Quantized FCA: Efficient Zero-Shot Texture Anomaly Detection

2025-10-17

AI Summary

To address the poor real-time performance and industrial deployability of existing zero-shot anomaly detection methods for texture images, this paper proposes an efficient unsupervised detection framework. Our method introduces three key innovations: (1) Quantized Feature Correspondence Analysis (QFCA), which accelerates computation by ~10× via histogram-based feature quantization; (2) PCA-driven feature preprocessing to enhance discriminability between normal and anomalous regions; and (3) a unified detection strategy integrating quantized representations, histogram statistics, and dimensionality-reduced contrastive learning for fast, pixel-level localization—without supervision or prior anomaly samples. Evaluated on multiple benchmark datasets, our approach substantially outperforms state-of-the-art methods, achieving real-time inference speed (>30 FPS) while maintaining high accuracy on complex textured surfaces, demonstrating strong potential for industrial deployment.

Technology Category

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

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Zero-shot anomaly localization is a rising field in computer vision research, with important progress in recent years. This work focuses on the problem of detecting and localizing anomalies in textures, where anomalies can be defined as the regions that deviate from the overall statistics, violating the stationarity assumption. The main limitation of existing methods is their high running time, making them impractical for deployment in real-world scenarios, such as assembly line monitoring. We propose a real-time method, named QFCA, which implements a quantized version of the feature correspondence analysis (FCA) algorithm. By carefully adapting the patch statistics comparison to work on histograms of quantized values, we obtain a 10x speedup with little to no loss in accuracy. Moreover, we introduce a feature preprocessing step based on principal component analysis, which enhances the contrast between normal and anomalous features, improving the detection precision on complex textures. Our method is thoroughly evaluated against prior art, comparing favorably with existing methods. Project page: https://reality.tf.fau.de/pub/ardelean2025quantized.html

Problem

• Detecting texture anomalies violating stationarity assumption
• Addressing high computational cost in existing methods
• Enabling real-time deployment for industrial applications

Innovation

• Quantized feature correspondence analysis for speed
• Histogram-based patch statistics comparison method
• PCA preprocessing enhances normal-anomaly feature contrast