EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Lightweight Multi-Scale Framework for Human Pose and Action Classification

Alireza Saber iD, Mohammad-Mehdi Hosseini iD, Amirreza Fateh iD, Mansoor Fateh iD, Vahid Abolghasemi iD

Sensors · 2026

DOI: 10.3390/s26041102 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, 2026. Verfasst von Alireza Saber und 4 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of Computer Engineering, Shahrood University of Technology, Shahrood 36199-95161, Iran. Hauptthemen laut MEDLINE: Posture. Die Arbeit nennt 42 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Human pose classification, along with related tasks such as action recognition, is a crucial area in deep learning due to its wide range of applications in assisting human activities. Despite significant progress, it remains a challenging problem because of high inter-class similarity, dataset noise, and the large variability in human poses. In this paper, we propose a lightweight yet highly effective modular attention-based architecture for human pose classification, built upon a Swin Transformer backbone for robust multi-scale feature extraction. The proposed design integrates the Spatial Attention module, the Context-Aware Channel Attention Module, and a novel Dual Weighted Cross Attention module, enabling effective fusion of spatial and channel-wise cues. Additionally, explainable AI techniques are employed to improve the reliability and interpretability of the model. We train and evaluate our approach on two distinct datasets: Yoga-82 (in both main-class and subclass configurations) and Stanford 40 Actions. Experimental results show that our model outperforms state-of-the-art baselines across accuracy, precision, recall, F1-score, and mean average precision, while maintaining an extremely low parameter count of only 0.79 million. Specifically, our method achieves accuracies of 90.40% and 87.44% for the 6-class and 20-class Yoga-82 configurations, respectively, and 94.28% for the Stanford 40 Actions dataset.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

PostureHumansAlgorithmsPattern Recognition, AutomatedDeep LearningHuman Activities

Stichwörter der Autor:innen

classificationhuman poselightweightmulti-scale

Autor:innen und Institutionen

  • Faculty of Computer Engineering, Shahrood University of Technology, Shahrood 36199-95161, Iran.
  • Department of Computer Engineering, Sha.C., Islamic Azad University, Shahrood 43189-36199, Iran.
  • School of Computer Engineering, Iran University of Science and Technology (IUST), Tehran 13114-16846, Iran.
  • School of Computer Science and Electronic Engineering, University of Essex, Colchester CO4 3SQ, UK.

ORCID-Profile: Alireza Saber, Mohammad-Mehdi Hosseini, Amirreza Fateh, Mansoor Fateh, Vahid Abolghasemi

Zitieren

APA

Saber, A., Hosseini, M. M., Fateh, A., Fateh, M., & Abolghasemi, V. (2026). Lightweight Multi-Scale Framework for Human Pose and Action Classification. Sensors. https://doi.org/10.3390/s26041102

Vancouver

Saber A, Hosseini MM, Fateh A, Fateh M, Abolghasemi V. Lightweight Multi-Scale Framework for Human Pose and Action Classification. Sensors. 2026. doi:10.3390/s26041102

BibTeX
@article{Saber2026,
  author = {Saber, Alireza and Hosseini, Mohammad-Mehdi and Fateh, Amirreza and Fateh, Mansoor and Abolghasemi, Vahid},
  title = {Lightweight Multi-Scale Framework for Human Pose and Action Classification},
  journal = {Sensors},
  year = {2026},
  doi = {10.3390/s26041102}
}
Kostenlos starten →