Accurate Detection of Wake Word Start and End Using a CNN

Christin Jose, Yuriy Mishchenko, Thibaud Sénéchal, Anish Shah, Alex Escott, Shiv Naga Prasad Vitaladevuni

Small footprint embedded devices require keyword spotters (KWS) with small model size and detection latency for enabling voice assistants. Such a keyword is often referred to as wake word as it is used to wake up voice assistant enabled devices. Together with wake word detection, accurate estimation of wake word endpoints (start and end) is an important task of KWS. In this paper, we propose two new methods for detecting the endpoints of wake words in neural KWS that use single-stage word-level neural networks. Our results show that the new techniques give superior accuracy for detecting wake words’ endpoints of up to 50 msec standard error versus human annotations, on par with the conventional Acoustic Model plus HMM forced alignment. To our knowledge, this is the first study of wake word endpoints detection methods for single-stage neural KWS.

 DOI: 10.21437/Interspeech.2020-1491

Cite as: Jose, C., Mishchenko, Y., Sénéchal, T., Shah, A., Escott, A., Vitaladevuni, S.N.P. (2020) Accurate Detection of Wake Word Start and End Using a CNN. Proc. Interspeech 2020, 3346-3350, DOI: 10.21437/Interspeech.2020-1491.

  author={Christin Jose and Yuriy Mishchenko and Thibaud Sénéchal and Anish Shah and Alex Escott and Shiv Naga Prasad Vitaladevuni},
  title={{Accurate Detection of Wake Word Start and End Using a CNN}},
  booktitle={Proc. Interspeech 2020},