Enhancing Student Engagement in Distance Learning:Integrating Facial Expression Analysis and Text Analysis.

This paper presents a unique system that seeks to improve student engagement in distance learning using facial expression recognition and natural language processing (NLP) with sentiment analysis. Among them, one of the significant issues as online education progresses is the monitoring and enhancement of student engagement in real-time. The proposed system incorporates Vision Transformers (ViT) for recognizing emotions based on facial expressions and BERT-based models for recognizing sentiment from the written and spoken responses of students. Whereas facial expressions and head movements are detected by analysing the video content, the engagerment level is determined based on textual inputs and serves as a complement to the video analysis by offering a more bottom-up approach. Thus, this dual-modality approach can enhance the learning-efficiency and effectiveness of students and teachers in distance-modelled virtual classes towards optimizing the education system.