Adaptation to changes in multimodal biometric authentication

Multimodal biometric authentication is gaining more and more attention recently. In [l], a Reduced Multivariate Polynomial model has heen used to combine the hiometric classifiers with good performance. Although thc model parameter can he solved very quickly, it may not be suitahk in caw when ncw registration is a frequent task. Moreover, due to wear and tcar of sensors, the distribution of match scorcs can change with time and hence the performance of a fixed classifier may be affected. In this paper, we introduce an adaptive algorithm to track these changes. The new algorithm is demonstrated to address both the registration and match scores distribution changing problems using two biometrics.