Publication Date


Document Type



With the current COVID-19 pandemic, group communication is often restricted to virtual video-conferencing platforms like Zoom in order to inhibit the spread of the virus. The virtual communication environment affects our ability to assess group emotion and support verbal messages through nonverbal communication. Because virtual meetings create visibility restrictions due to limited camera view, body language is occluded, and faces are now at the forefront of social interactions within groups. Since faces are still visible, it allows for some key components of interpersonal interactions to still occur, such as facial mimicry. Facial mimicry occurs when one person mirrors another person's facial expressions. Most research on facial mimicry has been conducted on face-to-face interactions. Further studies have also shown that facial mimicry exists when an individual is reacting to a recorded video containing different expressions. However, there is limited research on facial mimicry within video-conferencing conversations. Our research aims to use facial expression recognition techniques to analyze if facial mimicry exists during group conversations over virtual platforms through facial action units and expressions. For this purpose, we used current state-of-the-art methods to recognize and analyze the activation of eye gaze, seven universal facial expressions, and seventeen commonly presented facial action units over time for each participant within various Zoom meetings that were uploaded on Youtube to measure facial mimicry. From observing the simultaneous activation of facial action units, our findings suggest that facial mimicry, specifically in reaction to smiling and positive facial expressions, does exist in video-conferencing group conversations. We plan to conduct future research to determine whether this positive facial mimicry improves group emotion and productivity.



Download Full Text (1.5 MB)

Recognizing Facial Mimicry In Virtual Group Conversations