Robust Anomaly Detection in Industrial Environments via Meta-Learning

2025-08-25

AI Summary

In industrial anomaly detection, severe label noise during training critically degrades the performance of existing methods, yet robust solutions remain unexplored. This paper proposes a robust anomaly detection framework integrating Normalizing Flows with Model-Agnostic Meta-Learning (MAML). We introduce an uncertainty-quantification-driven L2 adaptive regularization mechanism that dynamically adjusts regularization strength in a noise-aware manner. Coupled with multi-scale feature extraction and precise likelihood modeling, our approach enhances discriminability for subtle anomalies under high label noise. On MVTec-AD and KSDD2, our method achieves I-AUROC scores of 95.4% and 94.6%, respectively. Remarkably, even with 50% label corruption, it retains strong performance—86.8% on MVTec-AD and 92.1% on KSDD2—significantly outperforming state-of-the-art methods. The key contributions include: (i) the first integration of meta-learning with flow-based modeling for robust anomaly detection; (ii) an uncertainty-guided adaptive regularization strategy; and (iii) superior generalization under extreme label noise.

Technology Category

• Machine Learning: Adversarial Learning & Robustness
• Computer Vision: Adversarial Attacks & Robustness
• Data Mining & Knowledge Management: Anomaly/Outlier Detection

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Anomaly detection is fundamental for ensuring quality control and operational efficiency in industrial environments, yet conventional approaches face significant challenges when training data contains mislabeled samples-a common occurrence in real-world scenarios. This paper presents RAD, a robust anomaly detection framework that integrates Normalizing Flows with Model-Agnostic Meta-Learning to address the critical challenge of label noise in industrial settings. Our approach employs a bi-level optimization strategy where meta-learning enables rapid adaptation to varying noise conditions, while uncertainty quantification guides adaptive L2 regularization to maintain model stability. The framework incorporates multiscale feature processing through pretrained feature extractors and leverages the precise likelihood estimation capabilities of Normalizing Flows for robust anomaly scoring. Comprehensive evaluation on MVTec-AD and KSDD2 datasets demonstrates superior performance, achieving I-AUROC scores of 95.4% and 94.6% respectively under clean conditions, while maintaining robust detection capabilities above 86.8% and 92.1% even when 50% of training samples are mislabeled. The results highlight RAD's exceptional resilience to noisy training conditions and its ability to detect subtle anomalies across diverse industrial scenarios, making it a practical solution for real-world anomaly detection applications where perfect data curation is challenging.

Problem

• Addresses label noise in industrial anomaly detection training data
• Enhances model robustness against mislabeled samples via meta-learning
• Improves anomaly detection accuracy under noisy real-world conditions

Innovation

• Meta-learning with Normalizing Flows integration
• Bi-level optimization for noise adaptation
• Uncertainty-guided adaptive regularization technique