mlsanityData to 14 Oct 2026

GATE-AD: Graph Attention Network Encoding For Few-Shot Industrial Visual Anomaly Detection

arXiv 2603.15300 ↗First posted 16 Mar 2026cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Few-Shot Industrial Visual Anomaly Detection (FS-IVAD) comprises a critical task in modern manufacturing settings, where automated product inspection systems need to identify rare defects using only a handful of normal/defect-free training samples. In this context, the current study introduces a novel reconstruction-based approach termed GATE-AD. In particular, the proposed framework relies on the employment of a masked, representation-aligned Graph Attention Network (GAT) encoding scheme to learn robust appearance patterns of normal samples. By leveraging dense, patch-level, visual feature tokens as graph nodes, the model employs stacked self-attentional layers to adaptively encode complex, irregular, non-Euclidean, local relations. The graph is enhanced with a representation alignment component grounded on a learnable, latent space, where high reconstruction residual areas (i.e., defects) are assessed using a Scaled Cosine Error (SCE) objective function. Extensive comparative evaluation on the MVTec AD, VisA, and MPDD industrial defect detection benchmarks demonstrates that GATE-AD achieves state-of-the-art performance across the 1- to 8-shot settings, combining the highest detection accuracy (increase up to 1.8% in image AUROC in the 8-shot case in MPDD) with the lowest per-image inference latency (at least 25.05% faster), compared to the best-performing literature methods. In order to facilitate reproducibility and further research, the source code of GATE-AD is available at https://github.com/gthpapadopoulos/GATE-AD.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 25 of its 33 references that are in the map. First posted on arXiv 16 Mar 2026.

Builds on

25 of 33 references.

References · 25

  1. 2017Graph Attention Networkscited by 8,923
  2. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  3. 2021Masked Autoencoders Are Scalable Vision Learnerscited by 7,403
  4. 2025DINOv3cited by 1,414
  5. 2021Towards Total Recall in Industrial Anomaly Detectioncited by 776
  6. 2022GraphMAE: Self-Supervised Masked Graph Autoencoderscited by 448
  7. 2023WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentationcited by 315
  8. 2020Sub-Image Anomaly Detection with Deep Pyramid Correspondencescited by 300
  9. 2023Deep Industrial Image Anomaly Detection: A Surveycited by 148
  10. 2022Registration based Few-Shot Anomaly Detectioncited by 133
  11. 2024Toward Generalist Anomaly Detection via In-context Residual Learning with Few-shot Sample Promptscited by 89
  12. 2023Pushing the Limits of Fewshot Anomaly Detection in Industry Vision: Graphcorecited by 75
13 more
  1. 2022A Survey on Graph Representation Learning Methodscited by 66
  2. 2019k-hop Graph Neural Networkscited by 57
  3. 2024AnomalyDINO: Boosting Patch-based Few-shot Anomaly Detection with DINOv2cited by 53
  4. 2022ADTR: Anomaly Detection Transformer with Feature Reconstructioncited by 31
  5. 2024UniVAD: A Training-free Unified Model for Few-shot Visual Anomaly Detectioncited by 28
  6. 2025AnomalyR1: A GRPO-based End-to-end MLLM for Industrial Anomaly Detectioncited by 28
  7. 2022A Survey on Unsupervised Anomaly Detection Algorithms for Industrial Imagescited by 21
  8. 2024Kernel-Aware Graph Prompt Learning for Few-Shot Anomaly Detectioncited by 15
  9. 2024FADE: Few-shot/zero-shot Anomaly Detection Engine using Large Vision-Language Modelcited by 7
  10. 2025One-to-Normal: Anomaly Personalization for Few-shot Anomaly Detectioncited by 7
  11. 2025FastRef:Fast Prototype Refinement for Few-Shot Industrial Anomaly Detectioncited by 7
  12. 2025Foundation Visual Encoders Are Secretly Few-Shot Anomaly Detectorscited by 5
  13. 2025Search is All You Need for Few-shot Anomaly Detectioncited by 1

Built on by

Not cited yet.