mlsanityData to 14 Oct 2026

Explainable Visual Anomaly Detection via Concept Bottleneck Models

arXiv 2511.20088 ↗First posted 25 Nov 2025cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

In recent years, VAD (Visual Anomaly Detection) has attracted significant attention due to its high interpretability, as it provides visual explanations by highlighting anomalous regions within an image. However, although these visual explanations can be helpful, they lack a direct and semantically meaningful interpretation for users. To address this limitation, we propose extending CBMs (Concept Bottleneck Models) to the VAD setting. By learning meaningful concepts, the network can provide human-interpretable descriptions of anomalies, offering a novel and more insightful way to explain them. Our main contributions are: (i) we introduce a concept-based framework for anomaly explanation by extending CBMs to the VAD setting for the first time; (ii) we evaluate multiple supervision regimes, ranging from supervised to unsupervised anomaly settings; (iii) we propose a dual-branch architecture that combines a CBM-based Concept branch for concept-level explanations with a Visual branch for pixel-level anomaly localization, bridging semantic and spatial interpretability. When evaluated across three well-established VAD benchmarks, our approach, Concept-Aware Visual Anomaly Detection (ourmethod), achieves performance comparable to classic VAD methods (up to 0.97 I-AUROC on MVTec AD) while providing richer, concept-driven explanations. A user study (582 responses) shows that ourmethod explanations are rated effective in 73.5% of cases (vs. 60.7% for the SOTA VLM-based explainer Anomaly-OV) and misleading in only 3.1% (vs. 11.7%). Beyond explanation quality, CBMs enable human-in-the-loop correction: concept-level interventions further boost detection in the challenging unsupervised setting, closing part of the gap to fully supervised performance.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 34 of its 36 references that are in the map. First posted on arXiv 25 Nov 2025.

Builds on

34 of 36 references.

References · 34

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  4. 2018MobileNetV2: Inverted Residuals and Linear Bottleneckscited by 4,795
  5. 2019Optuna: A Next-generation Hyperparameter Optimization Frameworkcited by 2,501
  6. 2025Gemma 3 Technical Reportcited by 1,824
  7. 2020Concept Bottleneck Modelscited by 1,032
  8. 2021Towards Total Recall in Industrial Anomaly Detectioncited by 776
  9. 2020PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localizationcited by 594
  10. 2017Unsupervised Anomaly Detection with Generative Adversarial Networks to Guide Marker Discoverycited by 566
  11. 2021CutPaste: Self-Supervised Learning for Anomaly Detection and Localizationcited by 531
  12. 2021DRAEM -- A discriminatively trained reconstruction embedding for surface anomaly detectioncited by 514
22 more
  1. 2022SPot-the-Difference Self-Supervised Pre-training for Anomaly Detection and Segmentationcited by 400
  2. 2022Language in a Bottle: Language Model Guided Concept Bottlenecks for Interpretable Image Classificationcited by 312
  3. 2023Label-Free Concept Bottleneck Modelscited by 285
  4. 2022A Unified Model for Multi-class Anomaly Detectioncited by 268
  5. 2021FastFlow: Unsupervised Anomaly Detection and Localization via 2D Normalizing Flowscited by 227
  6. 2023AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Modelscited by 171
  7. 2021SegmentMeIfYouCan: A Benchmark for Anomaly Segmentationcited by 150
  8. 2023Tree-Structured Parzen Estimator: Understanding Its Algorithm Components and Their Roles for Better Empirical Performancecited by 147
  9. 2021Student-Teacher Feature Pyramid Matching for Anomaly Detectioncited by 146
  10. 2024Real-IAD: A Real-World Multi-View Dataset for Benchmarking Versatile Industrial Anomaly Detectioncited by 139
  11. 2023VadCLIP: Adapting Vision-Language Models for Weakly Supervised Video Anomaly Detectioncited by 123
  12. 2023BMAD: Benchmarks for Medical Anomaly Detectioncited by 82
  13. 2024Adapting Visual-Language Models for Generalizable Anomaly Detection in Medical Imagescited by 69
  14. 2024VLG-CBM: Training Concept Bottleneck Models with Vision-Language Guidancecited by 65
  15. 2023A Closer Look at the Intervention Procedure of Concept Bottleneck Modelscited by 64
  16. 2025LogicAD: Explainable Anomaly Detection via VLM-based Text Feature Extractioncited by 25
  17. 2024Unseen Visual Anomaly Generationcited by 23
  18. 2023Component-aware anomaly detection framework for adjustable and logical industrial visual inspectioncited by 22
  19. 2022Concept-based Explanations for Out-Of-Distribution Detectorscited by 16
  20. 2024PaSTe: Improving the Efficiency of Visual Anomaly Detection at the Edgecited by 16
  21. 2023Transparent Anomaly Detection via Concept-based Explanationscited by 5
  22. 2025No Label Left Behind: A Unified Surface Defect Detection Model for all Supervision Regimescited by 4

Built on by

Not cited yet.