Unsupervised anomaly detection methods are widely used in industrial anomaly detection, primarily since there is a lack of anomalous data available for training. However, these methods still struggle to effectively detect and localize anomalies due to the diverse types of anomalies and frequent variations in their sizes and shapes. To address this challenge, we propose a new flow-based framework called Two-Hierarchy Normalizing Flow (THF) for anomaly detection and localization. THF consists of two main components: a masked flow and a constant flow. In the masked flow, we improve the traditional convolutional kernel by masking partial regions of the kernel to prevent excessive minimization of negative log-likelihood. This enhancement greatly helps in effectively detecting and localizing anomalies. The constant flow incorporates dual units and a volume-preserving flow module. The dual units consist of a characterization unit and a mixture unit. The characterization unit accurately captures both local and global feature information, while the mixture unit learns neighboring feature representations. Unlike traditional flow-based methods, the volume-preserving flow module maintains the volume of the probability distribution invariable, facilitating precise detection and localization of high-scoring anomalies. Extensive experiments conducted on multiple widely used anomaly detection datasets demonstrate that THF outperforms state-of-the-art methods in both anomaly detection and localization.

With production growth and improvements in production speed in modern industries, accurate anomaly detection is becoming increasingly important to improve quality inspection efficiency and help minimize production costs. Existing reconstruction-based methods achieve promising anomaly detection results in some scenarios. However, when large-scale anomalies exist, the generalization ability of these methods is limited, and it is difficult to reconstruct the anomalies effectively into normal areas, which may lead to unsatisfactory results. To address this issue, this paper proposes a novel network architecture called SeRNet, which comprises three components: a segmentation sub-network, a reconstruction sub-network, and a repair module. SeRNet addresses the challenge of large-scale anomaly reconstruction by utilizing the segmentation sub-network for pre-segmentation, the repair module for repairing the large-scale anomalies using normal images similar to the test images, and the reconstruction sub-network for processing small-scale anomalies and anomalies in the repaired splices. Additionally, this paper introduces two methods for generating pseudo-anomalies at different scales. SeRNet leverages the advantages of both the segmentation sub-network, which is effective at segmenting large-scale anomalies, and the reconstruction sub-network, which can effectively reconstruct small-scale anomalies. Experiments on the MVTec AD industrial dataset demonstrate that SeRNet delivers outstanding performance, achieving an image-level AUROC score of 99.6.