Due to the complexity and diversity of production environments,it is essential to understand the robustness of unsupervised anomaly detection models to common corruptions.To explore this issue systematically,we propose a data-set named MVTec-C to evaluate the robustness of unsupervised anomaly detection models.Based on this dataset,we ex-plore the robustness of approaches in five paradigms,namely,reconstruction-based,representation similarity-based,nor-malizing flow-based,self-supervised representation learning-based,and knowledge distillation-based paradigms.Further-more,we explore the impact of different modules within two optimal methods on robustness and accuracy.This includes the multi-scale features,the neighborhood size,and the sampling ratio in the PatchCore method,as well as the multi-scale features,the MMF module,the OCE module,and the multi-scale distillation in the Reverse Distillation method.Finally,we propose a feature alignment module(FAM)to reduce the feature drift caused by corruptions and combine PatchCore and the FAM to obtain a model with both high performance and high accuracy.We hope this work will serve as an evalu-ation method and provide experience in building robust anomaly detection models in the future.
维普
万方数据