Methods and systems for providing noise filtering using speech recognition

• US 20060282264A1
• Filed: 06/09/2005
• Published: 12/14/2006
• Est. Priority Date: 06/09/2005
• Status: Abandoned Application

• Current Assignee: AT&T Intellectual Property I LP (AT&T, Inc.)
• Original Assignee: Bellsouth Intellectual Property Corporation (AT&T, Inc.)
• Inventors: Anderson, David J., Denny, Michael S.
• Application Number:
  US11/148,457
• Publication Number:
  US 20060282264A1
• Time in Patent Office: Days
• US Class Current: 704/233
• CPC Class Codes: G10L 21/0208 Noise filtering
• CPC Class Codes: G10L 21/0232 Processing in the frequency domain

Abstract

Systems and methods are disclosed for providing noise filtering using speech recognition. The disclosed systems and methods may include performing a speech recognition process on a signal. Furthermore, the disclosed systems and methods may include determining a first frequency range based on the performed speech recognition process, the first frequency range corresponding to an expected voice associated with the signal. In addition, the disclosed systems and methods may include attenuating a second frequency range in the signal, the second frequency range being outside the first frequency range and the extemporaneous noise in the first frequency.

Claims

1. A method for providing noise filtering using speech recognition, the method comprising:

performing a speech recognition process on a signal;

determining a first frequency range based on the performed speech recognition process, the first frequency range corresponding to an expected voice signal associated with the signal; and

attenuating a second frequency range in the signal containing noise and being outside the first frequency range, the second frequency range being outside the first frequency range.

2. The method of claim 1, wherein performing the speech recognition process further comprises performing the speech recognition process in response to at least one of a caller initiated input, a caller initiated voice command, a characteristic of a calling identifier, a test for noise in the signal.

3. The method of claim 1, wherein performing the speech recognition process further comprises performing the speech recognition process using at least one of a directed language model, an open language model, a statistical language model, and a statistical semantic model with related phonetic based acoustic models.

4. The method of claim 1, wherein performing the speech recognition process further comprises performing the speech recognition process that considers at least one of the environment of a caller, an accent of the caller, the identity of the caller, a national region associated with the caller, a caller initiated input, a caller initiated voice command, a characteristic a calling identifier, a test for noise in the signal.

5. The method of claim 1, wherein performing the speech recognition process further comprises performing the speech recognition process that considers data previously gather and associated with a caller'"'"'s voice.

6. The method of claim 1, wherein the first frequency range and the second frequency range are in a frequency range audible by humans.

7. The method of claim 1, further comprises adding to the signal in order to improve voice information transmitted by the signal.

8. A system for providing noise filtering using speech recognition, the system comprising:

a memory storage for maintaining a database; and

a processing unit coupled to the memory storage, wherein the processing unit is operative to;

perform a speech recognition process on a signal;

determine a first frequency range based on the performed speech recognition process, the first frequency range corresponding to an expected voice signal associated with the signal; and

attenuate a second frequency range in the signal containing noise and being outside the first frequency range, the second frequency range being outside the first frequency range.

9. The system of claim 8, wherein the processing unit being operative to perform the speech recognition process further comprises the processing unit being operative to perform the speech recognition process in response to at least one of a caller initiated input, a caller initiated voice command, a characteristic a calling identifier, a test for noise in the signal.

10. The system of claim 8, wherein the processing unit being operative to perform the speech recognition process further comprises the processing unit being operative to perform the speech recognition process using at least one of a directed language model, an open language model, a statistical language model, and a statistical semantic model with related phonetic based acoustic models.

11. The system of claim 8, wherein the processing unit being operative to perform the speech recognition process further comprises the processing unit being operative to perform the speech recognition process that considers at least one of the environment of a caller, an accent of the caller, the identity of the caller, a national region associated with the caller, a caller initiated input, a caller initiated voice command, a characteristic of a calling identifier, a test for noise in the signal.

12. The system of claim 8, wherein the processing unit being operative to perform the speech recognition process further comprises the processing unit being operative to perform the speech recognition process that considers data previously gather and associated with a caller'"'"'s voice.

13. The system of claim 8, wherein the first frequency range and the second frequency range are in a frequency range audible by humans.

14. A computer-readable medium which stores a set of instructions which when executed performs a method for providing noise filtering using speech recognition, the method executed by the set of instructions comprising:

performing a speech recognition process on a signal;

determining a first frequency range based on the performed speech recognition process, the first frequency range corresponding to an expected voice signal associated with the signal; and

attenuating a second frequency range in the signal containing noise and being outside the first frequency range, the second frequency range being outside the first frequency range.

15. The computer-readable medium of claim 14, wherein performing the speech recognition process further comprises performing the speech recognition process in response to at least one of a caller initiated input, a caller initiated voice command, a characteristic of a calling identifier, a test for noise in the signal.

16. The computer-readable medium of claim 14, wherein performing the speech recognition process further comprises performing the speech recognition process using at least one of a directed language model, an open language model, a statistical language model, and a statistical semantic model with related phonetic based acoustic models.

17. The computer-readable medium of claim 14, wherein performing the speech recognition process further comprises performing the speech recognition process that considers at least one of the environment of a caller, an accent of the caller, the identity of the caller, a national region associated with the caller, a caller initiated input, a caller initiated voice command, a characteristic a calling identifier, a test for noise in the signal.

18. The computer-readable medium of claim 14, wherein performing the speech recognition process further comprises performing the speech recognition process that considers data previously gather and associated with a caller'"'"'s voice.

19. The computer-readable medium of claim 14, wherein the first frequency range and the second frequency range are in a frequency range audible by humans.

20. The computer-readable medium of claim 14, further comprises adding to the signal in order to improve voice information transmitted by the signal.