Face Recognition Issues in a Border Control Environment

Face recognition has greatly matured since the earliest forms, but still improvements must be made before it can be applied in high security or large scale applications. We conducted an experiment in order to estimate percentage of Norwegian people having one or more look-alikes in Norwegian population. The results indicate that the face recognition technology may not be adequate for identity verification in large scale applications. To survey the additional value of a human supervisor, we conducted an experiment where we investigated whether a human guard would detect false acceptances made by a computerized system, and the role of hair in human recognition of faces. The study showed that, on average, the human guard was able to detect almost 80% of the errors made by the computerized system. More over, the study showed that the ability of human guard to recognize a human face is a function of hair: false acceptance rate was significantly higher for the images where the hair was removed compared to where it was present.