Revisiting Deep Feature Reconstruction for Logical and Structural Industrial Anomaly Detection

TL;DR

This work refine the DFR training objective to improve performance in structural anomaly detection, while introducing an attention-based loss mechanism using a global autoencoder-like network to handle logical anomaly detection.

Abstract

Industrial anomaly detection is crucial for quality control and predictive maintenance, but it presents challenges due to limited training data, diverse anomaly types, and external factors that alter object appearances. Existing methods commonly detect structural anomalies, such as dents and scratches, by leveraging multi-scale features from image patches extracted through deep pre-trained networks. However, significant memory and computational demands often limit their practical application. Additionally, detecting logical anomalies-such as images with missing or excess elements-requires an understanding of spatial relationships that traditional patch-based methods fail to capture. In this work, we address these limitations by focusing on Deep Feature Reconstruction (DFR), a memory- and compute-efficient approach for detecting structural anomalies. We further enhance DFR into a unified framework, called ULSAD, which is capable of detecting both structural and logical anomalies. Specifically, we refine the DFR training objective to improve performance in structural anomaly detection, while introducing an attention-based loss mechanism using a global autoencoder-like network to handle logical anomaly detection. Our empirical evaluation across five benchmark datasets demonstrates the performance of ULSAD in detecting and localizing both structural and logical anomalies, outperforming eight state-of-the-art methods. An extensive ablation study further highlights the contribution of each component to the overall performance improvement. Our code is available at https://github.com/sukanyapatra1997/ULSAD-2024.git

Figures (9)

• Figure 1: Types of anomalies for the categories "breakfast box" and "pushpin" in MVTecLOCO dataset. Two normal samples (Left) along with structural (Middle) and logical anomalies (Right) where the anomalous regions are highlighted in blue and red, respectively.
• Figure 2: Overview of the end-to-end architecture of ULSAD.
• Figure 3: Feature Reconstruction Network
• Figure 4: Global Branch of ULSAD
• Figure 5: (First) Example image belonging to the category "breakfast box" in the MVTecLOCO dataset. (Rest) Visualization of attention maps computed using the intermediate features from a pre-trained model.