Applied cryptography (2nd ed.): protocols, algorithms, and source code in C
Applied cryptography (2nd ed.): protocols, algorithms, and source code in C
Eigenfaces vs. Fisherfaces: Recognition Using Class Specific Linear Projection
IEEE Transactions on Pattern Analysis and Machine Intelligence
CCS '99 Proceedings of the 6th ACM conference on Computer and communications security
An Analysis of Minutiae Matching Strength
AVBPA '01 Proceedings of the Third International Conference on Audio- and Video-Based Biometric Person Authentication
Combining Crypto with Biometrics Effectively
IEEE Transactions on Computers
Discriminant analysis via support vectors
Neurocomputing
Secure sketch for biometric templates
ASIACRYPT'06 Proceedings of the 12th international conference on Theory and Application of Cryptology and Information Security
Hi-index | 0.00 |
The use of biometric systems is becoming an important solution to replace traditional authentication. However, biometric systems are vulnerable to attacks. When biometric data is compromised, unlike a password, it can't be changed. Therefore, the security of biometrics models is essential in designing an authentication system. To achieve this protection of biometric models, two categories of approaches are proposed in the literature, namely, methods based on transformation of characteristics and biometric cryptosystems. For the first type of approaches, a study is made to assess the security of biometric systems. In biometric cryptosystems the realized works are hampered by the lack of formal security analysis. Hence the purpose of this paper is to propose standard criteria for a formal security analysis of biometric cryptosystems. The proposed measures take into account the specific effect of key binding cryptosystems. The security analysis is illustrated by experiments on the techniques of Fuzzy Commitment and Fuzzy Vault which we use in this work for the protection of biometric face recognition system. Our analysis indicates that both techniques are vulnerable to intrusion and binding attacks because of the ease of obtaining the user's model using the elements known to the attacker.