Enhancing Engagement in Virtual Classrooms: A Contactless Multi-Modal Emotion Detection Model

This research addresses the growing need for effective assessment of student engagement and emotional states in virtual learning environments. Recognizing challenges in accurately detecting student participation and emotional states through costly contact methods, the study develops a Multi-Modal Emotion Detection System for a Virtual Classroom using a non-intrusive, contactless approach. To overcome the limitations of contactless methods, the system employs remote Photoplethysmography (rPPG) to capture physiological attributes via real-time video streaming from webcams. The integrated system utilizes a deep learning model for speech and facial emotion detection. The findings contribute to online education by providing a non-intrusive and reliable method for monitoring students' emotional states and engagement levels. Future research directions may involve refining the model, expanding the dataset, and exploring the system's potential in detecting critical health challenges in online classrooms.