Exploring AI Facial Recognition for Real-time Emotion Detection: Assessing Student Engagement in Online Learning Environments

Sherin Eliyas,J. Shan

DOI: https://doi.org/10.1109/AIIoT58432.2024.10574587

2024-05-03

Abstract:AI facial recognition for real-time emotion detection in online learning environments intends to analyze students’ sentiments throughout online classes, recognizing emotions such as focused, neutral, distracted, enraged, thrilled, or dissatisfied states. The proposed approach uses a generative model trained by conditional Generative Adversarial Networks (cGAN) to accomplish De-expression Residue Learning (DeRL), a facial expression recognition technique. DeRL prioritizes gathering residue for delicate emotion recognition by filtering expressive material and storing it in intermediary layers. This technology gives educators and researchers real-time insights into students’ emotional states, allowing for a more thorough understanding of collective student involvement through the lens of socially shared control of learning. Experiments on seven facial expression databases, including two pre-training datasets, indicate that DeRL outperforms other methods in accurately recognizing and classifying students’ emotions. The findings emphasize its ability to capture subtle feelings during online learning. The discussion dives into theoretical and practical implications, highlighting AI’s potential to enhance the evaluation of student participation in online learning environments. The study concludes by providing relevant information for building tailored learning experiences, indicating a possible avenue for incorporating AI in the classroom.

Computer Science,Education