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

Self-Attention Enhanced Student–Teacher Framework for Visual Anomaly Detection in Industrial Images

Hafiz A. Amjad, Irida Shallari, Mattias O’ Nils, Faisal Z. Qureshi

IEEE Access · 2026 · 14 · 36318-36330

DOI: 10.1109/access.2026.3670056 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Access, 2026. Verfasst von Hafiz A. Amjad und 3 weiteren Autor:innen. Die Arbeit nennt 48 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Sciences and Engineering Research Council of Canada (NSERC) Discovery, Mitacs Accelerate.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

In this work we introduce AeCSAD, a student–teacher framework designed to detect two types of anomalies in images of industrial components: structural anomalies (physical defects) and logical anomalies (incorrect relationships between components). Unlike prior methods, AeCSAD extends the component segmentation–based logical anomaly detection scheme (c. 2024) with self-attention mechanisms, enabling more effective relational modeling. We demonstrate consistent improvements on the MVTec LOCO Anomaly Detection benchmark. Specifically, AeCSAD employs a global student network with self-attention for reasoning across distant components, complemented by a local student network for fine-grained analysis. Additionally, a patch histogram module measures the frequency distribution of components, allowing the system to detect irregularities in their occurrence. During inference, anomaly scores from the histogram module and the fused local–global networks are combined to produce the final anomaly score. Experiments show that AeCSAD achieves superior average AUROC performance on both structural and logical anomaly detection tasks compared to prior approaches.

Förderung

  • National Sciences and Engineering Research Council of Canada (NSERC) Discovery
  • Mitacs Accelerate

Zitieren

APA

Amjad, H. A., Shallari, I., Nils, M. O., & Qureshi, F. Z. (2026). Self-Attention Enhanced Student–Teacher Framework for Visual Anomaly Detection in Industrial Images. IEEE Access, 14, 36318-36330. https://doi.org/10.1109/access.2026.3670056

Vancouver

Amjad HA, Shallari I, Nils MO, Qureshi FZ. Self-Attention Enhanced Student–Teacher Framework for Visual Anomaly Detection in Industrial Images. IEEE Access. 2026;14:36318-36330. doi:10.1109/access.2026.3670056

BibTeX
@article{Amjad2026,
  author = {Amjad, Hafiz A. and Shallari, Irida and Nils, Mattias O’ and Qureshi, Faisal Z.},
  title = {Self-Attention Enhanced Student–Teacher Framework for Visual Anomaly Detection in Industrial Images},
  journal = {IEEE Access},
  year = {2026},
  volume = {14},
  pages = {36318--36330},
  doi = {10.1109/access.2026.3670056}
}
Kostenlos starten →