无监督异常检测模型的鲁棒性基准

Robustness benchmark for unsupervised anomaly detection models

由于生产环境的复杂性和多样性,了解无监督异常检测模型对常见降质的鲁棒性是至关重要的。为了系统地探索这个问题,我们提出一个名为MVTec-C的数据集来评估无监督异常检测模型的鲁棒性。基于这个数据集,我们探索了五种不同范式的方法的鲁棒性,包括基于重建的、基于表征相似度的、基于归一化流的、基于自监督表征学习的和基于知识蒸馏的范式。此外,我们还探讨了两种最佳的方法中不同模块对鲁棒性和准确性的影响,包括Patch Core方法中的多尺度特征、邻域大小、采样比例和Reverse Distillation方法中的多尺度特征、MMF模块与OCE模块、多尺度蒸馏。最后,我们提出了一个特征对齐模块(FAM),以减少降质带来的特征偏移,并将Patch Core和FAM结合起来,得到一个同时具备高准确率和高鲁棒性的模型。我们希望这项工作能够作为一种鲁棒性评估手段,并在将来为构建鲁棒的异常检测模型提供经验。

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 dataset named MVTec-C to evaluate the robustness of unsupervised anomaly detection models.Based on this dataset,we explore the robustness of approaches in five paradigms,namely,reconstruction-based,representation similarity-based,normalizing flow-based,self-supervised representation learning-based,and knowledge distillation-based paradigms.Furthermore,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 evaluation method and provide experience in building robust anomaly detection models in the future.