Comparison of Biometric Authentication Software Techniques: GEFE vs. Angle Based Metrics

Science Center Auditorium, University of Dayton

In this paper, we explore three alternatives for developing a biometric authentication software system. The first approach we will consider is a computer vision technique optimized by Genetic and Evolutionary Feature Extraction (GEFE); the second is Angle Based Metrics (ABM); and the third is Angle Based Metrics combined with Genetic and Evolutionary Computation (ABM + GEC). Each of these techniques are research areas which show promise in regards to being able to authenticate users based on their natural mouse movements. When applied to the same data set, the results of our experimentation indicate that both the ABM and ABM + GEC techniques are more accurate than GEFE in correctly verifying genuine users, as well as correctly rejecting impostors.