William L Gale

US Patent:

20210304769, Sep 30, 2021

Filed:

May 14, 2020

Appl. No.:

15/931788

Inventors:

- Redmond WA, US
Yan Huang - Redmond WA, US
Wenning Wei - Beijing, CN
Lei He - Beijing, CN
Eva Sharma - Vancouver, CA
Jian Wu - Bellevue WA, US
Yao Tian - Beijing, CN
Edward C. Lin - Beijing, CN
Yifan Gong - Sammamish WA, US
Rui Zhao - Bellevue WA, US
Jinyu Li - Redmond WA, US
William Maxwell Gale - Sunnyvale CA, US

International Classification:

G10L 15/26
G10L 15/16
G10L 15/06
G10L 13/08

Abstract:

Systems, methods, and devices are provided for generating and using text-to-speech (TTS) data for improved speech recognition models. A main model is trained with keyword independent baseline training data. In some instances, acoustic and language model sub-components of the main model are modified with new TTS training data. In some instances, the new TTS training is obtained from a multi-speaker neural TTS system for a keyword that is underrepresented in the baseline training data. In some instances, the new TTS training data is used for pronunciation learning and normalization of keyword dependent confidence scores in keyword spotting (KWS) applications. In some instances, the new TTS training data is used for rapid speaker adaptation in speech recognition models.