Wake-up word detection models are widely used in real life, but suffer from severe performance degradation when encountering adversarial samples. In this paper we discuss the concept of confusing words in adversarial samples. Confusing words are commonly encountered, which are various kinds of words that sound similar to the predefined keywords. To enhance the robustness of the wake-up word detection system against confusing words, we propose several methods to generate the adversarial confusing samples for simulating real confusing words scenarios in which we usually do not have any real confusing samples in the training set. The generated samples include concatenated audio, synthesized data, and partially masked keywords. Moreover, we use a domain embedding concatenated system to improve the performance. Experimental results show that the adversarial samples generated in our approach help improve the system’s robustness in both the common scenario and the confusing words scenario. In addition, we release the confusing words testing database called HI-MIA-CW for future research.
Cite as: Wang, H., Jia, Y., Zhao, Z., Wang, X., Wang, J., Li, M. (2022) Generating TTS Based Adversarial Samples for Training Wake-Up Word Detection Systems Against Confusing Words. Proc. The Speaker and Language Recognition Workshop (Odyssey 2022), 402-406, doi: 10.21437/Odyssey.2022-56
@inproceedings{wang22c_odyssey,
author={Haoxu Wang and Yan Jia and Zeqing Zhao and Xuyang Wang and Junjie Wang and Ming Li},
title={{Generating TTS Based Adversarial Samples for Training Wake-Up Word Detection Systems Against Confusing Words}},
year=2022,
booktitle={Proc. The Speaker and Language Recognition Workshop (Odyssey 2022)},
pages={402--406},
doi={10.21437/Odyssey.2022-56}
}