Resting State EEG-Based Biometric System Using Concatenation of Quadrantal Functional Networks

Electroencephalography (EEG) signal-based biometric authentication systems have received remarkable attention as a potential candidate to replace or complement conventional recognition systems in various applications, such as healthcare, neuro-gaming platform, and military industries. Although several resting EEG-based biometrics have been proposed, the feasibility of the system based on local functional networks (FNs) considering the structural brain characteristics has not been explored yet. In this paper, we provide the resting state EEG-based biometric framework exploiting concatenation of quadrantal FNs, indeed, we evaluate the proposed approach using the public dataset in three different resting conditions; eye-open, eye-closed, and waiting states. Notably, the proposed scheme achieves a more robust and better performance compared to the conventional global FN paradigm. We provide the visualization applying the t-SNE technique to comprehend the effects of the proposed approach, furthermore, we represent the potentials of the augmented FN method complementarily combining both global FN and local FNs. Our findings imply the viability of concatenating regional FNs for resting the EEG-based biometrics in the verification problem, hence we suggest the proposed schemes for improving biometric systems.