FACIAL EXPRESSION ANALYSIS FOR ONLINE LEARNING ENGAGEMENT USING DEEP LEARNING

ABSTRACT: This paper presents a Convolutional Neural Network (CNN)-based system for detecting and analyzing student engagement in online learning environments using facial expressions. The system leverages real-time webcam input or static image uploads to classify emotions that are then mapped to engagement levels such as “Engaged,” “Not Engaged,” or “Distracted.” Unlike traditional systems that use gaze tracking or manual observation, our approach automates emotion recognition using CNN trained on facial data. The application is deployed through a Django-based web interface, integrated with OpenCV for real-time face detection, and utilizes SQLite for backend data handling. Results show improved classification accuracy and offer practical utility for instructors to monitor learner attentiveness and participation during virtual sessions. Keywords — Facial Expression, Engagement Detection, CNN, Online Learning, Emotion Recognition, Real-Time Analysis, Django, OpenCV