Patent · US Active

Method and system for training language models to reduce recognition errors

US10176799B2 · kind B2 · utility

5Cited by
2References
20Claims
0Family size

Assignee

Inventors

Key dates

Filing dateFeb 2, 2016
Grant dateJan 8, 2019
Priority date
Expiry dateFeb 2, 2036

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N3/02
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A method and for training a language model to reduce recognition errors, wherein the language model is a recurrent neural network language model (RNNLM) by first acquiring training samples. An automatic speech recognition system (ASR) is applied to the training samples to produce recognized words and probabilites of the recognized words, and an N-best list is selected from the recognized words based on the probabilities. determining word errors using reference data for hypotheses in the N-best list. The hypotheses are rescored using the RNNLM. Then, we determine gradients for the hypotheses using the word errors and gradients for words in the hypotheses. Lastly, parameters of the RNNLM are updated using a sum of the gradients.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.