Facial expression image mapping for brain-computer interface using El type classification

Recently, brain signal processing techniques related to brain computer interface (BCI) have been focused on rather entertainment areas for healthy users than alternative communication means for motor-impaired people. In this paper, we present a method for El type classification and facial expression mapping on expression pictures from EEG data using fuzzy rule based models. In this study, we suggest links between facial expression image space and personal tendency such as extrovert or introvert determined by EEG data from users. Extrovert or introvert (El) types including very extrovert, extrovert, medium, introvert, and very introvert are defined as building El fuzzy model based on alpha and beta values from EEG input. Also, EEG fuzzy model taking a membership value of El is designed to match a facial expression image among six facial expression image samples via defuzzification process. Finally, we have shown experimental results to validate the proposed methods.