Transferable Attention for Domain Adaptation
Transferable Attention for Domain Adaptation
复制标题
DOI：
10.1609/aaai.v33i01.33015345
复制
发表时间：
2019-07
期刊：
影响因子：
--
通讯作者：
Ximei Wang;Liang Li;Weirui Ye;Mingsheng Long;Jianmin Wang
中科院分区：
文献类型：
--
作者：
Ximei Wang;Liang Li;Weirui Ye;Mingsheng Long;Jianmin Wang
最近的工作领域适应桥梁不同的领域，通过逆向学习的领域不变的表示，不能区分的领域的竞争。现有的对抗域自适应方法主要是在源域和目标域之间对齐全局图像。然而，很明显，并非图像的所有区域都是可转移的，而强制对齐不可转移的区域可能导致负转移。此外，一些图像在域之间是显著不同的，导致弱的图像级可转移性。为此，我们提出了可转移注意力域适应（TADA），我们的适应模型集中在可转移的区域或图像。我们实现了两种类型的互补可转移注意力：可转移的局部注意力产生的多个区域级域鉴别器突出可转移的区域，和可转移的全局注意力产生的单个图像级域鉴别器突出可转移的图像。大量的实验验证了我们提出的模型超过了标准域自适应数据集上的最新结果。
Recent work in domain adaptation bridges different domains by adversarially learning a domain-invariant representation that cannot be distinguished by a domain discriminator. Existing methods of adversarial domain adaptation mainly align the global images across the source and target domains. However, it is obvious that not all regions of an image are transferable, while forcefully aligning the untransferable regions may lead to negative transfer. Furthermore, some of the images are significantly dissimilar across domains, resulting in weak image-level transferability. To this end, we present Transferable Attention for Domain Adaptation (TADA), focusing our adaptation model on transferable regions or images. We implement two types of complementary transferable attention: transferable local attention generated by multiple region-level domain discriminators to highlight transferable regions, and transferable global attention generated by single image-level domain discriminator to highlight transferable images. Extensive experiments validate that our proposed models exceed state of the art results on standard domain adaptation datasets.