Attribute Restoration Framework for Anomaly Detection
This addresses anomaly detection in computer vision, offering a novel approach to improve performance on benchmarks like ImageNet and MVTec AD, but it is incremental as it builds on reconstruction-based methods.
The paper tackles the problem of anomaly detection in multimedia by proposing an attribute restoration framework that erases selected attributes from data and learns to restore them, distinguishing anomalies based on restoration errors. It significantly outperforms state-of-the-art methods, achieving a 10.1% increase in AUROC on ImageNet.
With the recent advances in deep neural networks, anomaly detection in multimedia has received much attention in the computer vision community. While reconstruction-based methods have recently shown great promise for anomaly detection, the information equivalence among input and supervision for reconstruction tasks can not effectively force the network to learn semantic feature embeddings. We here propose to break this equivalence by erasing selected attributes from the original data and reformulate it as a restoration task, where the normal and the anomalous data are expected to be distinguishable based on restoration errors. Through forcing the network to restore the original image, the semantic feature embeddings related to the erased attributes are learned by the network. During testing phases, because anomalous data are restored with the attribute learned from the normal data, the restoration error is expected to be large. Extensive experiments have demonstrated that the proposed method significantly outperforms several state-of-the-arts on multiple benchmark datasets, especially on ImageNet, increasing the AUROC of the top-performing baseline by 10.1%. We also evaluate our method on a real-world anomaly detection dataset MVTec AD and a video anomaly detection dataset ShanghaiTech.