Deployed end-to-end speech recognition

US10319374B2 · kind B2 · utility

Assignee

• BAIDU USA LLC · US

Inventors

• Bryan Catanzaro · Cupertino, US
• Jingdong Chen · Beijing, CN
• Mike Chrzanowski · Sunnyvale, US
• Erich Konrad Elsen · Mountain View, US
• Jesse Engel · Oakland, US
• Christopher Fougner · London, GB
• Xu Han · Santa Clara, US
• Awni Hannun · Los Altos, US
• Ryan Prenger · Oakland, US
• Sanjeev Satheesh · Sunnyvale, US
• Shubhabrata Sengupta · Menlo Park, US
• Dani Yogatama · Sunnyvale, US
• Chong Wang · Redmond, US
• Jun Zhan · Sunnyvale, US
• Zhenyao Zhu · Mountain View, US
• Dario Amodei · San Francisco, US

Key dates

• Filing date: Nov 21, 2016
• Grant date: Jun 11, 2019
• Expiry date: Nov 21, 2036

Classification

• Technology area (CPC G): Physics
• CPC primary: G10L2015/0635
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

Embodiments of end-to-end deep learning systems and methods are disclosed to recognize speech of vastly different languages, such as English or Mandarin Chinese. In embodiments, the entire pipelines of hand-engineered components are replaced with neural networks, and the end-to-end learning allows handling a diverse variety of speech including noisy environments, accents, and different languages. Using a trained embodiment and an embodiment of a batch dispatch technique with GPUs in a data center, an end-to-end deep learning system can be inexpensively deployed in an online setting, delivering low latency when serving users at scale.