Abstract
The COVID-19 pandemic has led to a surge in video content consumption, but measuring viewers' empathy towards the content has been limited to subjective evaluations or attached physiological apparatus. In this study, we introduced a novel non-contact physiological method for measuring empathy towards video content by assessing the synchronization of facial micromovements between the subject and object (i.e., person) within the media. We recorded facial micromovements and heart rate variability (HRV) remotely using a camera while 62 subjects watched one video each, designed and validated to elicit one of four two-dimensional emotions: pleasant-aroused, pleasant-relaxed, unpleasant-aroused, and unpleasant-relaxed. We also collected the subjects' self-assessed emotions and empathy using a questionnaire. The results confirmed that the stimuli effectively induced the intended arousal in the subjects, as evidenced by both self-reported emotions and HRV responses that suggested higher arousal was associated with stronger activity in the sympathetic nervous system. A closer examination of HRV indicators, such as SDNN and Total Power values, showed an amplification during the unpleasant state. We interpret this as the body's dynamic response to stressors, underlining the autonomic nervous system's proactive role in responding to such stimuli. In a broader context, our results emphasized that while subjects showcased augmented empathy during aroused conditions, the introduction of stressors, represented by unpleasant content, led to a dampening of this empathetic response. This findings demonstrate the potential of non-contact physiological methods for measuring empathy toward video content.