标题： 开放集领域自适应的理论界限与深度算法之间的桥梁 显示英文标题

标题： Bridging the Theoretical Bound and Deep Algorithms for Open Set Domain Adaptation

摘要： 在无监督开放集域适应（UOSDA）中，目标域包含源域中未观察到的未知类别。该领域的研究人员旨在训练一个分类器以准确地：1）识别未知目标数据（具有未知类别的数据），以及2）对其他目标数据进行分类。为了实现这一目标，之前的一项研究证明了目标域风险的上界，而开放集差异作为上界中的一个重要项，用于衡量未知目标数据的风险。通过最小化上界，可以训练一个浅层分类器以达到目标。然而，如果分类器非常灵活（例如，深度神经网络（DNNs）），在最小化上界时开放集差异将收敛到负值，这会导致大多数目标数据被识别为未知数据的问题。为了解决这个问题，我们为UOSDA提出了一个新的目标域风险上界，其中包括四个项：源域风险，$\epsilon$-开放集差异（$\Delta_\epsilon$），域之间的分布差异和一个常数。与开放集差异相比，$\Delta_\epsilon$在被最小化时对这个问题更具鲁棒性，因此我们可以使用非常灵活的分类器（即DNNs）。然后，我们提出了一种基于新原则引导的深度UOSDA方法，该方法通过最小化新的上界来训练DNNs。具体而言，通过梯度下降最小化源域风险和$\Delta_\epsilon$，并通过一种新颖的开放集条件对抗训练策略来最小化分布差异。最后，与现有的浅层和深层UOSDA方法相比，我们的方法在多个基准数据集上表现出最先进的性能，包括数字识别（MNIST，SVHN，USPS），物体识别（Office-31，Office-Home）和人脸识别（PIE）。 显示英文摘要

摘要： In the unsupervised open set domain adaptation (UOSDA), the target domain contains unknown classes that are not observed in the source domain. Researchers in this area aim to train a classifier to accurately: 1) recognize unknown target data (data with unknown classes) and, 2) classify other target data. To achieve this aim, a previous study has proven an upper bound of the target-domain risk, and the open set difference, as an important term in the upper bound, is used to measure the risk on unknown target data. By minimizing the upper bound, a shallow classifier can be trained to achieve the aim. However, if the classifier is very flexible (e.g., deep neural networks (DNNs)), the open set difference will converge to a negative value when minimizing the upper bound, which causes an issue where most target data are recognized as unknown data. To address this issue, we propose a new upper bound of target-domain risk for UOSDA, which includes four terms: source-domain risk, $\epsilon$-open set difference ($\Delta_\epsilon$), a distributional discrepancy between domains, and a constant. Compared to the open set difference, $\Delta_\epsilon$ is more robust against the issue when it is being minimized, and thus we are able to use very flexible classifiers (i.e., DNNs). Then, we propose a new principle-guided deep UOSDA method that trains DNNs via minimizing the new upper bound. Specifically, source-domain risk and $\Delta_\epsilon$ are minimized by gradient descent, and the distributional discrepancy is minimized via a novel open-set conditional adversarial training strategy. Finally, compared to existing shallow and deep UOSDA methods, our method shows the state-of-the-art performance on several benchmark datasets, including digit recognition (MNIST, SVHN, USPS), object recognition (Office-31, Office-Home), and face recognition (PIE).