Fooling a Liveness-Detecting Capacitive Fingerprint Scanner

Abstract

Biometrics are increasingly being deployed as solutions to the security problems of authentication, identification and to some extent, non-repudiation. Biometrics are also publicized by proponents to be more secure than conventional mechanisms such as passwords and tokens, while also being more convenient too since there is no need to remember passwords nor carry anything around. Yet the security of biometrics lies on the assumption that biometric traits are unique to an individual and are unforgeable; once this assumption is invalidated, the security of biometrics collapses. Therefore, it is crucial to ensure that biometric traits are indeed unforgeable. In scientific literature, proponents have invented different ways for liveness detection, in order to differentiate forged traits from real ones, based on the premise that forged traits should not have liveness. In this paper, we show that a celebrated capacitive fingerprint scanner with liveness detection claims, can be fooled by fake fingers produced by amateurs from cheap commercially available materials. This brings into question that a gap may exist between what scientific literature has proposed for liveness detection and the actual robustness of liveness-detecting fingerprint scanners available in the market against fake fingers.