BriefGPT.xyz
Jul, 2022
无需索引中间域的渐进式领域自适应
Gradual Domain Adaptation without Indexed Intermediate Domains
HTML
PDF
Hong-You Chen, Wei-Lun Chao
TL;DR
本研究提出了一种新的逐步领域自适应方法,称为Intermediate DOmain Labeler(IDOL),应用了粗到细的框架,通过逐步领域鉴别器训练和循环一致性损失函数,发现中间的领域序列,并提升了自适应性能。
Abstract
The effectiveness of
unsupervised domain adaptation
degrades when there is a large discrepancy between the source and target domains.
gradual domain adaptation
(GDA) is one promising way to mitigate such an issue
→