Source-free unsupervised domain adaptation (SF-UDA) aims to adapt a well-trained source model to an unlabeled target domain without access to the source data. One key challenge is the lack of source data during domain adaptation. To handle this, we propose to mine the hidden knowledge
本文探讨了一种实际的领域自适应任务,称为无源域自适应(SFUDA),在此任务中,源预训练模型在没有访问源数据的情况下适应于目标域。我们介绍了一种新的SFUDA范例Divide and Contrast(DaC),使用自适应对比学习框架,通过预测的置信度将目标数据分为类似源域和特定于目标域的样本,并针对每个组别进行调整目标,以在全局和局部层面上提高性能