cs.CVNov 16, 2026

Defect-Aware Masked Autoencoders for Unsupervised Clustering of Steel Surface Defects

Authors: Shrey Patel

Organizations: University of Maryland, College Park United States of America

Abstract

Reliable recognition of steel surface defects is essential for automated quality inspection, but supervised methods require manually annotated images that are costly to produce. This creates a need for methods that can learn useful defect representations directly from unlabelled surface images. This work presents Defect aware MAE, a label free masked autoencoder for clustering steel surface defects. Each image is divided into 196 patches, with 147 masked and 49 processed by a Vision Transformer encoder. Texture energy based masking preferentially hides structurally informative patches, while a shallow decoder reconstructs HOG descriptors that capture local texture and edge orientation. Strong augmentation and an intermediate layer readout are used to form the final image features. Pretraining uses no class labels, bounding boxes, detection branch, or contrastive loss. On the held out NEU DET split, UMAP followed by Ward clustering achieves 98.6% Hungarian matched accuracy, 0.968 NMI, 0.967 ARI, and 0.838 silhouette score. Under an identical encoder and training schedule, it exceeds a SimCLR contrastive baseline in accuracy, NMI, ARI and silhouette scores across 5 seeds.

Figures & tables

Explore similar work

CardsList
  1. A Masked Autoencoder Approach to Unsupervised Steel Surface Defect Recognition

    Jul 14, 2026Shrey PatelMasked AutoencodersIndustrial Anomaly Detection

  2. SPDCN: Strip-based Deformable Convolutional Network for Steel Surface Defect Segmentation

    Jul 23, 2026Zhongming Liu, Bingbing Jiang, Guangxin Wan +1Image SegmentationIndustrial Inspection

  3. Spatial Attention Supervision for Defect Localization: Exploiting Ground-Truth Masks as Training Signal in Diffusion-Augmented Defect Detection

    Sep 5, 2026Sajjad Rezvani Boroujeni, Muskan Saraf, Gnana Tulasi Makineni +2Industrial Anomaly DetectionAnomaly Localization