Recently, there has been a growing interest in the use of brain activity for biometric systems. However, so far these studies have focused mainly on basic features of the Electroencephalography. In this study we propose an approach based on phase synchronization, to investigate personal distinctive brain network organization. To this end, the importance, in terms of centrality, of different regions was determined on the basis of EEG recordings. We hypothesized that nodal centrality enables the accurate identification of individuals. EEG signals from a cohort of 109 64-channels EEGs were band-pass filtered in the classical frequency bands and functional connectivity between the sensors was estimated using the Phase Lag Index. The resulting connectivity matrix was used to construct a weighted network, from which the nodal Eigenvector Centrality was computed. Nodal centrality was successively used as feature vector. Highest recognition rates were observed in the gamma band (equal error rate ( EER) = 0.044) and high beta band ( EER = 0.102). Slightly lower recognition rate was observed in the low beta band ( EER = 0.144), while poor recognition rates were observed for the others frequency bands. The reported results show that resting-state functional brain network topology provides better classification performance than using only a measure of functional connectivity, and may represent an optimal solution for the design of next generation EEG based biometric systems. This study also suggests that results from biometric systems based on high-frequency scalp EEG features should be interpreted with caution.

In this article, we review previous work on biometric security under a recent framework proposed in the field of adversarial machine learning. This allows us to highlight novel insights on the security of biometric systems when operating in the presence of intelligent and adaptive attackers that manipulate data to compromise normal system operation. We show how this framework enables the categorization of known and novel vulnerabilities of biometric recognition systems, along with the corresponding attacks, countermeasures, and defense mechanisms. We report two application examples, respectively showing how to fabricate a more effective face spoofing attack, and how to counter an attack that exploits an unknown vulnerability of an adaptive face-recognition system to compromise its face templates.

Adaptive biometric systems update clients' templates during operation to account for natural changes over time (e.g., aging of biometric templates). Recently, it has been shown that this update can be exploited by an attacker to compromise the clients' templates: by presenting a proper sequence of fake biometric traits to the sensor, the attacker may eventually impersonate the targeted clients without any fake trait, and even force the system to deny access to them. This attack has however been shown only for PCA-based face verification, with one template per client, under worstcase assumptions about the attacker's knowledge of the sys-tem. In this paper, we show that it can be successful even in the case of multiple templates per client, for different matchers, and under more realistic scenarios, and validate it by experiments to highlight its practical relevance.