FAPM: Fast Adaptive Patch Memory for Real-time Industrial Anomaly Detection

TL;DR

FAPM introduces a fast, adaptive patch memory approach for real-time industrial anomaly detection, significantly improving speed while maintaining high accuracy by utilizing patch-wise and layer-wise memory banks.

Contribution

The paper proposes FAPM, a novel method that employs patch-wise and layer-wise memory banks with adaptive sampling to enable real-time anomaly detection.

Findings

FAPM achieves superior speed compared to existing methods.

FAPM maintains high detection accuracy.

FAPM outperforms state-of-the-art in real-time industrial anomaly detection.

Abstract

Feature embedding-based methods have shown exceptional performance in detecting industrial anomalies by comparing features of target images with normal images. However, some methods do not meet the speed requirements of real-time inference, which is crucial for real-world applications. To address this issue, we propose a new method called Fast Adaptive Patch Memory (FAPM) for real-time industrial anomaly detection. FAPM utilizes patch-wise and layer-wise memory banks that store the embedding features of images at the patch and layer level, respectively, which eliminates unnecessary repetitive computations. We also propose patch-wise adaptive coreset sampling for faster and more accurate detection. FAPM performs well in both accuracy and speed compared to other state-of-the-art methods