Source-Free Domain Adaptation (SFDA) aims to adapt a pre-trained source model to an unlabeled target domain without access to source data. Recent advances in Foundation Models (FMs) have introduced new opportunities for leveraging external semantic knowledge to guide SFDA. However, relying on a single FM is often insufficient, as it tends to bias adaptation toward a restricted semantic coverage, failing to capture diverse contextual cues under domain shift. To overcome this limitation, we propose a Collaborative Multi-foundation Adaptation (CoMA) framework that jointly leverages two different FMs (e.g., CLIP and BLIP) with complementary properties to capture both global semantics and local contextual cues. Specifically, we employ a bidirectional adaptation mechanism that (1) aligns different FMs with the target model for task adaptation while maintaining their semantic distinctiveness, and (2) transfers complementary knowledge from the FMs to the target model. To ensure stable adaptation under mini-batch training, we introduce Decomposed Mutual Information (DMI) that selectively enhances true dependencies while suppressing false dependencies arising from incomplete class coverage. Extensive experiments demonstrate that our method consistently outperforms existing state-of-the-art SFDA methods across four benchmarks, including Office-31, Office-Home, DomainNet-126, and VisDA, under the closed-set setting, while also achieving best results on partial-set and open-set variants.

翻译：无源域适应（SFDA）旨在将预训练的源模型适配到无标注的目标域，且无需访问源数据。基础模型（FMs）的最新进展为利用外部语义知识指导SFDA提供了新机遇。然而，依赖单一FM往往不足，因其易使适配偏向受限的语义覆盖，难以捕捉域偏移下的多样化上下文线索。为克服此局限，我们提出协同多基础适配（CoMA）框架，联合利用两个具有互补特性的不同FMs（例如CLIP和BLIP），以同时捕获全局语义和局部上下文线索。具体而言，我们采用双向适配机制：（1）将不同FMs与目标模型对齐以进行任务适配，同时保持其语义独特性；（2）将互补知识从FMs迁移至目标模型。为确保小批量训练下的稳定适配，我们引入分解互信息（DMI），选择性增强真实依赖关系，同时抑制因类别覆盖不全而产生的虚假依赖。大量实验表明，在闭集设置下，我们的方法在四个基准数据集（包括Office-31、Office-Home、DomainNet-126和VisDA）上均持续优于现有最先进的SFDA方法，同时在部分集和开集变体上也取得了最佳结果。