Database, protocols and tools for evaluating score-level fusion algorithms in biometric authentication

Abstract

Fusing the scores of several biometric systems is a very promising approach to improve the overall system's accuracy. Despite many works in the literature, it is surprising that there is no coordinated effort in making a benchmark database available. It should be noted that fusion in this context consists not only of multimodal fusion, but also intramodal fusion, i.e., fusing systems using the same biometric modality but different features, or same features but using different classifiers. Building baseline systems from scratch often prevents researchers from putting more efforts in understanding the fusion problem. This paper describes a database of scores taken from experiments carried out on the XM2VTS face and speaker verification database. It then proposes several fusion protocols and provides some state-of-the-art tools to evaluate the fusion performance.

A Novel Approach to Combining Client-Dependent and Confidence Information in Multimodal Biometrics

“…The publicly available 1 XM2VTS benchmark database for score-level fusion [11] is used. There are altogether 32 fusion data sets and each data set contains a fusion task of two experts.…”

“…This is done by calculating the global false acceptance and false rejection errors over the 32 experiments for each of the α values. The pooled EPC curve and its implementation can be found in [11].…”

“…These baseline experiments are then compared to fusion using all the available information sources. Based on 32 fusion data sets taken from the publicly available XM2VTS score fusion benchmark database [11], it is concluded that fusion with both sources of information is more beneficial than using either one or none of them.…”

Biosec baseline corpus: A multimodal biometric database

“…Some of them consist of only matching scores produced by several biometric systems operating on different modalities [2]. While these databases encourage research on multimodal fusion, they prevent research on individual systems and even on fusion at other levels than score fusion.…”

“…In this communication we focus on the more general and common case of multimodal databases consisting of biometric signals. In this respect, prominent examples are: XM2VTS [2], including face and voice; MCYT [3], including fingerprint and handwritten signature; and BIOMET [4], which contains samples of face, voice, fingerprint, hand and handwritten signature. These previously existing databases had several limitations that the BioSec baseline corpus tries to overcome.…”