Deep Learning-Based Interactive Dashboard for Enhancing Online Classroom Experience Through Student Emotion Analysis

An interactive analytical dashboard that analyzes students’ facial expressions during online lectures is crucial for digital learning platforms. This research addresses the need for educational institutions to analyze individual students’ emotions to improve teaching standards. Given the challenge of occluded facial data, we employ a regenerative Generative Adversarial Network (GAN) to reconstruct these occluded regions. Subsequently, the emotions of the students are predicted and analyzed using our proposed interactive dashboard, which incorporates additional inputs such as subject name and teaching faculty. The dashboard visualizes various charts and analytics to support informed decision-making. We validated our deep learning model using the CK+ dataset, achieving notable accuracy in classifying each type of emotion. Our results demonstrate that the model can effectively interpret student emotions, even in the presence of occlusions, thereby providing educators with precise, real-time emotional insights to tailor their teaching methodologies effectively.