Adapted user-dependent multimodal biometric authentication exploiting general information

Abstract

This is the author’s version of a work that was accepted for publication in Pattern Recognition Letters. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in Pattern Recognition Letters 26.16 (2005): 2628 – 2639, DOI: 10.1016/j.patrec.2005.06.008A novel adapted strategy for combining general and user-dependent knowledge at the decision-level in multimodal biometric authentication is presented. User- independent, user-dependent, and adapted fusion and decision schemes are com- pared by using a bimodal system based on ¯ngerprint and written signature. The adapted approach is shown to outperform the other strategies considered in this pa- per. Exploiting available information for training the fusion function is also shown to be better than using existing information for post-fusion trained decisions.This work has been supported by the Spanish Ministry for Science and Tech- nology under projects TIC2003-09068-C02-01 and TIC2003-08382-C05-01

Combining different biometric traits with one-class classification

“…We can also observe that the weighted sum is the best combination rule for both databases. Some justification for this behavior can be found in the work of Jain and Ross [9] and of Fierrez et al [4]. They argue, and show through experimentation, that the performance of biometric systems can be further improved by learning user-specific parameters.…”

“…The second database used in this work is the MCYT scores database [17,4], which contains a dataset of bimodal matching scores (fingerprint and signature). The database was built with the fingerprint (N and Q) and signature (S) data of 75 subjects from the MCYT database, together with scalar fingerprint quality measures labeled by a human expert.…”

“…The MCYT database provides only normalized scores [4]. The fingerprint scores were normalized according to the following equation:…”

“…The limitation of this strategy is that very often the amount of data available for training is not sufficient and not representative enough to guarantee good parameter estimation and generalization capabilities [4]. In some cases, only the patterns from one class (genuine or impostor) are available or the data are seriously unbalanced.…”

Fusion of biometric systems using one-class classification

“…The drawback of this strategy is that very often the amount of data available for training is not sufficient and not very representative to guarantee good parameter estimation and generalization capabilities [3]. In some cases just the patterns coming from one class (genuine or impostor) are available or the data is hugely unbalanced.…”