Author(s): Basavaprasad B and Chandrashekhar S

Abstract: Emotion recognition plays a crucial role in enhancing personalized learning experiences and engagement in online education. This research explores the application of Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) for detecting and analysing students’ emotions in virtual learning environments. Drawing on recent studies, the paper highlights how CNNs effectively capture spatial features from learners’ facial expressions, while RNNs, including Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU), capture temporal patterns over time. Hybrid CNN-RNN architectures, combined with attention mechanisms and multimodal inputs such as facial expressions, audio, and gaze tracking, have demonstrated superior performance compared to standalone models. The results of the literature review reveal the advantages of deep learning in affective computing, while also addressing challenges related to dataset variability, privacy, and deployment in real-time settings. By leveraging CNN-RNN-based models, online learning platforms can provide adaptive and emotionally responsive feedback, thereby improving engagement and academic performance. The study concludes with future directions emphasizing lightweight architectures, ethical considerations, and explainable AI.

DOI: 10.33545/26633582.2025.v7.i2a.205

Pages: 82-91 | Views: 541 | Downloads: 334

Download Full Article: Click Here