Biometrics - Biometrics Fusion

In highly secure systems, reinforcement of evidence from multiple independent biometric identifiers offers increasingly irrefutable proof of the identity of the authorized person. Multi-modal (e.g., multiple biometrics) systems provably deliver better performance than any single constituent biometrics

Multiple biometrics can also alleviate several practical problems in biometrics-based personal identification. First, although a biometric identifier is supposed to be universal (each person in the target population should possess it), in practice, no biometric identifier is truly universal. For instance, some people are missing fingers due to accidents. Second, the relevant biometric identifiers can not always be sensed or measured by a practical biometric identification system. For instance, a small fraction of the population possess fingerprints which are very shallow, extremely dry, or roughened from work. Consequently, the authentication system can not handle this fraction of population based on that particular biometric identifier. Third, some biometrics may not be accepted by certain segments of the population. For instance, retinal identification is fairly intimidating to the casual user.

We have integrated fingerprint and voice features in a novel way to improve the performance of Bio-Barrier - a biometrics based door access control system currently installed in our lab (Hawthorne 1S-F54).

Selected publications:

Can Multi-biometrics Improve Performance
L. Hong, A. K. Jain, and S. Pankanti
Proc. 1999 IEEE Workshop on Automatic Identification Advanced Technologies (WAIAT-99), Morristown NJ, pp. 59-64, October 1999.

While it is widely acknowledged that the performance improvement in current biometrics�based personal authentication systems is necessary, it is not clear what mechanisms could be used to improve the performance. In this paper, we formulate the problem of multiple biometrics integration and examine whether the improvement in performance could be achieved from integrating multiple biometrics. For two practical and commonly used situations of multi-biometric integration, we analyze the performance gains. We also demonstrate empirically that integration of multiple biometrics does indeed result in a consistent and significant performance improvement.