Back to results
US 2003/0144835 A1Published application
Correlation domain formant enhancement
OwnerRichard L. Zinser JR., Steven R. Kochinventors
Life of this patent
Estimating the term…
Abstract
The present invention comprises a system and method for correlation domain formant enhancement. The method includes attenuating an output speech signal in areas of low spectral amplitude. The attenuated signals are subsequently analyzed, and then converted to a format suitable for transmission.
Claims4
A method for enhancing prediction coefficient signals, comprising:
- a) receiving a set of prediction coefficient signals;
- b) generating a set of bandwidth expanded coefficient signals based on said prediction coefficient signals;
- c) converting at least some of said prediction coefficient signals and said bandwidth expanded coefficient signals into corresponding autocorrelation domain coefficient signals;
- d) convolving said autocorrelation domain coefficient signals;
- e) converting said convolved autocorrelation domain coefficient signals into reflection coefficient signals;
- f) determining the stability of said reflection coefficient signals;
- and
- g) converting said reflection coefficient signals into enhanced prediction coefficient signals.
5 more elements
A device for enhancing speech prediction coefficient signals, comprising:
- a) means for receiving a set of prediction coefficient signals;
- b) means for generating a set of bandwidth expanded coefficient signals based on said prediction coefficient signals;
- c) means for converting at least some of said prediction coefficient signals and said bandwidth expanded coefficient signals into corresponding autocorrelation domain coefficient signals;
- d) means for convolving said autocorrelation domain coefficient signals;
- e) means for converting said convolved autocorrelation domain coefficient signals into reflection coefficient signals;
- f) means for determining the stability of said reflection coefficient signals;
- and
- g) means for converting said reflection coefficient signals into enhanced prediction coefficient signals.
5 more elements
A method for enhancing coded speech signals, comprising:
- a) receiving a set of prediction coefficient signals;
- b) generating a set of bandwidth expanded coefficient signals based on said prediction coefficient signals;
- c) converting at least some of said prediction coefficient signals and said bandwidth expanded coefficient signals into corresponding autocorrelation domain coefficient signals;
- d) convolving said autocorrelation domain coefficient signals;
- e) converting said convolved autocorrelation domain coefficient signals into reflection coefficient signals;
- f) determining the stability of said reflection coefficient signals;
- and converting said reflection coefficient signals into enhanced prediction coefficient signals.
- g) receiving coded speech signals;
- and
- h) processing said coded speech signals based on said enhanced prediction coefficient signals.
7 more elements
A device for enhancing coded speech signals, comprising:
- a) means for receiving a set of prediction coefficient signals;
- b) means for generating a set of bandwidth expanded coefficient signals based on said prediction coefficient signals;
- c) means for converting at least some of said prediction coefficient signals and said bandwidth expanded coefficient signals into corresponding autocorrelation domain coefficient signals;
- d) means for convolving said autocorrelation domain coefficient signals;
- e) means for converting said convolved autocorrelation domain coefficient signals into reflection coefficient signals;
- f) means for determining the stability of said reflection coefficient signals;
- and
- converting said reflection coefficient signals into enhanced prediction coefficient signals.
- g) means for receiving coded speech signals;
- and
- h) means for processing said coded speech signals based on said enhanced prediction coefficient signals.
8 more elements
Dependencies are inferred from the claim text.