Generating and Reweighting Dense Contrastive Patterns for Unsupervised Anomaly Detection

TL;DR

This paper introduces GRAD, a novel unsupervised anomaly detection framework that generates contrastive patterns without prior assumptions, using a diffusion model and self-supervised reweighting to improve detection accuracy and speed.

Contribution

The paper proposes a prior-less anomaly generation paradigm and develops GRAD, integrating a diffusion model, reweighting mechanism, and patch-level detector for improved unsupervised anomaly detection.

Findings

GRAD achieves competitive accuracy on MVTec datasets.

GRAD demonstrates superior inference speed.

PatchDiff effectively exposes various anomaly types.

Abstract

Recent unsupervised anomaly detection methods often rely on feature extractors pretrained with auxiliary datasets or on well-crafted anomaly-simulated samples. However, this might limit their adaptability to an increasing set of anomaly detection tasks due to the priors in the selection of auxiliary datasets or the strategy of anomaly simulation. To tackle this challenge, we first introduce a prior-less anomaly generation paradigm and subsequently develop an innovative unsupervised anomaly detection framework named GRAD, grounded in this paradigm. GRAD comprises three essential components: (1) a diffusion model (PatchDiff) to generate contrastive patterns by preserving the local structures while disregarding the global structures present in normal images, (2) a self-supervised reweighting mechanism to handle the challenge of long-tailed and unlabeled contrastive patterns generated by PatchDiff, and (3) a lightweight patch-level detector to efficiently distinguish the normal patterns and reweighted contrastive patterns. The generation results of PatchDiff effectively expose various types of anomaly patterns, e.g. structural and logical anomaly patterns. In addition, extensive experiments on both MVTec AD and MVTec LOCO datasets also support the aforementioned observation and demonstrate that GRAD achieves competitive anomaly detection accuracy and superior inference speed.