SubspaceAD: Training-Free Few-Shot Anomaly Detection via Subspace Modeling

TL;DR

SubspaceAD is a simple, training-free method for few-shot anomaly detection that uses PCA on features from foundation models to identify anomalies, achieving state-of-the-art results.

Contribution

It introduces a novel, training-free approach that leverages PCA on frozen foundation model features for effective anomaly detection in industrial images.

Findings

Achieves 97.1% image AUROC on MVTec-AD with one-shot setting.

Outperforms prior methods in few-shot anomaly detection.

Operates without training, memory banks, or prompt tuning.

Abstract

Detecting visual anomalies in industrial inspection often requires training with only a few normal images per category. Recent few-shot methods achieve strong results employing foundation-model features, but typically rely on memory banks, auxiliary datasets, or multi-modal tuning of vision-language models. We therefore question whether such complexity is necessary given the feature representations of vision foundation models. To answer this question, we introduce SubspaceAD, a training-free method, that operates in two simple stages. First, patch-level features are extracted from a small set of normal images by a frozen DINOv2 backbone. Second, a Principal Component Analysis (PCA) model is fit to these features to estimate the low-dimensional subspace of normal variations. At inference, anomalies are detected via the reconstruction residual with respect to this subspace, producing interpretable and statistically grounded anomaly scores. Despite its simplicity, SubspaceAD achieves state-of-the-art performance across one-shot and few-shot settings without training, prompt tuning, or memory banks. In the one-shot anomaly detection setting, SubspaceAD achieves image-level and pixel-level AUROC of 97.1% and 97.5% on the MVTec-AD dataset, and 93.2% and 98.2% on the VisA dataset, respectively, surpassing prior state-of-the-art results. Code and demo are available at https://github.com/CLendering/SubspaceAD.