Analysing the robust EEG channel set for person authentication

Abstract

In this paper, we present the findings on the EEG channel selection and its impact on the robustness for EEG based person authentication. We test the effect of the enhancement threshold value (Te), EEG frequency rhythms, mental task and the person identity on the selected EEG channels. Experimental validation of the work with publicly available EEG dataset, showed that the idle mental task provides the highest accuracy rates compared to other considered mental tasks. Moreover, we noticed that imaginary movement tasks provide better accuracy than actual movement tasks. Also for the frequency rhythm effect, the combined frequency rhythms increase the authentication accuracy better than using a single rhythm, so no single rhythm contains all the related identity information. Also for the Te value, we found that the less Te we consider, the more EEG channels to be included. Further, for the final part of this work, we tested if the selected channel are person specific. As a result, we found that EEG channel set, if selected for each person differently does enhance the authentication accuracy.