Patent · US Expired

Method for reducing noise distortions in a speech recognition system

US6173258A · kind A · utility

31Cited by
10References
54Claims
0Family size

Assignees

Inventors

Key dates

Filing dateOct 22, 1998
Grant dateJan 9, 2001
Priority date
Expiry dateOct 22, 2018

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG10L21/0264
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A method for reducing noise distortions in a speech recognition system comprises a feature extractor that includes a noise-suppressor, one or more time cosine transforms, and a normalizer. The noise-suppressor preferably performs a spectral subtraction process early in the feature extraction procedure. The time cosine transforms preferably operate in a centered-mode to each perform a transformation in the time domain. The normalizer calculates and utilizes normalization values to generate normalized features for speech recognition. The calculated normalization values preferably include mean values, left variances and right variances.

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