Recent works of multi-source domain adaptation focus on learning a domain-agnostic model, of which the parameters are static. However, such a static model is difficult to handle conflicts across multiple domains, and suffers from a performance degradation in both source domains and target domain. In this paper, we present dynamic transfer to address domain conflicts, where the model parameters are adapted to samples. The key insight is that adapting model across domains is achieved via adapting model across samples. Thus, it breaks down source domain barriers and turns multi-source domains into a single-source domain. This also simplifies the alignment between source and target domains, as it only requires the target domain to be aligned with any part of the union of source domains. Furthermore, we find dynamic transfer can be simply modeled by aggregating residual matrices and a static convolution matrix. Experimental results show that, without using domain labels, our dynamic transfer outperforms the state-of-the-art method by more than 3% on the large multi-source domain adaptation datasets -- DomainNet. Source code is at https://github.com/liyunsheng13/DRT.
翻译:多个源域适应的近期工作侧重于学习一个域-不可知模型,其参数是静态的。然而,这种静态模型难以处理多个域的冲突,且在源域和目标域中都存在性能退化。在本文中,我们提出动态转移,以解决域冲突,模型参数适合样本。关键的洞察力是,通过对不同样本的模型进行调整,实现跨域的适应模式。因此,它打破了源域屏障,将多源域变成一个单一源域。这也简化了源域和目标域之间的对齐,因为它只要求目标域与源域联盟的任何部分保持一致。此外,我们发现动态转移可以通过汇总剩余矩阵和一个静态共变矩阵进行简单的建模。实验结果表明,不使用域标签,我们的动态转移在大型多源域适应数据集 -- DomainNet 上超过了3%。源代码在 https://github.com/liyunsheng13/DRT 上 https://github.com/linsheng13/DRT。