A Hierarchical Transformation-Discriminating Generative Model for Few Shot Anomaly Detection
This work addresses the problem of detecting anomalies with limited training data for applications in image analysis and defect detection, representing an incremental improvement over existing methods.
The paper tackled few-shot anomaly detection in images by developing a hierarchical generative model that uses image transformations and patch-discriminators to capture multi-scale distributions, achieving superior performance over recent baselines on datasets like Paris, CIFAR10, MNIST, FashionMNIST, and MVTec.
Anomaly detection, the task of identifying unusual samples in data, often relies on a large set of training samples. In this work, we consider the setting of few-shot anomaly detection in images, where only a few images are given at training. We devise a hierarchical generative model that captures the multi-scale patch distribution of each training image. We further enhance the representation of our model by using image transformations and optimize scale-specific patch-discriminators to distinguish between real and fake patches of the image, as well as between different transformations applied to those patches. The anomaly score is obtained by aggregating the patch-based votes of the correct transformation across scales and image regions. We demonstrate the superiority of our method on both the one-shot and few-shot settings, on the datasets of Paris, CIFAR10, MNIST and FashionMNIST as well as in the setting of defect detection on MVTec. In all cases, our method outperforms the recent baseline methods.