Different stimuli and machine learning methods for person identification

This study delves into the identification of individuals through gaze patterns, employing a variety of stimuli and machine learning techniques. The objective is to discern the impact of diverse stimuli on the precision of identifying individuals via their eye movements. By harnessing MATLAB and neural networks, the coordinates for each frame were meticulously extracted, facilitating the classification of individuals with a classification learner. The research encompassed 55 participants. The results show significant differences in accuracy, with stimuli performing better and person identification based on eye movements alone being successful. This approach provides insights for use in biometric authentication and contributes to the further development of gaze-based identification systems.