Unseen Visual Anomaly Generation

TL;DR

This paper introduces Anomaly Anything, a framework using Stable Diffusion to generate realistic unseen anomalies from normal samples, improving anomaly detection by creating diverse synthetic anomalies with minimal data.

Contribution

The work presents a novel method leveraging Stable Diffusion conditioned on a single normal sample to generate diverse, realistic anomalies for various object types, with attention-guided and prompt-guided refinement.

Findings

Effective generation of high-quality unseen anomalies.

Enhanced anomaly detection performance on benchmark datasets.

Ability to generate diverse anomalies from minimal input data.

Abstract

Visual anomaly detection (AD) presents significant challenges due to the scarcity of anomalous data samples. While numerous works have been proposed to synthesize anomalous samples, these synthetic anomalies often lack authenticity or require extensive training data, limiting their applicability in real-world scenarios. In this work, we propose Anomaly Anything (AnomalyAny), a novel framework that leverages Stable Diffusion (SD)'s image generation capabilities to generate diverse and realistic unseen anomalies. By conditioning on a single normal sample during test time, AnomalyAny is able to generate unseen anomalies for arbitrary object types with text descriptions. Within AnomalyAny, we propose attention-guided anomaly optimization to direct SD attention on generating hard anomaly concepts. Additionally, we introduce prompt-guided anomaly refinement, incorporating detailed descriptions to further improve the generation quality. Extensive experiments on MVTec AD and VisA datasets demonstrate AnomalyAny's ability in generating high-quality unseen anomalies and its effectiveness in enhancing downstream AD performance.