EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Swin-HSSAM: A green coffee bean grading method by Swin transformer

Yujie Jiao, Yuqing Zhao, Aoying Jia, Tianyun Wang, Jiashun Li, Kaiming Xiang, Hangyu Deng, Maochang He, Rui Jiang, Yue Zhang iD

PLOS One · 14.05.2025 · 20 (5) · e0322198

DOI: 10.1371/journal.pone.0322198 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS One, online erschienen am 14.05.2025. Verfasst von Yujie Jiao und 9 weiteren Autor:innen. Beteiligte Einrichtung: Faculty of Mechanical and Electrical Engineering, Yunnan Agriculture University, Kunming, China. Hauptthemen laut MEDLINE: Coffee, Coffea. Die Arbeit wurde bisher 10-mal zitiert (Crossref) und nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Major Science and Technology Projects in Yunnan Province, Special Fund for Scientific Innovation Strategy-Construction of High-level Academy of Agriculture Science, Yunnan Key Laboratory of Coffee.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

A novel shifted window (Swin) Transformer coffee bean grading model called Swin-HSSAM has been proposed to address the challenges of accurately classifying green coffee beans and low identification accuracy. This model integrated the Swin Transformer as the backbone network; fused features from the second, third, and fourth stages using the high-level screening-feature pyramid networks module; and incorporated the selective attention module (SAM) for discriminative power enhancement to enhance the feature outputs before classification. Fusion Loss was employed as the loss function. Experimental results on a proprietary coffee bean dataset demonstrate that the Swin-HSSAM model achieved an average grading accuracy of 96.34% for the three grading as well as the nine defect subdivision levels, outperforming the AlexNet, VGG16, ResNet50, MobileNet-v2, Vision Transformer (ViT), and CrossViT models by 3.86%, 2.56%, 0.44%, 4.05%, 5.36%, and 5.40% percentage points, respectively. Evaluations on a public coffee bean dataset revealed that, compared with the aforementioned models, the Swin-HSSAM model improved the average grading accuracy by 1.01%, 0.13%, 4.75%, 0.85%, 0.73%, and 0.27% percentage points, respectively. These results indicate that the Swin-HSSAM model not only achieved high grading accuracy but also exhibited broad applicability, providing a novel solution for the automated grading and identification of green coffee beans.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

CoffeeCoffeaAlgorithmsSeeds

Substanzen

Coffee

Autor:innen und Institutionen

  • Faculty of Mechanical and Electrical Engineering, Yunnan Agriculture University, Kunming, China.
  • Key Laboratory for Crop Production and Smart Agriculture of Yunnan Province, Kunming, China.
  • Yunnan Key Laboratory of Coffee, Kunming, China.
  • Faculty of Transportation Engineering, Kunming University of Science and Technology, Kunming, China.
  • College of Big Data, Yunnan Agricultural University, Kunming, China.

ORCID-Profile: Yue Zhang

Förderung

  • Major Science and Technology Projects in Yunnan Province
  • Special Fund for Scientific Innovation Strategy-Construction of High-level Academy of Agriculture Science
  • Yunnan Key Laboratory of Coffee

Zitieren

APA

Jiao, Y., Zhao, Y., Jia, A., Wang, T., Li, J., Xiang, K., Deng, H., He, M., Jiang, R., & Zhang, Y. (2025). Swin-HSSAM: A green coffee bean grading method by Swin transformer. PLOS One, 20(5), e0322198. https://doi.org/10.1371/journal.pone.0322198

Vancouver

Jiao Y, Zhao Y, Jia A, Wang T, Li J, Xiang K, et al. Swin-HSSAM: A green coffee bean grading method by Swin transformer. PLOS One. 2025;20(5):e0322198. doi:10.1371/journal.pone.0322198

BibTeX
@article{Jiao2025,
  author = {Jiao, Yujie and Zhao, Yuqing and Jia, Aoying and Wang, Tianyun and Li, Jiashun and Xiang, Kaiming and Deng, Hangyu and He, Maochang and Jiang, Rui and Zhang, Yue},
  title = {Swin-HSSAM: A green coffee bean grading method by Swin transformer},
  journal = {PLOS One},
  year = {2025},
  volume = {20},
  number = {5},
  pages = {e0322198},
  doi = {10.1371/journal.pone.0322198}
}
Kostenlos starten →