. This thesis will examine what effect non-zero effort impos-tors will have on different face recognition products FAR when used in a environment similar to that of a border control environment