Social Network Analysis for Biometric Template Protection

Abstract

In this book chapter, novel cancelable biometric template generation algorithm using Social Network Analysis is presented. Two sets of features are fused using Social Network. Fusion technique is cancelable. Eigenvector centrality is used to generate final sets of features from the Virtual Social Network (VSN). From the features, Virtual Social Network generation and analysis is one of the important tasks. Main difficulty of this process is to find the validity of social network feature. Two sets of feature keep the relation among them for the classification. Finding the relation among the features could be done by using social network analysis. A new algorithm to generate VSN from the features is also presented in this chapter. Generated virtual network keeps the discriminability among the features. Once the features are different, the generated networks are different from the previous one. Social network based feature analysis confirms the domain transformation. It is computationally very hard to regenerate the original features from the social network matric value. This domain transformation confirms the cancelability in biometric template generation. To ensure the multi-level cancelability random projection is used to project social network based feature. Performance analysis for biometric domain data is also presented in this chapter.