Edge-Based and Privacy-Preserving Multi-Modal Monitoring of Student Engagement in Online Learning Environments

With engagement being an early predictor for a student's learning achievements, it is paramount that teachers can observe the behavior of their audience to keep them engaged, for example, with interactive lectures. In order to address this concern, we present an edge-based multimodal engagement analysis solution for teachers to maintain an engagement overview of their entire audience, including those in distance learning settings. We designed and evaluated an edge-based browser solution for the analysis of different behavior modalities with cross-user aggregation through secure multiparty computation.