mlsanityData to 14 Oct 2026

Rethinking Continual Anomaly Detection on the Edge: Benchmarking Under Realistic Industrial Conditions

arXiv 2605.24251 ↗First posted 22 May 2026cs.LG cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Continual anomaly detection (CAD) addresses the need for industrial inspection systems to adapt to evolving production conditions, yet existing methods share three critical gaps: unrealistic evaluation, no systematic comparison, and no consideration of edge deployment constraints. We introduce a unified benchmark combining discrete-task evaluation on structural and logical anomalies, a novel continuous drift protocol, the first head-to-head comparison of all published CAD methods, and computational efficiency profiling on edge hardware. Our results reveal that existing CAD methods do not consistently outperform traditional approaches with simple experience replay. Thus motivated, we propose DINOSaur, a training-free method combining a frozen DINOv3 backbone with spatially-indexed coreset memory and neighborhood-restricted anomaly scoring. DINOSaur achieves zero forgetting by construction, outperforms all evaluated methods across all five protocols, and runs at sub-100,ms inference on an NVIDIA Jetson Orin Nano, with on-device adaptation to new tasks in under 30 seconds.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 30 of its 39 references that are in the map. First posted on arXiv 22 May 2026.

Builds on

30 of 39 references.

References · 30

  1. 2023Segment Anythingcited by 8,234
  2. 2016Overcoming catastrophic forgetting in neural networkscited by 4,787
  3. 2016Progressive Neural Networkscited by 1,361
  4. 2017Memory Aware Synapses: Learning what (not) to forgetcited by 1,105
  5. 2018Experience Replay for Continual Learningcited by 1,095
  6. 2020Dark Experience for General Continual Learning: a Strong, Simple Baselinecited by 911
  7. 2021Towards Total Recall in Industrial Anomaly Detectioncited by 776
  8. 2020PaDiM: a Patch Distribution Modeling Framework for Anomaly Detection and Localizationcited by 594
  9. 2021DRAEM -- A discriminatively trained reconstruction embedding for surface anomaly detectioncited by 514
  10. 2021Gradient Projection Memory for Continual Learningcited by 348
  11. 2023AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detectioncited by 271
  12. 2022A Unified Model for Multi-class Anomaly Detectioncited by 268
18 more
  1. 2020PANDA: Adapting Pretrained Features for Anomaly Detection and Segmentationcited by 175
  2. 2023AnomalyGPT: Detecting Industrial Anomalies Using Large Vision-Language Modelscited by 171
  3. 2023EfficientAD: Accurate Visual Anomaly Detection at Millisecond-Level Latenciescited by 160
  4. 2023Investigating the Catastrophic Forgetting in Multimodal Large Language Modelscited by 119
  5. 2023Preventing Zero-Shot Transfer Degradation in Continual Learning of Vision-Language Modelscited by 112
  6. 2024Toward Generalist Anomaly Detection via In-context Residual Learning with Few-shot Sample Promptscited by 89
  7. 2022TRGP: Trust Region Gradient Projection for Continual Learningcited by 82
  8. 2023IM-IAD: Industrial Image Anomaly Detection Benchmark in Manufacturingcited by 64
  9. 2022Beyond Not-Forgetting: Continual Learning with Backward Knowledge Transfercited by 59
  10. 2024Unsupervised Continual Anomaly Detection with Contrastively-learned Promptcited by 31
  11. 2023An Incremental Unified Framework for Small Defect Inspectioncited by 25
  12. 2025One-for-More: Continual Diffusion Model for Anomaly Detectioncited by 19
  13. 2024PaSTe: Improving the Efficiency of Visual Anomaly Detection at the Edgecited by 16
  14. 2024Unveiling the Anomalies in an Ever-Changing World: A Benchmark for Pixel-Level Anomaly Detection in Continual Learningcited by 13
  15. 2023Lifelong Continual Learning for Anomaly Detection: New Challenges, Perspectives, and Insightscited by 8
  16. 2025Memory Efficient Continual Learning for Edge-Based Visual Anomaly Detectioncited by 4
  17. 2025Continual-MEGA: A Large-scale Benchmark for Generalizable Continual Anomaly Detectioncited by 4
  18. 2025Towards Continual Visual Anomaly Detection in the Medical Domaincited by 3

Built on by

Not cited yet.