Fusion of Multiple Biometric For Photo-Attack Detection in Face Recognition Systems

Abstract

A spoofing attack is a situation in which one person successfully masquerades as another by falsifying data and gaining illegitimate access. Spoofing attacks are of several types such as photograph, video or mask. Biometrics are playing the role of a password which cannot be replaced if stolen, so there is the necessity of counter-measures to biometric spoofing attacks. Face biometric systems are vulnerable to spoofing attack. Regardless of the biometric mode, the typical approach of anti-spoofing systems is to classify the biometric evidence which are based on features discriminating between real accesses and spoofing attacks. A number of biometric characteristics are in use in various applications. This system will be based on face recognition and lip movement recognition systems. This system will make use of client-specific information to build client-specific anti-spoofing solution, depending on a generative model. In this system, we will implement the client identity to detect spoofing attack. With this, it increases efficiency of authentication. The image will be captured and registered with its client identity. When user has to be authenticated, the image will be captured with his identity manually entered. Now system will check the image with respect to client identity only. Lip movement recognition will be done at time of authentication to identify whether client is spoof or not. If client is authenticated, then it will check for captured image dimension using Gaussian Mixture Model (GMM). This system also encrypts and decrypts a file by extracting parameter values of a registered face.