EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Multimodal Academic Emotion Detection using Hybrid CNN–BiLSTM Networks with Attention Mechanisms

Shaik Khaja Mohiddin, Khadija Slimani

Procedia Computer Science · 2026 · 283 · 3387-3396

DOI: 10.1016/j.procs.2026.06.402 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Procedia Computer Science, 2026. Verfasst von Shaik Khaja Mohiddin und 1 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

To improve adaptive and personalized learning, matching academic feelings are accurate and essential to intelligent tutoring systems. The Multimodal Academic Emotion Detection Network offered in this paper, named MAED-Net, is based on Convolutional Neural Networks (CNN), Bidirectional Long Short-Term Memory (BiLSTM), and a mechanism of attention in order to efficiently extract both spatial, temporal, and contextual attributes, related to text, audio, and visual representations. Compared to the currently used unimodal or two-modal models that are unable to address cross-modal connections, MAED-Net accomplishes hierarchical integration by weighting with attention, which makes it more interpretable and stronger. The model was tested and trained on the IEMOCAP dataset with 89.4% accuracy, 88.2% precision, 87.9% recall, and an F1-score of 88.0% performing well compared to recent state-of-the-art models including MSER, AFEFS, MTAF, and AVaTER. The suggested framework is a scalable and realtime means of emotion-sensitive e-learning systems, which includes a new hybrid deep learning method to improve interactions, flexibility, and emotional intelligence in contemporary educational settings.

Zitieren

APA

Mohiddin, S. K., & Slimani, K. (2026). Multimodal Academic Emotion Detection using Hybrid CNN–BiLSTM Networks with Attention Mechanisms. Procedia Computer Science, 283, 3387-3396. https://doi.org/10.1016/j.procs.2026.06.402

Vancouver

Mohiddin SK, Slimani K. Multimodal Academic Emotion Detection using Hybrid CNN–BiLSTM Networks with Attention Mechanisms. Procedia Computer Science. 2026;283:3387-3396. doi:10.1016/j.procs.2026.06.402

BibTeX
@article{Mohiddin2026,
  author = {Mohiddin, Shaik Khaja and Slimani, Khadija},
  title = {Multimodal Academic Emotion Detection using Hybrid CNN–BiLSTM Networks with Attention Mechanisms},
  journal = {Procedia Computer Science},
  year = {2026},
  volume = {283},
  pages = {3387--3396},
  doi = {10.1016/j.procs.2026.06.402}
}
Kostenlos starten →