In unsupervised domain adaptation (UDA), a classifier for the target domain is trained with massive true-label data from the source domain and unlabeled data from the target domain. However, collecting fully-true-label data in the source domain is high-cost and sometimes impossible. Compared to the true labels, a complementary label specifies a class that a pattern does not belong to, hence collecting complementary labels would be less laborious than collecting true labels. Thus, in this paper, we propose a novel setting that the source domain is composed of complementary-label data, and a theoretical bound for it is first proved. We consider two cases of this setting, one is that the source domain only contains complementary-label data (completely complementary unsupervised domain adaptation, CC-UDA), and the other is that the source domain has plenty of complementary-label data and a small amount of true-label data (partly complementary unsupervised domain adaptation, PC-UDA). To this end, a complementary label adversarial network} (CLARINET) is proposed to solve CC-UDA and PC-UDA problems. CLARINET maintains two deep networks simultaneously, where one focuses on classifying complementary-label source data and the other takes care of source-to-target distributional adaptation. Experiments show that CLARINET significantly outperforms a series of competent baselines on handwritten-digits-recognition and objects-recognition tasks.
翻译:在未经监督的域适应(UDA)中,目标域的分类员接受来自源域的大量真实标签数据和来自目标域的未贴标签数据的培训。然而,在源域收集完全真实标签数据的成本很高,有时是不可能的。与真实标签相比,一个补充标签指定了一个不属于一个模式的类别,因此,收集补充标签比收集真实标签更不费力。因此,在本文件中,我们提议一个新颖的设置,即源域由补充标签数据组成,并首先证明它的理论约束。我们考虑了这一设置的两个实例,即源域仅包含补充标签数据(完全互补且不受监督的域适应,CC-UDA),而另一个是源域与真正标签数据(部分是辅助性、不受监督的域适应,PC-UDA)。为此,我们提议一个补充标签对抗网络(CLACRINET) 以CC-UDA和PC-UDA 目的的理论约束。CLARINET同时将两个补充性网络的基线来源归类为CLA-CLA-ILA-CLA-IAL listal ladal ladal ladal ladal ladal ladal ladal laut ladal laut ladal ladal ladal laut laut ladal ladal ladal laut ladal ladal ladal laut lades lades ladal lades ladal laut lades laut laut laut laut ladre su lauts lades lauts subildaldal subaldre labal subal labal ladal ladal ladal labal ladal labaldaldaldaldal lades lades su ladal ladal ladal ladal ladal ladal ladal ladal ladal ladal ladal ladal lades lades lades