Anti-spoofing in Action: Joint Operation with a Verification System

Abstract

Besides the recognition task, today's biometric systems need to cope with additional problem: spoofing attacks. Up to date, academic research considers spoofing as a binary classification problem: systems are trained to discriminate between real accesses and attacks. However, spoofing counter-measures are not designated to operate stand-alone, but as a part of a recognition system they will protect. In this paper, we study techniques for decisionlevel and score-level fusion to integrate a recognition and anti-spoofing systems, using an open-source framework that handles the ternary classification problem (clients, impostors and attacks) transparently. By doing so, we are able to report the impact of different spoofing counter-measures, fusion techniques and thresholding on the overall performance of the final recognition system. For a specific usecase covering face verification, experiments show to what extent simple fusion improves the trustworthiness of the system when exposed to spoofing attacks.

Enhanced on-line signature verification based on skilled forgery detection using Sigma-LogNormal Features

“…Also the fusion of the knowledge provided by the skilled detector and the verification score will be improved, following the examples of previous works on anti-spoofing and verification systems fusion in [4,14], where different sequential-and classifier-based approaches, as well as decision-and scorelevel fusion of anti-spoofing and verification systems, for face and fingerprint, are explored. Other signature matchers and databases will be taken into account as well.…”

Joint Speaker Verification and Antispoofing in the <inline-formula> <tex-math notation="LaTeX">$i$ </tex-math></inline-formula>-Vector Space

“…False acceptance is the opposite case when an impostor is misclassified as a target. The evaluation of the ASV system is done in terms of both LICIT and SPOOF protocols [29]. The LICIT protocol, involving zero-effort impostors, is the typical evaluation protocol used in verification 2 http://sp-tk.sourceforge.net/ scenarios, whereas the SPOOF protocol contains only spoofed impostors and is used to evaluate system performance when spoofing attacks are present.…”

“…We use zero-effort FAR (ZFAR) as a metric for the LICIT protocol and spoofing FAR (SFAR) [29] for the SPOOF protocol. The only difference between them is that the former considers zero-effort impostors while the latter considers their spoofed versions.…”

Robust multimodal face and fingerprint fusion in the presence of spoofing attacks

“…Consequently, there are numerous studies [1,2] and open challenges [3,4] on anti-spoofing techniques assessing the spoofing detector's ability to distinguish between genuine and fake attempts for especially these two traits. Recently, the integration of anti-spoofing scores with recognition scores has received considerable attention [5][6][7]. The standard approach, as outlined in [5], has been to reject spoofed samples before comparing them against the gallery template.…”