Human implicit intent discrimination using EEG and eye movement

In this paper, we propose a new human implicit intent understanding model based on multi-modal information, which is a combination of eye movement data and brain wave signal obtained from eye-tracker and Electroencephalography (EEG) sensors respectively. From the eye movement data, we extract human implicit intention related to features such as fixation count and fixation duration corresponding to the areas of interest (AOI). Also, we analyze the EEG signals based on phase synchrony method. Combining the eye movement and EEG information, we train several classifiers such as support vector machine classifier, Gaussian Mixture Model and Naïve Bayesian, which can successfully identify the human's implicit intention into two defined categories, i.e. navigational and informational intentions. Experimental results show that the human implicit intention can be better understood using multimodal information.