Fingerprint Recognition With Embedded Presentation Attacks Detection: Are We Ready?

Abstract

on the probabilistic modeling of the relationships among the Receiver Operating Characteristics (ROC) of the two individual systems when PAD and verification stages are implemented sequentially. As a matter of fact, this is the most straightforward, flexible, and widespread approach. We carry out simulations on the PAD algorithms' ROCs submitted to the most recent editions of LivDet (2017-2019), the state-of-the-art NIST Bozorth3, and the top-level Veryfinger 12 matchers. Reported experiments explore significant scenarios to get the conditions under which fingerprint matching with embedded PAD can improve, rather than degrade, the overall personal verification performance.

Balancing Accuracy and Error Rates in Fingerprint Verification Systems Under Presentation Attacks With Sequential Fusion

“…However, the current version of Bio-WISE and, in general, the findings of Ref. [12] and related previous works [6]- [10], [13], [14] do not allow assessing for which PAD's operational points the overall GAR degradation can be still acceptable, with the advantage of handling presentation attacks. The key question is: what is the best way to embed a PAD into a recognition system so that the final product is robust to spoofing (IAPAR low enough), without suffering from significantly reduced gen-uine recognition accuracy (GAR still acceptable)?…”

“…Up to date, it is difficult to draw the effect on several possible scenarios; in other words, having a clear idea of the conditions for which this embedding may lead to a real gain, that is, an advantage with respect to neglect the presentation attack problem. To fill this gap, we presented in [12] a novel simulation approach based on the probability modeling of receiver operating characteristic (ROC) curves of PADs and verification systems. The goal was to simulate the performance of a sequentially integrated system.…”

“…Ref. [12] exemplified the performance in two possible operational points of state-of-the-art fingerprint PADs and analyzed the performance of a benchmark and a toplevel comparator when embedding such PADs. We noticed that the PAD significantly impacts the performance of the integrated system, regardless of the comparison algorithm used.…”

“…Section 2 makes the point about the current research on fingerprint comparison and PAD fusion, with specific reference to other biometrics where this problem was put on the table. Section 3 firstly summarizes the basic model fully described in [12]. Then, it introduces the concept of "trade-off", which is well known in many engineering fields, and in particular how it has been used so far in pattern recognition applications.…”

Introduction to Presentation Attack Detection in Fingerprint Biometrics

“…• assessing the robustness of anti-spoofing methods against novel presentation attacks in terms of fabrication strategy, adopted materials, sensor technology; for instance, in [27] it has been shown that the PAD error rates of software-based approaches can show a three-fold increase when tested on PA materials not seen during training; • designing effective methods to embed PAD in fingerprint verification systems [68], including the need for computationally efficient PAD techniques, to be used on low-resources systems such as embedded devices and low-cost smartphones; • improving explainability of PAD systems; the use of CNNs is providing great benefits to fingerprint PAD performance, but such solutions are usually considered as "black boxes" shedding little light on how and why they actually work.…”