Construction of a Bayesian decision theory‐based secure multimodal fusion framework for soft biometric traits

Abstract

In a soft biometric-based model, multiple soft biometric characteristics are fused with one or more primary biometric traits in a multimodal environment. In this study, the authors have reviewed a Bayesian decision theory-based fusion technique and considerably improved its performance by first identifying some of its limitations and subsequently modifying it accordingly. Specifically speaking, they have utilised the notion of Gaussian functions and a novel dynamic soft biometric weight assignment (DSWA) scheme for achieving these objectives. They have tested the modified framework on real-life data, which resulted in improved performances over the basic fusion model. They have also attempted to address here some security and privacy concerns associated with such frameworks. Although the soft biometric characteristics possess much lower uniqueness in comparison to a primary trait, they can be exploited by an active adversary to mine sensitive information about any individual. As such, the authors have proposed a secure fusion technique which performs one-way transformations of the soft biometric characteristics. They have also tested this secure design on real-life data and found the results satisfying.

A novel hybrid score level and decision level fusion scheme for cancelable multi-biometric verification

“…To validate the performance of our method, we compare our proposed hybrid fusion scheme with other recent methodologies in literature. Besides hybrid fusion methods [12,37], we include few other recent state-of-the-art fusion approaches based on score level [25,18,15,22,9] and decision level fusion [31,17]. As described in performance evaluation, it can be observed that the proposed method performs optimally than the other approaches with respect to EER (see Figure 4 -6).…”

“…In case of Virtual A database, the technique proposed in [12] performs better than the proposed method, but it involves complex evaluation for global error optimization using PSO. The decision fusion methods involve AND rule and OR rule-based fusion proposed by Kelkboom et al [17] and Bayesian classifier fusion proposed by Sadhya et al [31]. Table 3 reports the EER and GMR @ 0.01% FMR, obtained using the proposed and existing weighting techniques.…”

“…Further, the appropriate weights for these four rules are evaluated for score fusion. Sadhya et al [31] considered four soft biometric traits for combination i.e. height, weight, age, gender.…”

“…The limitation lies in the complex implementation and it may also suffer from the local minimum problem. Sadhya et al [31] utilized Bayesian decision theory to combine soft biometric traits. In this method [31], the performance degrades if matcher's characteristics or distributions are not known.…”

“…Sadhya et al [31] utilized Bayesian decision theory to combine soft biometric traits. In this method [31], the performance degrades if matcher's characteristics or distributions are not known. Nguyen et al [25] proposed DS theory based score fusion for iris and fingerprint.…”