Unsupervised Insulator Defect Detection Method Based on Masked Autoencoder

With the rapid expansion of high-speed rail infrastructure, maintaining the structural integrity of insulators is critical to operational safety. However, conventional defect detection techniques typically rely on extensive labeled datasets, struggle with class imbalance, and often fail to capture l...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	Sensors (Basel, Switzerland) Ročník 25; číslo 14; s. 4271
Hlavní autori:	Song, Yanying, Xiong, Wei
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	Switzerland MDPI AG 09.07.2025 MDPI
Predmet:	Accuracy Analysis Annotations Datasets Deep learning Defects Ethical aspects insulator defect detection interval masking inference Localization masked autoencoder (MAE) Methods Neural networks Passenger rail services random masking training Semantics unsupervised learning vision transformer (ViT) interval masking inference insulator defect detection vision transformer (ViT) masked autoencoder (MAE) random masking training defect detection and localization unsupervised learning
ISSN:	1424-8220, 1424-8220
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Popis
Shrnutí:	With the rapid expansion of high-speed rail infrastructure, maintaining the structural integrity of insulators is critical to operational safety. However, conventional defect detection techniques typically rely on extensive labeled datasets, struggle with class imbalance, and often fail to capture large-scale structural anomalies. In this paper, we present an unsupervised insulator defect detection framework based on a masked autoencoder (MAE) architecture. Built upon a vision transformer (ViT), the model employs an asymmetric encoder-decoder structure and leverages a high-ratio random masking scheme during training to facilitate robust representation learning. At inference, a dual-pass interval masking strategy enhances defect localization accuracy. Benchmark experiments across multiple datasets demonstrate that our method delivers competitive image- and pixel-level performance while significantly reducing computational overhead compared to existing ViT-based approaches. By enabling high-precision defect detection through image reconstruction without requiring manual annotations, this approach offers a scalable and efficient solution for real-time industrial inspection under limited supervision.
Bibliografia:	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23
ISSN:	1424-8220 1424-8220
DOI:	10.3390/s25144271