Feature Consistency Learning for Anomaly Detection

Anomaly detection in industrial image is a challenging automated vision inspection task. Given an input image, it is essential to know not only whether or not it is abnormal, but also to locate the anomaly. Currently, knowledge distillation-based teacher-student networks have nearly saturated image-level anomaly detection on publicly available datasets, however, pixel-level anomaly localization is still very tough. To address this problem, we propose a spatial neighbouring coding (SNC) module that facilitates the localization of anomalies by encoding the contextual information of the neighbourhood of each element in feature space. Our SNC can be easily plugged into diverse teacher-student network-based anomaly detectors for improving their performance. Subsequently, we propose a feature consistency learning (FCL) method that learns low-level and high-level feature consistencies in a unified framework for anomaly detection task. The proposed FCL is capable of achieving more accurate anomaly localization by imposing consistency constraints on the features extracted from the teacher-student network. Experimental results on benchmark datasets for industrial image anomaly detection show that our FCL method achieves image-level AUROC of 99.1% and 98.4%, pixel-level AUROC of 97.7% and 99.0%, and pixel-level AUPRO of 97.4% and 95.5% on MVTec AD and VisA datasets respectively, demonstrating the effectiveness of our FCL.