PromptAD: Learning Prompts with only Normal Samples for Few-Shot Anomaly Detection

TL;DR

PromptAD introduces a novel one-class prompt learning approach for few-shot industrial anomaly detection, effectively utilizing normal samples and explicit margin control to outperform existing methods in diverse settings.

Contribution

It proposes a new one-class prompt learning method with semantic concatenation and explicit anomaly margin, addressing challenges in anomaly detection without anomaly images.

Findings

Achieves top performance in 11/12 few-shot settings on MVTec and VisA datasets.

Introduces semantic concatenation to generate negative samples for prompt learning.

Utilizes explicit anomaly margin to improve detection accuracy.

Abstract

The vision-language model has brought great improvement to few-shot industrial anomaly detection, which usually needs to design of hundreds of prompts through prompt engineering. For automated scenarios, we first use conventional prompt learning with many-class paradigm as the baseline to automatically learn prompts but found that it can not work well in one-class anomaly detection. To address the above problem, this paper proposes a one-class prompt learning method for few-shot anomaly detection, termed PromptAD. First, we propose semantic concatenation which can transpose normal prompts into anomaly prompts by concatenating normal prompts with anomaly suffixes, thus constructing a large number of negative samples used to guide prompt learning in one-class setting. Furthermore, to mitigate the training challenge caused by the absence of anomaly images, we introduce the concept of explicit anomaly margin, which is used to explicitly control the margin between normal prompt features and anomaly prompt features through a hyper-parameter. For image-level/pixel-level anomaly detection, PromptAD achieves first place in 11/12 few-shot settings on MVTec and VisA.