Uncertainty-guided Open-Set Source-Free Unsupervised Domain Adaptation with Target-private Class Segregation
This addresses a challenging real-world scenario in domain adaptation for machine learning applications where source data access is restricted and target classes are unknown, though it is incremental in improving existing methods.
The paper tackles the problem of source-free open-set domain adaptation, where target data includes unknown classes and source data is unavailable, by proposing a method that segregates target-private samples into multiple unknown classes using uncertainty-guided pseudo-label refinement and a novel contrastive loss. The result is state-of-the-art performance on benchmark datasets, with the method also enabling novel class discovery.
Standard Unsupervised Domain Adaptation (UDA) aims to transfer knowledge from a labeled source domain to an unlabeled target but usually requires simultaneous access to both source and target data. Moreover, UDA approaches commonly assume that source and target domains share the same labels space. Yet, these two assumptions are hardly satisfied in real-world scenarios. This paper considers the more challenging Source-Free Open-set Domain Adaptation (SF-OSDA) setting, where both assumptions are dropped. We propose a novel approach for SF-OSDA that exploits the granularity of target-private categories by segregating their samples into multiple unknown classes. Starting from an initial clustering-based assignment, our method progressively improves the segregation of target-private samples by refining their pseudo-labels with the guide of an uncertainty-based sample selection module. Additionally, we propose a novel contrastive loss, named NL-InfoNCELoss, that, integrating negative learning into self-supervised contrastive learning, enhances the model robustness to noisy pseudo-labels. Extensive experiments on benchmark datasets demonstrate the superiority of the proposed method over existing approaches, establishing new state-of-the-art performance. Notably, additional analyses show that our method is able to learn the underlying semantics of novel classes, opening the possibility to perform novel class discovery.