Recently, Google launched YouTube Kids, a mobile application for children, that uses a speech recognizer built specifically for recognizing children's speech. In this paper we present techniques we explored to build such a system. We describe the use of a neural network classifier to identify matched acoustic training data, filtering data for language modeling to reduce the chance of producing offensive results. We also compare long short-term memory (LSTM) recurrent networks to convolutional, LSTM, deep neural networks (CLDNN). We found that a CLDNN acoustic model outperforms an LSTM across a variety of different conditions, but does not specifically model child speech relatively better than adult. Overall, these findings allow us to build a successful, state-of-the-art large vocabulary speech recognizer for both children and adults.
Cite as: Liao, H., Pundak, G., Siohan, O., Carroll, M.K., Coccaro, N., Jiang, Q.-M., Sainath, T.N., Senior, A., Beaufays, F., Bacchiani, M. (2015) Large vocabulary automatic speech recognition for children. Proc. Interspeech 2015, 1611-1615, doi: 10.21437/Interspeech.2015-373
@inproceedings{liao15_interspeech, author={Hank Liao and Golan Pundak and Olivier Siohan and Melissa K. Carroll and Noah Coccaro and Qi-Ming Jiang and Tara N. Sainath and Andrew Senior and Françoise Beaufays and Michiel Bacchiani}, title={{Large vocabulary automatic speech recognition for children}}, year=2015, booktitle={Proc. Interspeech 2015}, pages={1611--1615}, doi={10.21437/Interspeech.2015-373} }