Exploring the Potential of Eye-Tracking Technology for Emotion Recognition: A Preliminary Investigation

Emotion recognition has garnered significant attention across diverse research domains, prompting the adoption of various methodologies to discern and classify emotional states. In recent years, implicit measures such as heart-rate variability (HRV), electrodermal activity (EDA), and electroencephalogram (EEG) have gained prominence as valuable tools for capturing neural and physiological correlates of emotions. Nevertheless, the advent of novel technologies has opened up new avenues for advancing emotion recognition techniques. In this study, a preliminary investigation of the potential of eye-tracking (ET) technology as a means to recognize emotional states is presented. In particular, a reduced set of ET-derived features is extracted from a dataset obtained after an experimental campaign involving 50 healthy subjects under different stimulation conditions. The experimental findings exhibit promising results, suggesting that eye-tracking holds the potential for facilitating a deeper understanding of human emotions.