Hedegaard Lukas, Sheikh-Omar Omar Ali, Iosifidis Alexandros
IEEE Trans Image Process. 2021;30:8619-8631. doi: 10.1109/TIP.2021.3118978. Epub 2021 Oct 20.
Domain Adaptation is the process of alleviating distribution gaps between data from different domains. In this paper, we show that Domain Adaptation methods using pair-wise relationships between source and target domain data can be formulated as a Graph Embedding in which the domain labels are incorporated into the structure of the intrinsic and penalty graphs. Specifically, we analyse the loss functions of three existing state-of-the-art Supervised Domain Adaptation methods and demonstrate that they perform Graph Embedding. Moreover, we highlight some generalisation and reproducibility issues related to the experimental setup commonly used to demonstrate the few-shot learning capabilities of these methods. To assess and compare Supervised Domain Adaptation methods accurately, we propose a rectified evaluation protocol, and report updated benchmarks on the standard datasets Office31 (Amazon, DSLR, and Webcam), Digits (MNIST, USPS, SVHN, and MNIST-M) and VisDA (Synthetic, Real).
域适应是缓解来自不同域的数据之间分布差距的过程。在本文中，我们表明，使用源域和目标域数据之间成对关系的域适应方法可以被表述为一种图嵌入，其中域标签被纳入内在图和惩罚图的结构中。具体而言，我们分析了三种现有的先进监督域适应方法的损失函数，并证明它们执行图嵌入。此外，我们强调了一些与通常用于展示这些方法的少样本学习能力的实验设置相关的泛化和可重复性问题。为了准确评估和比较监督域适应方法，我们提出了一种修正的评估协议，并报告了在标准数据集Office31（亚马逊、数码单反相机和网络摄像头）、Digits（MNIST、USPS、SVHN和MNIST-M）和VisDA（合成、真实）上更新的基准。
