2DSig-Detect : a semi-supervised framework for anomaly detection on image data using 2D-signatures

Tools

Xie, Xinheng and Yamaguchi, Kureha and Leblanc, Margaux and Malzard, Simon and Chhabra, Varun and Nockles, Victoria and Wu, Yue (2026) 2DSig-Detect : a semi-supervised framework for anomaly detection on image data using 2D-signatures. Pattern Recognition, 180 (Pt. B). 114172. ISSN 0031-3203 (https://doi.org/10.1016/j.patcog.2026.114172)

[thumbnail of Xie-etal-PR-2026-2DSig-Detect-a-semi-supervised-framework-for-anomaly-detection-on-image-data]
Preview
 Text. Filename: Xie-etal-PR-2026-2DSig-Detect-a-semi-supervised-framework-for-anomaly-detection-on-image-data.pdf
Final Published Version
License: Creative Commons Attribution 4.0 logo
Download (3MB)| Preview

Abstract

The rapid and widespread deployment of AI within critical systems raises serious questions about their security against adversarial attacks. Models performing image-related tasks, are vulnerable to integrity violations causing misclassifications that do not compromise normal system operation, but rather, produce undesirable outcomes for targeted inputs. This paper introduces a novel technique for anomaly detection in images called 2DSig-Detect, which is a 2D-signature-embedded semi-supervised framework rooted in rough path theory. We demonstrate that 2D-signatures can be applied to detect both adversarial examples and backdoored data, mitigating against test-time and training-time integrity attacks. Our results demonstrate both the efficacy and superior computational efficiency of our 2D-signature method for detecting adversarial manipulations to the training and test data.

ORCID iDs

Xie, Xinheng, Yamaguchi, Kureha, Leblanc, Margaux, Malzard, Simon, Chhabra, Varun, Nockles, Victoria and Wu, Yue ORCID logoORCID: https://orcid.org/0000-0002-6281-2229;
CORE (COnnecting REpositories)