A proper representation that can well express the characteristics of a word plays an important role in wake-up word detection (WWD). However, it may be easily corrupted due to various types of environmental noise occurred in the place where WWD typically works, causing unreliable performance. To deal with this practical issue, we propose a novel strategy called cross-informed domain adversarial training (CiDAT) for noise-robust WWD. In the method, additional paths were introduced to conventional domain adversarial training (DAT) to encourage its ability to generate domain-invariant representation. Experiments on the Aurora4 corpus verified that CiDAT significantly outperformed the baselines as well as conventional DAT.

中文翻译：

---

用于噪声鲁棒唤醒词检测的交叉知情域对抗训练

能够很好地表达单词特征的适当表示在唤醒词检测（WWD）中起着重要作用。但是，由于 WWD 通常工作的地方会出现各种类型的环境噪音，因此它可能很容易损坏，从而导致性能不可靠。为了解决这个实际问题，我们提出了一种新的策略，称为用于噪声鲁棒 WWD 的交叉知情域对抗训练 (CiDAT)。在该方法中，额外的路径被引入到传统的域对抗训练 (DAT) 中，以鼓励其生成域不变表示的能力。在 Aurora4 语料库上的实验证实，CiDAT 显着优于基线以及传统 DAT。