enVolve+: Inertial Sensing to Reinforce Involvement of Silent Listeners during an Online Interaction

Garvit Chugh, Suchetana Chakraborty, Sandip Chakraborty

Abstract: Learning how engaged a remote participant is in an online meeting can lead to more productive discussions. If online meeting attendees turn off their mics and cameras, the presenter can't tell if they're paying attention. The research in [1] shows how earables could help remedy the issue at hand. While the prior study focused on offering speakers feedback of the classroom, the same results may be utilized to help students with their attention spans and bring back their attention to the screen during a lecture. Therefore, the proposed study adds a listener interface to previous work. The premise is, that if listeners start to doze off or drift away in an interaction, vibrations from the smartwatch might wake them up. Wearable IMU data from an in-lab user study was utilized to train machine learning models that obtained an F1 Score of 80% or above across the board. Additionally, we give persuasive evidence for the adoption of earables over other sensory modalities by analyzing experimental findings, user analysis, and current research needs in this field.