Maximum Likelihood Parameter Estimate Decision Level Fusion Based Face Biometrics Authentication Systems

Abstract

Face recognition has long been a goal of computer vision, but only in recent years reliable automated face recognition has become a realistic target of biometrics research. In this paper the contribution of classifier analysis to the Face Biometrics Verification performance is examined. It refers to the paradigm that in classification tasks, the use of multiple observations and their judicious fusion at the data, hence the decision fusions at different levels improve the correct decision performance. The fusion tasks reported in this work were carried through fusion of two well-known face recognizers, ICA I and ICA II. It incorporates the decision at matching score level, a novel fusion strategy is employed; the Likelihood Ratio Fusion within scores. This strategy
increases the accuracy of the face recognition system and at the same time reduces the limitations of individual recognizer. The performances of the analysis studies were tested based Indian face database (IFD2002) and the simulation results are showed significant performance achievements.