EUVIMEDEUVIMED entdecken
Open Access · CC BY-SA 4.0

Comparative study of energy and accuracy in spatio-temporal models for engagement classification

Fahrur Aslami, Peeraya Sripian

International Journal of Advances in Intelligent Informatics · 31.08.2026 · 12 (3) · 693

DOI: 10.26555/ijain.v12i3.2402 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Advances in Intelligent Informatics, online erschienen am 31.08.2026. Verfasst von Fahrur Aslami und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Energy consumption is a growing concern in deep learning, motivating the Green AI paradigm, where models are evaluated not only based on predictive performance but also on energy efficiency. Most existing evaluations focus on spatial tasks and do not fully capture the computational and temporal costs of video-based workflows. This study presents a systematic energy evaluation of spatio-temporal deep learning for video-based student engagement classification. Using the DAiSEE dataset, we compare EfficientNet variants (B0–B7) as frame-level feature extractors and model temporal dynamics with LSTMs on fixed-length sequences. GPU power consumption is measured during training using nvidia-smi, and energy efficiency is quantified with the Kappa Energy Index (KEI), defined as Cohen's Kappa divided by energy consumption (kWh). The results show a clear trade-off between accuracy and energy: EfficientNet-B7 achieves the highest accuracy (0.62) but incurs the highest energy cost (≈5 kWh), resulting in a low KEI, while EfficientNet-B0 achieves competitive accuracy (0.59) with the highest KEI (2.147) due to its low energy consumption (≈0.38 kWh). EfficientNet-B3 strikes a favorable balance (accuracy ≈ 0.61, KEI = 0.747), outperforming larger models under resource constraints. These findings suggest that deeper models do not always maximize energy efficiency, and the KEI value provides a practical metric to guide the selection of energy-efficient models.

Zitieren

APA

Aslami, F., & Sripian, P. (2026). Comparative study of energy and accuracy in spatio-temporal models for engagement classification. International Journal of Advances in Intelligent Informatics, 12(3), 693. https://doi.org/10.26555/ijain.v12i3.2402

Vancouver

Aslami F, Sripian P. Comparative study of energy and accuracy in spatio-temporal models for engagement classification. International Journal of Advances in Intelligent Informatics. 2026;12(3):693. doi:10.26555/ijain.v12i3.2402

BibTeX
@article{Aslami2026,
  author = {Aslami, Fahrur and Sripian, Peeraya},
  title = {Comparative study of energy and accuracy in spatio-temporal models for engagement classification},
  journal = {International Journal of Advances in Intelligent Informatics},
  year = {2026},
  volume = {12},
  number = {3},
  pages = {693},
  doi = {10.26555/ijain.v12i3.2402}
}
Kostenlos starten →