PNI : Industrial Anomaly Detection using Position and Neighborhood Information
This work improves anomaly detection for industrial inspection, but it is incremental as it builds on pre-trained networks and non-parametric modeling with specific enhancements.
The paper tackled the problem of industrial anomaly detection by addressing the neglect of position and neighborhood information in existing methods, achieving state-of-the-art performance with 99.56% AUROC for detection and 98.98% for localization on the MVTec AD dataset.
Because anomalous samples cannot be used for training, many anomaly detection and localization methods use pre-trained networks and non-parametric modeling to estimate encoded feature distribution. However, these methods neglect the impact of position and neighborhood information on the distribution of normal features. To overcome this, we propose a new algorithm, \textbf{PNI}, which estimates the normal distribution using conditional probability given neighborhood features, modeled with a multi-layer perceptron network. Moreover, position information is utilized by creating a histogram of representative features at each position. Instead of simply resizing the anomaly map, the proposed method employs an additional refine network trained on synthetic anomaly images to better interpolate and account for the shape and edge of the input image. We conducted experiments on the MVTec AD benchmark dataset and achieved state-of-the-art performance, with \textbf{99.56\%} and \textbf{98.98\%} AUROC scores in anomaly detection and localization, respectively.