A Unified Anomaly Synthesis Strategy with Gradient Ascent for Industrial Anomaly Detection and Localization

TL;DR

This paper proposes GLASS, a novel unified framework designed to synthesize a broader coverage of anomalies under the manifold and hypersphere distribution constraints of Global Anomaly Synthesis (GAS) at the feature level and Local Anomaly Synthesis (LAS) at the image level.

Abstract

Anomaly synthesis strategies can effectively enhance unsupervised anomaly detection. However, existing strategies have limitations in the coverage and controllability of anomaly synthesis, particularly for weak defects that are very similar to normal regions. In this paper, we propose Global and Local Anomaly co-Synthesis Strategy (GLASS), a novel unified framework designed to synthesize a broader coverage of anomalies under the manifold and hypersphere distribution constraints of Global Anomaly Synthesis (GAS) at the feature level and Local Anomaly Synthesis (LAS) at the image level. Our method synthesizes near-in-distribution anomalies in a controllable way using Gaussian noise guided by gradient ascent and truncated projection. GLASS achieves state-of-the-art results on the MVTec AD (detection AUROC of 99.9\%), VisA, and MPDD datasets and excels in weak defect detection. The effectiveness and efficiency have been further validated in industrial applications for woven fabric defect detection. The code and dataset are available at: \url{https://github.com/cqylunlun/GLASS}.

Figures (8)

• Figure 1: Process flow and visualization of various anomaly synthesis strategies. (a) Image-level anomaly synthesis strategy (gray triangles) provides detailed textures but lacks diversity. (b) Feature-level anomaly synthesis strategy (pink diamonds) is more efficient but lacks directionality. (c) Our method (blue squares) controls the distribution of synthetic anomalies at image and feature levels by using gradient ascent.
• Figure 2: Schematic of the proposed GLASS. The training phase comprises three branches: (a) Normal branch obtains adapted normal features through a feature extractor and a feature adaptor. (b) GAS branch synthesizes global anomaly features from normal features in three steps based on gradient guidance. (c) LAS branch synthesizes local anomaly images from normal images in three steps based on texture overlay.
• Figure 3: Schematic illustration of Global Anomaly Synthesis (GAS) under different hypotheses. Assume that $r_m$ or $r_h$ represents the $L_{2}$ distance to manifold or hypersphere center, respectively. Green circles ($r_m < r_1$ or $r_h < r'_1$) represent normal features, gray triangles ($r_m > r_2$ or $r'_2 < r_h < r'_3$) represent local anomaly features, pink diamonds represent Gaussian anomaly features obtained by Gaussian noise from normal features, and blue squares ($r_1 < r_m < r_2$ or $r'_1 < r_h < r'_2$) represent global anomaly features obtained by gradient ascent and truncated projection from Gaussian anomaly features.
• Figure 4: Flowchart of Local Anomaly Synthesis (LAS) consisting of three steps: Step I: Anomaly Mask, Step II: Anomaly Texture, and Step III: Overlay Fusion.
• Figure 5: Anomaly score histograms of GLASS-j on each category of MVTec AD.