Statistical-analysis-based reset of recurrent neural networks for automatic speech recognition

US10255909B2 · kind B2 · utility

Assignee

• Intel IP Corporation · US

Inventors

• Joachim Hofer · Munich, DE
• Josef Bauer · Munich, DE
• Piotr Rozen · Gdańsk, PL
• Georg Stemmer · Munich, DE

Key dates

• Filing date: Jun 29, 2017
• Grant date: Apr 9, 2019
• Expiry date: Jun 29, 2037

Classification

• Technology area (CPC G): Physics
• CPC primary: G10L15/063
• WIPO field: Computer technology
• WIPO sector: Electrical engineering

Abstract

Techniques are provided for calculating reset parameters for recurrent neural networks (RNN). A methodology implementing the techniques according to an embodiment includes generating a sequence of statistics. The calculation of each statistic is based on outputs of an RNN that is periodically re-initialized at a selected RNN reset time such that each of the calculated statistics is associated with a unique RNN reset time selected from a pre-determined range of reset times. The method further includes analyzing the sequence to identify a maximum interval during which the sequence remains relatively constant. The method further includes selecting a reset time parameter and reset context duration parameter, for re-initialization of the RNN during operation. The reset time parameter is based on the duration of the identified maximum interval and the reset context duration parameter is based on a time associated with the starting point of the identified maximum interval.