This paper introduces a novel approach to leverage the generalizability capability of Diffusion Models for Source-Free Domain Adaptation (DM-SFDA). Our proposed DM-SFDA method involves fine-tuning a pre-trained text-to-image diffusion model to generate source domain images using features from the target images to guide the diffusion process. Specifically, the pre-trained diffusion model is fine-tuned to generate source samples that minimize entropy and maximize confidence for the pre-trained source model. We then apply established unsupervised domain adaptation techniques to align the generated source images with target domain data. We validate our approach through comprehensive experiments across a range of datasets, including Office-31, Office-Home, and VisDA. The results highlight significant improvements in SFDA performance, showcasing the potential of diffusion models in generating contextually relevant, domain-specific images.

翻译：本文提出了一种利用扩散模型泛化能力实现无源域适应（DM-SFDA）的新方法。我们提出的DM-SFDA方法包括微调预训练的文本到图像扩散模型，利用目标图像的特征引导扩散过程，生成源域图像。具体而言，我们微调预训练扩散模型，使其生成能使预训练源模型熵最小化且置信度最大化的源样本。随后，我们采用成熟的无监督域适应技术，将生成的源图像与目标域数据对齐。通过在Office-31、Office-Home和VisDA等多个数据集上的综合实验验证了所提方法的有效性。实验结果凸显了扩散模型在生成上下文相关域特异性图像方面的潜力，显著提升了SFDA的性能。