COFT-AD: COntrastive Fine-Tuning for Few-Shot Anomaly Detection

Page view(s)
22
Checked on Mar 19, 2025
COFT-AD: COntrastive Fine-Tuning for Few-Shot Anomaly Detection
Title:
COFT-AD: COntrastive Fine-Tuning for Few-Shot Anomaly Detection
Journal Title:
IEEE Transactions on Image Processing
Keywords:
Publication Date:
12 March 2024
Citation:
Liao, J., Xu, X., Nguyen, M. C., Goodge, A., & Foo, C. S. (2024). COFT-AD: COntrastive Fine-Tuning for Few-Shot Anomaly Detection. IEEE Transactions on Image Processing, 33, 2090–2103. https://doi.org/10.1109/tip.2024.3374048
Abstract:
Existing approaches towards anomaly detection (AD) often rely on a substantial amount of anomaly-free data to train representation and density models. However, large anomaly-free datasets may not always be available before the inference stage; in which case an anomaly detection model must be trained with only a handful of normal samples, a.k.a. fewshot anomaly detection (FSAD). In this paper, we propose a novel methodology to address the challenge of FSAD which incorporates two important techniques. Firstly, we employ a model pre-trained on a large source dataset to initialize model weights. Secondly, to ameliorate the covariate shift between source and target domains, we adopt contrastive training to fine-tune on the few-shot target domain data. To learn suitable representations for the downstream AD task, we additionally incorporate cross-instance positive pairs to encourage a tight cluster of the normal samples, and negative pairs for better separation between normal and synthesized negative samples. We evaluate few-shot anomaly detection on on 3 controlled AD tasks and 4 real-world AD tasks to demonstrate the effectiveness of the proposed method.
License type:
Publisher Copyright
Funding Info:
This research / project is supported by the Agency for Science, Technology and Research - AME Programmatic Funds
Grant Reference no. : A20H6b0151
Description:
© 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
ISSN:
1941-0042
1057-7149
Files uploaded:

File Size Format Action
tip-29729-2023-proof-hi.pdf 4.39 MB PDF Request a copy