IP Library Granted Patent US 7,457,750
Granted Patent B2
US 7,457,750 · App. 09/972,929 · Granted Nov 25, 2008

Systems and methods for dynamic re-configurable speech recognition

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 7,457,750
App. No.
09/972,929
Granted
Nov 25, 2008
Kind
B2
Abstract

Speech recognition models are dynamically re-configurable based on user information, background information such as background noise and transducer information such as transducer response characteristics to provide users with alternate input modes to keyboard text entry. The techniques of dynamic re-configurable speech recognition provide for deployment of speech recognition on small devices such as mobile phones and personal digital assistants as well environments such as office, home or vehicle while maintaining the accuracy of the speech recognition.

Claims (49)

1. A method of dynamic re-configurable speech recognition comprising:

determining parameters of a background model and a transducer model at a periodic time during a received voice request;

increasing the periodic time when successive changes in sample noise information and sample transducer information do not exceed a threshold value;

determining an adapted speech recognition model based on the background model and the transducer model;

recognizing the voice request using the adapted speech recognition model;

translating the recognized voice request into an HTTP protocol request; and

generating a response to the recognized voice request based on information from a database based on the HTTP protocol request.

2. The method of claim 1 , wherein,

the parameters of the background model are determined based on a first sample period; and

the parameters of the transducer model are determined based on a second sample period.

3. The method of claim 1 , further comprising:

saving at least one of the parameters of the background model or the parameters of the transducer model.

4. A system for dynamic re-configurable speech recognition comprising:

a background model estimation circuit for determining a background model during a voice request based on estimated background parameters determined at a periodic time during a reception of the voice request;

a transducer model estimation circuit for determining a transducer model of the voice request based on estimated transducer parameters determined at the periodic time during a reception of the voice request;

a threshold value circuit that increases the periodic time when successive in changes in sample noise information and sample transducer information do not exceed a threshold value;

an adaptation circuit for determining an adapted speech recognition model based on a speech recognition model, the background model and the transducer model;

a speech recognizer for recognizing the voice request;

a translator adapted to translate the recognized voice request into an HTTP protocol request; and

a controller adapted to generate a response to the recognized voice request based on information from a database based on the HTTP protocol request.

5. The system of claim 4 , wherein, the controller periodically activates the background model estimation circuit and the transducer model estimation circuit.

6. The system of claim 4 , wherein,

the background model is determined based on a first sample period; and

the transducer model is determined based on a second sample period.

7. The system of claim 6 , wherein the controller saves at least one of the background model or the transducer model into storage.

8. A computer readable memory medium comprising:

computer readable program code embodied on the computer readable memory medium, said computer readable program code usable to program a computer to perform a method for dynamic re-configurable speech recognition comprising:

determining parameters of a background model and a transducer model at a periodic time during a received voice request;

increasing the periodic time when successive changes in sample noise information and sample transducer information do not exceed a threshold value;

determining an adapted speech recognition model based on the background model and the transducer model;

recognizing the voice request using the adapted speech recognition model;

translating the recognized voice request into an HTTP protocol request; and

generating a response to the recognized voice request based on information from a database based on the HTTP protocol request.

9. The system of claim 7 , wherein the controller saves at least one of the background model or the transducer model into storage.

10. The system of claim 6 , wherein the controller is further adapted to adjust the periodic time based, at least in part, on a frequency or a magnitude of determined changes in successively sampled ones of the noise information.

11. A computer readable memory medium comprising:

computer readable program code embodied on the computer readable memory medium, said computer readable program code usable to program a computer to perform a method for dynamic re-configurable speech recognition comprising:

determining parameters of a background model and a transducer model at a periodic time during a received voice request;

increasing the periodic time when successes changes in sample noise information and sample transducer information do not exceed a threshold value:

determining an adapted speech recognition model based on the background model and the transducer model;

recognizing the voice request using the adapted speech recognition model;

translating the recognized voice request into an HTTP protocol request; and

generating a response to the recognized voice request based on information from a database based on the HTTP protocol request.

12. The computer readable memory medium of claim 11 , wherein:

the background model is determined based on a first sample period; and

the transducer model is determined based on a second sample period.

13. The computer readable memory medium of claim 11 , wherein the method further comprises saving at least one of the background model or the transducer model.

14. The computer readable memory medium of claim 11 , wherein determining parameters of the background model and a transducer model at a periodic time during a received voice request further comprises periodic sampling during periods of speech inactivity while receiving the voice request.

15. The computer readable memory medium of claim 11 , wherein the method further comprises dynamically determining the periodic time based, at least in part, on a frequency or a magnitude of determined changes in the sampled noise information.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 26, 2017
From: AT&T INTELLECTUAL PROPERTY II, L.P.
To: NUANCE COMMUNICATIONS, INC.
Reel/Frame 041512/0608 →