Classification of eye movements due to emotional stimuli

Pupil size variations due to different types of emotional stimuli were shown to have discriminative characteristics. Pupil dilates when a negative type of stimulus is presented to a person, and pupil size decreases due to a positive stimulus. In the concept of this thesis, it is aimed to inestigate the emotional state of a subject based on eye movements. International Affective Picture System was used to define 60 stimuli which are grouped as neutral, positive and negative emotions each having 20 images. Eye tracker is used to collect eye movements data from 13 volunteer students. As a result of the data collection procedure, pupil locations on the different stimuli, pupil sizes and fixation time values were computed. Several classification techniques were applied in order to distinguish the type of stimulus presented to the subjects and the best accuracy (68%) was found for k-NN. Finally, it has been shown that the physiological parameters such as pupil size and eye movements data can be used to discriminate the responses given to poisitive or negative type of stimulus.