Biometrics: limitations and future research

Location: CPD room (M413) , Merz Court
Date/Time: 5th May 2011, 15:00

Biometrics has been an active research subject for many years. In this talk, I'll first explain my involvement in this area in the two published papers by IEEE Transactions journals. The first paper addresses the problem that how to effectively integrate biometrics into cryptographic applications despite the fuzziness of data (Hao,Anderson,Daugmna'06). The second paper address the search problem that how to efficiently store and retrieve biometric data (Hao,Daugman,Zielinski'08). I'll share some of my thoughts about the future of biometric research. It seems to me that this research area has reached a bottleneck. Inventing new types of biometrics is highly unlikely; improving existing biometric systems to a substantial extent seems very difficult as well. However, in this world, not only humans have biometrics. In fact, an physical object has "biometrics" that naturally occur, are unique and can't be cloned. If we keep an open-mind, we might find there are plenty of new research avenues and challenges.

published on: 5th April 2011