We are working with a team at the “Universidad Complutense de Madrid” to develop our own Face Recognition System. Initially it will be a 1:1 verifier, with provision to eventually become a 1:N verifier.

The team comprise of 4 very skilled PhD qualified university staff who specialize in image processing, highly parallelized processing, and optimizing processes for GPUs (Graphics Processing Units).

They guarantee that the resulting verification engine will perform as well as the top end systems in the Face Recognition Grand Challenge (FRGC) contest:

www.frvt.org

The winning algorithms of this contest are one from Viisage, NJIT (New Jersey Institute of Technology), and CMU (Carnegie Mellon University). The algorithm from Viisage is unpublished and protected by a patent. The other two are public, and these are the two that will be researched as the basis of our system.

To date they have developed the infrastructure to access the FRGC database and read all of the meta-information (eye locations). They are currently focusing on testing the proposed algorithms and making sure that they work as expected before launching into the definitive development work.