Imbalanced Open Set Domain Adaptation via Moving-threshold Estimation and Gradual Alignment

TL;DR

This paper introduces OMEGA, a novel method for open set domain adaptation that effectively handles class imbalance and label shift by using moving thresholds and target clustering, significantly improving performance.

Contribution

The paper proposes a new architecture, OMEGA, which addresses label shift and class imbalance in open set domain adaptation through target clustering and adaptive thresholds.

Findings

Outperforms existing OSDA methods on benchmarks

Effectively reduces negative effects of label shift and class imbalance

Demonstrates robustness across IOSDA, OSDA, and OPDA tasks

Abstract

Multimedia applications are often associated with cross-domain knowledge transfer, where Unsupervised Domain Adaptation (UDA) can be used to reduce the domain shifts. Open Set Domain Adaptation (OSDA) aims to transfer knowledge from a well-labeled source domain to an unlabeled target domain under the assumption that the target domain contains unknown classes. Existing OSDA methods consistently lay stress on the covariate shift, ignoring the potential label shift problem. The performance of OSDA methods degrades drastically under intra-domain class imbalance and inter-domain label shift. However, little attention has been paid to this issue in the community. In this paper, the Imbalanced Open Set Domain Adaptation (IOSDA) is explored where the covariate shift, label shift and category mismatch exist simultaneously. To alleviate the negative effects raised by label shift in OSDA, we propose Open-set Moving-threshold Estimation and Gradual Alignment (OMEGA) - a novel architecture that improves existing OSDA methods on class-imbalanced data. Specifically, a novel unknown-aware target clustering scheme is proposed to form tight clusters in the target domain to reduce the negative effects of label shift and intra-domain class imbalance. Furthermore, moving-threshold estimation is designed to generate specific thresholds for each target sample rather than using one for all. Extensive experiments on IOSDA, OSDA and OPDA benchmarks demonstrate that our method could significantly outperform existing state-of-the-arts. Code and data are available at https://github.com/mendicant04/OMEGA.