EUVIMEDEUVIMED entdecken
Crossref

CPDD: A Cross-Scenario Photovoltaic Defect Detector Based on Fine-Grained Feature Autoencoding and Pseudo-Box Contrastive Learning

Zhaoyang Wang, Haiyong Chen, Zhen Cao

IEEE Transactions on Semiconductor Manufacturing · 2025-08 · 38 (3) · 612-623

DOI: 10.1109/tsm.2025.3570323 ↗

Kurz gefasst

Fachartikel in der Zeitschrift IEEE Transactions on Semiconductor Manufacturing, 2025-08. Verfasst von Zhaoyang Wang und 2 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 63 Literaturangaben. Gefördert von: National Natural Science Foundation of China, National Key Research and Development Program of China, Hebei Province Central Guidance Local Science and Technology Development Project.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The vision foundation model, relying on large-scale pre-training, has advanced image comprehension capabilities and excels in general scenarios. However, its performance remains suboptimal in specialized tasks, such as photovoltaic (PV) cells defect detection. This limitation stems from the models’ lack of domain-specific prior knowledge. To address this, we propose a two-stage pre-training framework comprising fine-grained feature autoencoding (FFA) and pseudo-box contrastive learning (PCL), which leverages extensive unlabeled raw images to inject domain expertise into the model. First, we investigate the fine-grained feature autoencoder, which trains a detail-sensitive vision transformer (ViT) backbone by reconstructing the histogram of oriented gradients (HOG) of masked images. Second, we pre-train the detection head through contrastive learning. Using selective search (SS) to generate pseudo-boxes, we treat paired boxes from two augmented views of an image as positive samples. The abundant unsupervised pseudo-boxes optimize the detection head’s local representation and localization capabilities. Finally, we fully fine-tune the model with labeled images. Based on this methodology, we build the cross-scenario photovoltaic defect detector (CPDD). The experimental results demonstrate that CPDD achieves state-of-the-art (SOTA) mAP50 scores on three benchmarks, outperforming detectors pre-trained on the COCO dataset as well as those specifically designed for PV defect detection.

Förderung

  • National Natural Science Foundation of China
  • National Key Research and Development Program of China
  • Hebei Province Central Guidance Local Science and Technology Development Project
  • Hebei Province Major Science and Technology Support Project
  • Interdisciplinary Postgraduate Training Program of Hebei University of Technology

Zitieren

APA

Wang, Z., Chen, H., & Cao, Z. (2025). CPDD: A Cross-Scenario Photovoltaic Defect Detector Based on Fine-Grained Feature Autoencoding and Pseudo-Box Contrastive Learning. IEEE Transactions on Semiconductor Manufacturing, 38(3), 612-623. https://doi.org/10.1109/tsm.2025.3570323

Vancouver

Wang Z, Chen H, Cao Z. CPDD: A Cross-Scenario Photovoltaic Defect Detector Based on Fine-Grained Feature Autoencoding and Pseudo-Box Contrastive Learning. IEEE Transactions on Semiconductor Manufacturing. 2025;38(3):612-623. doi:10.1109/tsm.2025.3570323

BibTeX
@article{Wang2025,
  author = {Wang, Zhaoyang and Chen, Haiyong and Cao, Zhen},
  title = {CPDD: A Cross-Scenario Photovoltaic Defect Detector Based on Fine-Grained Feature Autoencoding and Pseudo-Box Contrastive Learning},
  journal = {IEEE Transactions on Semiconductor Manufacturing},
  year = {2025},
  volume = {38},
  number = {3},
  pages = {612--623},
  doi = {10.1109/tsm.2025.3570323}
}
Kostenlos starten →