Attention Guided Anomaly Localization in Images

TL;DR

This paper introduces CAVGA, a novel unsupervised and weakly-supervised method for anomaly localization in images that does not require anomalous training data and leverages guided attention to improve localization accuracy.

Contribution

The paper proposes CAVGA, a convolutional adversarial variational autoencoder with guided attention, for effective anomaly localization without needing anomalous training images.

Findings

CAVGA outperforms state-of-the-art methods on multiple datasets.

It achieves high localization accuracy in unsupervised and weakly-supervised settings.

CAVGA is effective with only 2% anomalous images in training.

Abstract

Anomaly localization is an important problem in computer vision which involves localizing anomalous regions within images with applications in industrial inspection, surveillance, and medical imaging. This task is challenging due to the small sample size and pixel coverage of the anomaly in real-world scenarios. Most prior works need to use anomalous training images to compute a class-specific threshold to localize anomalies. Without the need of anomalous training images, we propose Convolutional Adversarial Variational autoencoder with Guided Attention (CAVGA), which localizes the anomaly with a convolutional latent variable to preserve the spatial information. In the unsupervised setting, we propose an attention expansion loss where we encourage CAVGA to focus on all normal regions in the image. Furthermore, in the weakly-supervised setting we propose a complementary guided attention loss, where we encourage the attention map to focus on all normal regions while minimizing the attention map corresponding to anomalous regions in the image. CAVGA outperforms the state-of-the-art (SOTA) anomaly localization methods on MVTec Anomaly Detection (MVTAD), modified ShanghaiTech Campus (mSTC) and Large-scale Attention based Glaucoma (LAG) datasets in the unsupervised setting and when using only 2% anomalous images in the weakly-supervised setting. CAVGA also outperforms SOTA anomaly detection methods on the MNIST, CIFAR-10, Fashion-MNIST, MVTAD, mSTC and LAG datasets.