References & Citations
Computer Science > Computer Vision and Pattern Recognition
Title: Fine-grained Anomaly Detection via Multi-task Self-Supervision
(Submitted on 20 Apr 2021 (v1), last revised 17 Mar 2022 (this version, v2))
Abstract: Detecting anomalies using deep learning has become a major challenge over the last years, and is becoming increasingly promising in several fields. The introduction of self-supervised learning has greatly helped many methods including anomaly detection where simple geometric transformation recognition tasks are used. However these methods do not perform well on fine-grained problems since they lack finer features. By combining in a multi-task framework high-scale shape features oriented task with low-scale fine features oriented task, our method greatly improves fine-grained anomaly detection. It outperforms state-of-the-art with up to 31% relative error reduction measured with AUROC on various anomaly detection problems.
Submission history
From: Loïc Jézéquel [view email][v1] Tue, 20 Apr 2021 14:19:08 GMT (207kb,D)
[v2] Thu, 17 Mar 2022 09:53:56 GMT (3235kb,D)
Link back to: arXiv, form interface, contact.