Fusing multimodal biometrics with quality estimates via a Bayesian belief network. (English) Zbl 1132.68649

Summary: Biometric systems for today’s high security applications must meet stringent performance requirements; fusing multiple biometrics can help lower system error rates. Fusion methods include processing biometric modalities sequentially until an acceptable match is obtained, using logical (AND/OR) operations, or summing similarity scores. More sophisticated methods combine scores from separate classifiers for each modality. This paper develops a novel fusion architecture based on Bayesian belief networks. Although Bayesian update methods have been used before, our approach more fully exploits the graphical structure of Bayes nets to define and explicitly model statistical dependencies between relevant variables: per sample measurements, such as match scores and corresponding quality estimates, and global decision variables. These statistical dependencies are in the form of conditional distributions which we model as Gaussian, gamma, log-normal or beta, each of which is determined by its mean and variance, thus significantly reducing training data requirements. Moreover, by conditioning decision variables on quality as well as match score, we can extract information from lower quality measurements rather than rejecting them out of hand. Another feature of our method is a global quality measure designed to be used as a confidence estimate supporting decision making. Preliminary studies using the architecture to fuse fingerprints and voice are reported.


68T10 Pattern recognition, speech recognition


Full Text: DOI


[1] Jain, A. K.; Prabhakar, S.; Chen, S., Combining multiple matchers for a high security fingerprint verification system, Pattern Recognition Lett., 20, 1371-1379 (1999)
[2] Ross, A.; Jain, A. K., Information fusion in biometrics, Pattern Recognition Lett., 24, 2115-2125 (2003)
[3] Kittler, J., On combining classifiers, IEEE Trans. Pattern Anal. Mach. Intell., 20, 3, 226-239 (1998)
[4] Bigun, J., Multimodal biometric authentication using quality signals in mobile communications, (Proceedings of IAPR International Conference on Image Analysis and Processing (ICIAP) (2003), IEEE CS Press), 2-13
[5] Rukhin, A. L.; Malioutov, I., Fusion of biometric algorithms in the recognition problem, Pattern Recognition Lett., 26, 679-684 (2005)
[6] Verlinde, P.; Chollet, G.; Acheroy, M., Multimodal identity verification using expert fusion, Inf. Fusion, 1, 1, 17-33 (2000)
[8] Fierrez-Aguilar, J., Kernel-based multimodal biometric verification using quality signals, (Jain, A. K.; Ratha, N. K., Biometric Technology for Human Identification, Proceedings of the SPIE, vol. 5404 (2004)), 544-554
[12] Wein, L. M.; Baveja, M., Using fingerprint image quality to improve the identification performance of the U.S. Visitor and Immigrant Status Indicator Technology Program, Proc. Natl. Acad. Sci., 102, 21, 7772-7775 (2005)
[14] Fierrez-Aguilar, J.; Ortega-Garcia, J.; Gonzales-Rodriguez, J., Discriminative multimodal biometric authentication based on quality measures, Pattern Recognition, 38, 777-779 (2005)
[15] Baker, J. P.; Maurer, D. E., Fusion of biometric data with quality estimates via a Bayesian belief network, (Proceedings of the Biometric Symposium (2005), Arlington: Arlington VA), 21-22
[17] Teoh, A. B.J.; Samad, S. A.; Hussain, A., A face and speech biometric verification system using a simple Bayesian structure, J. Inf. Sci. Eng., 21, 1121-1137 (2005)
[19] Jensen, F. V., Bayesian Networks and Decision Graphs (2001), Springer: Springer New York · Zbl 0973.62005
[20] Prabhakar, S.; Jain, A. K., Decision-level fusion in fingerprint verification, Pattern Recognition, 35, 861-874 (2002) · Zbl 0999.68579
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.