The Covid-19 pandemic has necessitated the wearing of face masks in public which undermines the ability to discern emotional cues from the entire face. Face masks typically cover regions around the nose, cheeks, and mouth, which undermines the visual dimension of face-to-face communication. Current studies use a variety of approaches to quantify the importance of facial action units in determining emotional expression, but none specifically focus exclusively on the top portion of the face. Here, we present a method that uses the Facial Action Coding System (FACs) to quantify the performance of muscle movements in the top portion of the face using time averaged facial action units (AUs) for classifying happiness and sadness.
