Brainwave based user identification system: A pilot study in robotics environment

Abstract

The world is seeing the emergence of brainwave computing and its application in our daily communication and security is on the horizon. One of the interesting areas of interaction modality is robotic environment where involvement of illegitimate users may cause severe threat to environment thereby, a continuous monitoring is mandatory. In this paper, we have presented a study of potential of brainwaves and their computing for providing higher security. This work utilizes EEG signal generated by visual stimuli for keying in the password. In order to construct a model for user authentication, we have selected the time–frequency analysis method called wavelet analysis to decompose original signal corresponding to EEG sub-band frequency, thereby extracting statistical measures and energy calculation of each decomposed waves. Furthermore, we have selected and employed the neural network based on learning vector quantization for correct classification. Classification rate has been calculated over the different combination of channels.