Traditional domain adaptation addresses the task of adapting a model to a novel target domain under limited or no additional supervision. While tackling the input domain gap, the standard domain adaptation settings assume no domain change in the output space. In semantic prediction tasks, different datasets are often labeled according to different semantic taxonomies. In many real-world settings, the target domain task requires a different taxonomy than the one imposed by the source domain. We therefore introduce the more general taxonomy adaptive domain adaptation (TADA) problem, allowing for inconsistent taxonomies between the two domains. We further propose an approach that jointly addresses the image-level and label-level domain adaptation. On the label-level, we employ a bilateral mixed sampling strategy to augment the target domain, and a relabelling method to unify and align the label spaces. We address the image-level domain gap by proposing an uncertainty-rectified contrastive learning method, leading to more domain-invariant and class discriminative features. We extensively evaluate the effectiveness of our framework under different TADA settings: open taxonomy, coarse-to-fine taxonomy, and partially-overlapping taxonomy. Our framework outperforms previous state-of-the-art by a large margin, while capable of adapting to target taxonomies.
翻译:传统领域适应工作涉及在有限或不增加额外监督下将模型适应到一个新目标领域的任务。在解决投入领域差距时,标准领域适应设置假定产出空间没有领域变化。在语义预测任务中,不同的数据集往往根据不同的语义分类加以标签。在许多现实世界环境中,目标领域任务需要一种不同于源领域规定的分类法。因此,我们引入了更普遍的分类适应领域适应性适应(TADA)问题,允许两个领域之间的分类不一致。我们进一步提出了一种共同处理图像水平和标签层面域适应的方法。在标签层面,我们采用双边混合抽样战略来扩大目标领域,并重新标注统一和调整标签空间的方法。我们通过提出一种不确定性和重新分类的对比性学习方法来解决图像层面的差距,从而导致更多的域性差异性和等级歧视性特征。我们广泛评估了我们在不同塔达环境下的框架的有效性:公开的税制、可分析的分类和标签层面域域适应。在先前的税制框架之外,我们采用了一个部分目标化的税制框架。