When Unsupervised Domain Adaptation meets One-class Anomaly Detection: Addressing the Two-fold Unsupervised Curse by Leveraging Anomaly Scarcity

ICLR 2026 Conference Submission18901 Authors

Keywords: Unsupervised Domain Adaptation

TL;DR: The paper introduces an unsupervised domain adaptation framework for anomaly detection by aligning source-domain normal data with the dominant cluster of the unlabeled target domain to address domain shift in the context of one-class classification.

Abstract: This paper introduces the first fully unsupervised domain adaptation (UDA) framework for unsupervised anomaly detection (UAD). The performance of UAD techniques degrades significantly in the presence of a domain shift, difficult to avoid in a real-world setting. While UDA has contributed to solving this issue in binary and multi-class classification, such a strategy is ill-posed in one-class UAD. This might be explained by the unsupervised nature of the two tasks, namely, domain adaptation and anomaly detection. Herein, we first formulate this problem that we call the two-fold unsupervised curse. Then, we propose a pioneering solution to this curse, considered intractable so far, by assuming that anomalies are rare. Specifically, we leverage clustering techniques to identify a dominant cluster in the target feature space. Posed as the normal cluster, the latter is aligned with the source normal features. Specifically, given a one-class source set and an unlabeled target set composed primarily of normal data and some anomalies, we fit the source features within a hypersphere while jointly aligning them with the features of the dominant cluster in the target set. The paper provides extensive experiments and analysis on common domain adaptation benchmarks, adapted to the one-class anomaly detection setting, demonstrating the relevance of both the newly introduced paradigm and the proposed approach. The code will be made publicly available.

Primary Area: transfer learning, meta learning, and lifelong learning

Submission Number: 18901