mlsanityData to 14 Oct 2026

Training-Free Human-in-the-Loop Anomaly Detection via Memory Bank Correction

arXiv 2608.17775 ↗First posted 18 Aug 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Anomaly detectors are hardest to deploy exactly where training data is scarcest: a newly commissioned production line has a handful of verified "golden" samples and no machine-learning engineer on the factory floor. We present a training-free human-in-the-loop framework in which a domain expert corrects a PatchCore detector by direct memory bank editing: no retraining, no gradients, no original training data. A false-positive correction inserts the reviewed image's normal patches through a self-calibrating novelty gate admitting only those beyond the median pool-normal nearest-neighbour distance. From a bank built on only ten golden samples, operator corrections close a median 66% of the gap to an uncorrected fully trained bank (mean 80%, raised by three categories that overshoot parity), significantly improving 12 of 15 MVTec AD categories and harming none: ten samples plus corrections outperform hundreds of samples without them. On already-trained banks the headroom is smaller and concentrated where the bank undersamples normal appearance (gated: toothbrush +0.10, metal nut +0.09, zipper +0.05, screw +0.05), and no category except grid is significantly harmed. Evaluation uses a held-out protocol (20 splits per category, Holm-corrected Wilcoxon), because corrected images entering the bank inflate naive evaluation toward AUROC 1.0 by memorisation. Passive and active querying are statistically indistinguishable; a matched-label-budget control attributes gains to deployment-time label production at 43% of exhaustive-review cost; a defect-memory extension fails decisively. Feedback is simulated from ground truth; live expert trials, where mislabelling is costliest on small banks, remain future work.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 23 of its 41 references that are in the map. First posted on arXiv 18 Aug 2026.

Builds on

23 of 41 references.

References · 23

  1. 2016Wide Residual Networkscited by 3,070
  2. 2021Towards Total Recall in Industrial Anomaly Detectioncited by 776
  3. 2020PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localizationcited by 594
  4. 2017Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discoverycited by 566
  5. 2023WinCLIP: Zero-/Few-Shot Anomaly Classification and Segmentationcited by 315
  6. 2020Sub-Image Anomaly Detection with Deep Pyramid Correspondencescited by 300
  7. 2023SimpleNet: A Simple Network for Image Anomaly Detection and Localizationcited by 256
  8. 2023AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Modelscited by 171
  9. 2021Student-Teacher Feature Pyramid Matching for Anomaly Detectioncited by 146
  10. 2022Registration based Few-Shot Anomaly Detectioncited by 133
  11. 2023Pushing the Limits of Fewshot Anomaly Detection in Industry Vision: Graphcorecited by 75
  12. 2024Unveiling the Anomalies in an Ever-Changing World: A Benchmark for Pixel-Level Anomaly Detection in Continual Learningcited by 13
11 more
  1. 2019Effectiveness of Tree-based Ensembles for Anomaly Discovery: Insights, Batch and Streaming Active Learningcited by 9
  2. 2024Taming False Positives in Out-of-Distribution Detection with Human Feedbackcited by 7
  3. 2018Active Anomaly Detection via Ensemblescited by 4
  4. 2022Little Help Makes a Big Difference: Leveraging Active Learning to Improve Unsupervised Time Series Anomaly Detectioncited by 4
  5. 2024A Reliable Framework for Human-in-the-Loop Anomaly Detection in Time Seriescited by 3
  6. 2025CADIC: Continual Anomaly Detection Based on Incremental Coresetcited by 2
  7. 2023Human in the AI loop via xAI and Active Learning for Visual Inspectioncited by 1
  8. 2025ASAT: Adaptive Scoring and Thresholding with Human Feedback for Robust Out-of-Distribution Detectioncited by 1
  9. 2025Leveraging Intermediate Representations of Time Series Foundation Models for Anomaly Detectioncited by 1
  10. 2025MECAD: A multi-expert architecture for continual anomaly detectioncited by 1
  11. 2026Normality-Preserving Continual Industrial Anomaly Detection via Orthogonal LoRA Bankscited by 1

Built on by

Not cited yet.