Adversarial Training (AT) is proposed to alleviate the adversarial vulnerability of machine learning models by extracting only robust features from the input, which, however, inevitably leads to severe accuracy reduction as it discards the non-robust yet useful features. This motivates us to preserve both robust and non-robust features and separate them with disentangled representation learning. Our proposed Adversarial Asymmetric Training (AAT) algorithm can reliably disentangle robust and non-robust representations without additional supervision on robustness. Empirical results show our method does not only successfully preserve accuracy by combining two representations, but also achieve much better disentanglement than previous work.

中文翻译：

---

无需（额外）监督的无解码器鲁棒性解开

提出对抗性训练 (AT) 以通过仅从输入中提取稳健特征来减轻机器学习模型的对抗性脆弱性，然而，由于丢弃了非稳健但有用的特征，不可避免地导致精度严重降低。这促使我们保留健壮和非健壮的特征，并用解开的表示学习将它们分开。我们提出的对抗性非对称训练 (AAT) 算法可以可靠地分离鲁棒和非鲁棒表示，而无需对鲁棒性进行额外监督。实证结果表明，我们的方法不仅通过组合两种表示成功地保持了准确性，而且比以前的工作实现了更好的解开。