Brainwaves as authentication method: Proving feasibility under two different approaches

Acquisition of EEG is becoming more and more affordable, and relatively high-quality signals are now digitizable even with consumer-grade hardware. The improvement in EEG hardware opens many possibilities for its applied usage. In this paper we explore the possibility of using the EEG as an secure authentication method. To achieve this goal, we process the EEG under two different methods: Cross-correlation and Divergent Autoencoding. The first method being a traditional approach to compare the similarity between two signals, and the second a classification tool originally developed to simulate aspects of human intelligence in the domain of cognitive psychology, but with potentially interesting applications in engineering.