Context Enhancement with Reconstruction as Sequence for Unified Unsupervised Anomaly Detection

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Yang, Hui-Yue, Chen, Hui, Liu, Lihao, Lin, Zijia, Chen, Kai, Wang, Liejun, Han, Jungong, Ding, Guiguang
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866917772299075584
author Yang, Hui-Yue
Chen, Hui
Liu, Lihao
Lin, Zijia
Chen, Kai
Wang, Liejun
Han, Jungong
Ding, Guiguang
author_facet Yang, Hui-Yue
Chen, Hui
Liu, Lihao
Lin, Zijia
Chen, Kai
Wang, Liejun
Han, Jungong
Ding, Guiguang
contents Unsupervised anomaly detection (AD) aims to train robust detection models using only normal samples, while can generalize well to unseen anomalies. Recent research focuses on a unified unsupervised AD setting in which only one model is trained for all classes, i.e., n-class-one-model paradigm. Feature-reconstruction-based methods achieve state-of-the-art performance in this scenario. However, existing methods often suffer from a lack of sufficient contextual awareness, thereby compromising the quality of the reconstruction. To address this issue, we introduce a novel Reconstruction as Sequence (RAS) method, which enhances the contextual correspondence during feature reconstruction from a sequence modeling perspective. In particular, based on the transformer technique, we integrate a specialized RASFormer block into RAS. This block enables the capture of spatial relationships among different image regions and enhances sequential dependencies throughout the reconstruction process. By incorporating the RASFormer block, our RAS method achieves superior contextual awareness capabilities, leading to remarkable performance. Experimental results show that our RAS significantly outperforms competing methods, well demonstrating the effectiveness and superiority of our method. Our code is available at https://github.com/Nothingtolose9979/RAS.
format Preprint
id arxiv_https___arxiv_org_abs_2409_06285
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Context Enhancement with Reconstruction as Sequence for Unified Unsupervised Anomaly Detection
Yang, Hui-Yue
Chen, Hui
Liu, Lihao
Lin, Zijia
Chen, Kai
Wang, Liejun
Han, Jungong
Ding, Guiguang
Computer Vision and Pattern Recognition
Unsupervised anomaly detection (AD) aims to train robust detection models using only normal samples, while can generalize well to unseen anomalies. Recent research focuses on a unified unsupervised AD setting in which only one model is trained for all classes, i.e., n-class-one-model paradigm. Feature-reconstruction-based methods achieve state-of-the-art performance in this scenario. However, existing methods often suffer from a lack of sufficient contextual awareness, thereby compromising the quality of the reconstruction. To address this issue, we introduce a novel Reconstruction as Sequence (RAS) method, which enhances the contextual correspondence during feature reconstruction from a sequence modeling perspective. In particular, based on the transformer technique, we integrate a specialized RASFormer block into RAS. This block enables the capture of spatial relationships among different image regions and enhances sequential dependencies throughout the reconstruction process. By incorporating the RASFormer block, our RAS method achieves superior contextual awareness capabilities, leading to remarkable performance. Experimental results show that our RAS significantly outperforms competing methods, well demonstrating the effectiveness and superiority of our method. Our code is available at https://github.com/Nothingtolose9979/RAS.
title Context Enhancement with Reconstruction as Sequence for Unified Unsupervised Anomaly Detection
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.06285