t-EER: Parameter-Free Tandem Evaluation of Countermeasures and Biometric Comparators

Presentation attack (spoofing) detection (PAD) typically operates alongside biometric verification to improve reliablity in the face of spoofing attacks. Even though the two sub-systems operate in tandem to solve the single task of reliable biometric verification, they address different detection tasks and are hence typically evaluated separately. Evidence shows that this approach is suboptimal. We introduce a new metric for the joint evaluation of PAD solutions operating in situ with biometric verification. In contrast to the tandem detection cost function proposed recently, the new tandem equal error rate (t-EER) is parameter free. The combination of two classifiers nonetheless leads to a set of operating points at which false alarm and miss rates are equal and also dependent upon the prevalence of attacks. We therefore introduce the concurrent t-EER, a unique operating point which is invariable to the prevalence of attacks. Using both modality (and even application) agnostic simulated scores, as well as real scores for a voice biometrics application, we demonstrate application of the t-EER to a wide range of biometric system evaluations under attack. The proposed approach is a strong candidate metric for the tandem evaluation of PAD systems and biometric comparators.

I. Introduction

Biometric recognition is nowadays in widespread use across forensic, civilian and consumer domains. Likewise, approaches for the testing and performance reporting of biometric systems [1], [2] under normal presentation mode¹

Normal or routine implies that the system is used in the fashion intended by the system designer [3]. Spoofed trials are considered to be outside of the normal presentation mode.

are well established — see e.g. the ISO/IEC 19795 standard [4]. Despite high reliability, biometric systems are unfortunately not infallible outside of the normal presentation mode, e.g. when they are attacked by an adversary. Since the identification of biometric system attack points more than two decades ago [5], the community has been active in addressing vulnerabilities, especially presentation or spoofing attacks, for all the major biometric modes such as fingerprints [6], face [7], and voice [8], [9]. Well-studied examples of spoofing attacks include printed photographs (face), gummy fingers (fingerprints), and audio replay (voice). Of particular concern are vulnerabilities to DeepFakes [10], stemming from rapid developments in deep learning [11], which can be used to implement face swapping and voice cloning.