Feature extraction for automatic speech recognition

TitleFeature extraction for automatic speech recognition
Publication TypePatent
Year of Publication2001
AuthorsKingsbury, B., Greenberg S., & Morgan N.
Other Numbers2150
Abstract

An automatic speech recognition apparatus and method with a front end feature extractor that improved recognition performance under adverse acoustic conditions are disclosed. The inventive feature extractor is characterized by a critical bandwidth spectral resolution, an emphasis on slow changes in the spectral structure o the speech signal, and adaptive automatic gain control. In one the feature extractor includes a feature generator configured to compute short-term parameters of the speech signal, a filter system configured to filter the time sequences of the short-term parameters, and a normalizer configured to normalize the filtered parameters with respect to one or more previous values of the filtered parameters. Accordingly, the feature extractor is operable to carry out the following steps: computing short-term parameters of the speech signal; filtering time sequences of the short-term parameters; and normalizing the filtered parameters with respect to one or more previous values of the filtered parameters. The step of filtering and normalizing preferably are performed independently of one another.

Bibliographic Notes

United States Patent 6,308,155. U.S. Class 704/256.1, 704/231, 704/251, 704/255. International Class G10L 15/00 (20060101), G10L 15/02 (20060101). Field of Search 704/206,224,225,236,234,231,251-256

Abbreviated Authors

B. Kingsbury, S. Greenberg, and N. Morgan

ICSI Publication Type

Patent