Open-Set Domain Adaptation (OSDA) assumes that a target domain contains unknown classes, which are not discovered in a source domain. Existing domain adversarial learning methods are not suitable for OSDA because distribution matching with $\textit{unknown}$ classes leads to negative transfer. Previous OSDA methods have focused on matching the source and the target distribution by only utilizing $\textit{known}$ classes. However, this $\textit{known}$-only matching may fail to learn the target-$\textit{unknown}$ feature space. Therefore, we propose Unknown-Aware Domain Adversarial Learning (UADAL), which $\textit{aligns}$ the source and the target-$\textit{known}$ distribution while simultaneously $\textit{segregating}$ the target-$\textit{unknown}$ distribution in the feature alignment procedure. We provide theoretical analyses on the optimized state of the proposed $\textit{unknown-aware}$ feature alignment, so we can guarantee both $\textit{alignment}$ and $\textit{segregation}$ theoretically. Empirically, we evaluate UADAL on the benchmark datasets, which shows that UADAL outperforms other methods with better feature alignments by reporting state-of-the-art performances.

翻译：开放- 网域适应 (OSDA) 假设目标域包含未知的类别, 而在源域中无法发现。 现有的域对抗学习方法不适合 OSDA, 因为发布匹配 $\ textit{ 未知的 $ 类导致负转移 。 以前的 OSDA 方法只使用 $\ textit{ 已知的 $ 类来匹配源和目标分配 。 但是, 仅以美元为名的匹配可能无法学习 $\ textit{ 未知的 $ 特性空间 。 因此, 我们提议了 未知 Aware Domain Aversarial 学习 (UADALL), 因为它是源和 目标 $\ textitleit{ 已知的 $ 导致负转移。 先前的配置方法只侧重于 $\ textitleitleitleitit{ unnn $ 。 然而, 我们提供对拟议的 $\ textitleitit{ un- un- adre} specal specal ad to spolitial coom suplading, 所以我们可以保证 $\ $\ swead $\ reviews wequalview ex ex ex ex 和 exparlations report the wequal be wequal deadds reports reviews reports abaltibaltibisgal exgal ex exgations abations exgilding exfolgy exgy ex exgy 。 exgment abs abalds abs abs ex abs abs abs abs abs abs abs ex ex ex ex ex exbal ex abaldaldaldal ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex ex