Security issues in face recognition

Data uncertainty in face recognition due to various environmental factors and spoofing face recognition with 3D- masks are two major threats to gain illegitimate access. A single image of face has got high uncertainty in representing the face since the face image varies with illumination, facial expression and pose. Thus a face image can only be considered as an observation but not an absolute accurate representation of the face. The accuracy of the face recognition can be improved by obtaining more face images from the same person. However in real-world we only have a limited number of available face images and thus there is high uncertainty. In this paper, we attempt to reduce the uncertainty of face image by synthesizing the virtual training samples. Similarly, spoofing is the act of gaining an illegitimate access by masquerading as a valid user by falsifying data. Among all biometric traits, face can be exposed to the most serious threats as it is easy to access and reproduce. In this paper various spoofing techniques have been have been examined and different algorithms have been put forward to detect them. We try to examine the spoofing potential of 3D facial masks for different recognition systems and address the detection problem for this complex attack.