In the pursuit of transferring a source model to a target domain without
access to the source training data, source-free domain adaptation (SFDA) has
been extensively explored across various scenarios, including closed-set,
open-set, partial-set, and generalized settings. Existing meth
本文提出了基于源分布估计方法的 Source-Free Domain Adaptation 模型 SFDA-DE,采用球形 k 均值聚类生成目标数据的假伪标签,并利用目标数据和锚点学习源域的类条件特征分布,通过最小化交叉适应损失函数来对齐两个数据域,在多个 DA 基准测试中实现了最先进的性能表现,并且优于需要大量源数据的传统 DA 方法。