UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Brittle Features May Help Anomaly Detection

Mai, KT; Davies, T; Griffin, LD; (2021) Brittle Features May Help Anomaly Detection. In: Proceedings of the 9th Women in Computer Vision workshop at CVPR (2021). Women in Computer Vision: Virtual conference. Green open access

[thumbnail of 2104.10453v1.pdf]
Preview
Text
2104.10453v1.pdf - Accepted Version

Download (2MB) | Preview

Abstract

One-class anomaly detection is challenging. A representation that clearly distinguishes anomalies from normal data is ideal, but arriving at this representation is difficult since only normal data is available at training time. We examine the performance of representations, transferred from auxiliary tasks, for anomaly detection. Our results suggest that the choice of representation is more important than the anomaly detector used with these representations, although knowledge distillation can work better than using the representations directly. In addition, separability between anomalies and normal data is important but not the sole factor for a good representation, as anomaly detection performance is also correlated with more adversarially brittle features in the representation space. Finally, we show our configuration can detect 96.4% of anomalies in a genuine X-ray security dataset, outperforming previous results.

Type: Proceedings paper
Title: Brittle Features May Help Anomaly Detection
Event: 9th Women in Computer Vision workshop at CVPR (2021)
Open access status: An open access version is available from UCL Discovery
Publisher version: https://sites.google.com/view/wicvcvpr2021/home
Language: English
Additional information: This version is the author accepted manuscript. For information on re-use, please refer to the publisher's terms and conditions.
UCL classification: UCL
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Security and Crime Science
URI: https://discovery.ucl.ac.uk/id/eprint/10132480
Downloads since deposit
47Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item