Learning Fair Domain Adaptation with Virtual Label Distribution
This addresses fairness issues in domain adaptation for machine learning applications, though it is incremental as it builds on existing UDA methods.
The paper tackles the problem of category fairness in Unsupervised Domain Adaptation (UDA), where classifiers often favor easy categories over difficult ones, and proposes the VILL framework to improve worst-case performance while maintaining high overall accuracy.
Unsupervised Domain Adaptation (UDA) aims to mitigate performance degradation when training and testing data are sampled from different distributions. While significant progress has been made in enhancing overall accuracy, most existing methods overlook performance disparities across categories-an issue we refer to as category fairness. Our empirical analysis reveals that UDA classifiers tend to favor certain easy categories while neglecting difficult ones. To address this, we propose Virtual Label-distribution-aware Learning (VILL), a simple yet effective framework designed to improve worst-case performance while preserving high overall accuracy. The core of VILL is an adaptive re-weighting strategy that amplifies the influence of hard-to-classify categories. Furthermore, we introduce a KL-divergence-based re-balancing strategy, which explicitly adjusts decision boundaries to enhance category fairness. Experiments on commonly used datasets demonstrate that VILL can be seamlessly integrated as a plug-and-play module into existing UDA methods, significantly improving category fairness.