An exploratory study of detecting emotion states using eye-tracking technology

Studying eye movement has proven to be useful in the study of detecting and understanding human emotional states. This paper aims to investigate eye movement features: pupil size, time of first fixation, first fixation duration, fixation duration and fixation count in clips emotional stimulation. Thirty seven subjects' pupil responses were measured while watching two pleasant and unpleasant emotional clips. The results showed that the fixation duration and fixation count significantly different between pleasant and unpleasant clip arousal. These results suggest that the measurement of eye fixation may be a potentially useful computer input for detecting positive and negative emotional state.