Action Units of Facial Expressions in Emotional Contagion

This dataset contains processed data from a two-phase experimental study that hypothesized that specific facial Action Units (AUs) could be identified and quantitatively measured to characterize and differentiate facial expressions associated with emotional contagion, specifically laughter, yawning, and mirror pain. The study aimed to determine whether morphometric measurements based on facial landmarks could statistically discriminate between these emotional expressions and neutral states.
The research examines facial Action Units (AUs) and their corresponding morphometric measurements through automated facial landmark analysis and electromyographic validation. The dataset includes quantitative measurements derived from facial recordings of participants viewing emotional stimuli without contextual information. Key features comprise normalized Euclidean distances between facial landmarks that characterize specific AUs for each emotional category: AU12 and AU25 for laughter; AU9, AU27, and AU43 for yawning; and AU4, AU7, and AU10 for mirror pain. All measurements were normalized to ensure size independence and comparability across participants.
Statistical analyses revealed that the selected facial distances effectively discriminate between the three emotional categories and neutral expressions, with particularly clear distinctions observed in lower face regions. The data corresponds to the analyses reported in the associated manuscript and may be useful for researchers interested in facial expression analysis, emotional contagion studies, or computational approaches to facial expression recognition.