Patent · US Active

Methods and systems for adapting a model for a speech recognition system

US7827032B2 · kind B2 · utility

40Cited by
96References
37Claims
0Family size

Assignee

Inventors

Key dates

Filing dateOct 6, 2006
Grant dateNov 2, 2010
Priority date
Expiry dateJul 14, 2029

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG10L15/065
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

Possible errors made by a speech recognition system may be identified without using a transcript of words input to the system. A method for model adaptation for a speech recognition system includes determining an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The method may further include adjusting an adaptation, of the model for the word or various models for the various words, based on the error rate. An apparatus identifies possible errors made by a speech recognition system without using a transcript of words input to the system. An apparatus for model adaptation for a speech recognition system includes a processor adapted to estimate an error rate, corresponding to either recognition of instances of a word or recognition of instances of various words, without using a transcript of words input to the system. The apparatus may further include a controller adapted to adjust an adaptation of the model for the word or various models for the various words, based on the error rate.

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