Enhancing model prediction confidence on target data is an important objective in Unsupervised Domain Adaptation (UDA). In this paper, we explore adversarial training on penultimate activations, i.e., input features of the final linear classification layer. We show that this strategy is more efficient and better correlated with the objective of boosting prediction confidence than adversarial training on input images or intermediate features, as used in previous works. Furthermore, with activation normalization commonly used in domain adaptation to reduce domain gap, we derive two variants and systematically analyze the effects of normalization on our adversarial training. This is illustrated both in theory and through empirical analysis on real adaptation tasks. Extensive experiments are conducted on popular UDA benchmarks under both standard setting and source-data free setting. The results validate that our method achieves the best scores against previous arts. Code is available at https://github.com/tsun/APA.

翻译：提升模型对目标数据的预测置信度是无监督领域自适应（UDA）的重要目标。本文探索在倒数第二层激活（即最终线性分类层的输入特征）上进行对抗训练。我们证明，与先前工作中使用的输入图像或中间特征对抗训练相比，该策略效率更高且与提升预测置信度的目标关联性更强。进一步地，结合领域自适应中常用的激活归一化方法以减少领域差异，我们推导出两种变体，并从理论上和实际自适应任务的实证分析中系统研究了归一化对对抗训练的影响。在标准设置和源数据不可获取设置下，我们于通用UDA基准上进行了大量实验。结果表明，我们的方法相较先前方法取得了最优性能。代码已开源至https://github.com/tsun/APA。