BriefGPT.xyz
Jul, 2020
无需源数据的无监督领域自适应
Unsupervised Domain Adaptation in the Absence of Source Data
HTML
PDF
Roshni Sahoo, Divya Shanmugam, John Guttag
TL;DR
该研究提出了一种针对亮度、对比度等自然变化轴的目标域适应方法，只需要无标签的目标数据和源分类器，有效地解决了预训练模型中源数据不可用的问题，并表明其在有限标记数据的情况下胜过微调基线。
Abstract
Current
unsupervised domain adaptation
methods can address many types of
distribution shift
, but they assume data from the source domain is freely available. As the use of
→