DictAS: A Framework for Class-Generalizable Few-Shot Anomaly Segmentation via Dictionary Lookup

TL;DR

DictAS introduces a class-generalizable few-shot anomaly segmentation framework that uses dictionary lookup and self-supervised learning to detect unseen anomalies without retraining, outperforming existing methods.

Contribution

The paper proposes a novel dictionary lookup-based framework for FSAS that generalizes to unseen classes using only normal reference images and self-supervised learning, eliminating the need for retraining.

Findings

Outperforms state-of-the-art FSAS methods on seven datasets.

Effectively detects unseen anomalies with only normal reference images.

Utilizes self-supervised dictionary lookup for anomaly detection.

Abstract

Recent vision-language models (e.g., CLIP) have demonstrated remarkable class-generalizable ability to unseen classes in few-shot anomaly segmentation (FSAS), leveraging supervised prompt learning or fine-tuning on seen classes. However, their cross-category generalization largely depends on prior knowledge of real seen anomaly samples. In this paper, we propose a novel framework, namely DictAS, which enables a unified model to detect visual anomalies in unseen object categories without any retraining on the target data, only employing a few normal reference images as visual prompts. The insight behind DictAS is to transfer dictionary lookup capabilities to the FSAS task for unseen classes via self-supervised learning, instead of merely memorizing the normal and abnormal feature patterns from the training set. Specifically, DictAS mainly consists of three components: (1) Dictionary Construction - to simulate the index and content of a real dictionary using features from normal reference images. (2) Dictionary Lookup - to retrieve queried region features from the dictionary via a sparse lookup strategy. When a query feature cannot be retrieved, it is classified as an anomaly. (3) Query Discrimination Regularization - to enhance anomaly discrimination by making abnormal features harder to retrieve from the dictionary. To achieve this, Contrastive Query Constraint and Text Alignment Constraint are further proposed. Extensive experiments on seven public industrial and medical datasets demonstrate that DictAS consistently outperforms state-of-the-art FSAS methods.