Method and System to Seed a Voice Database

• US 20070280436A1
• Filed: 05/29/2007
• Published: 12/06/2007
• Est. Priority Date: 04/14/2006
• Status: Abandoned Application

• Current Assignee: Victria Incorporated (Verint Systems Incorporated)
• Original Assignee: Victria Incorporated (Verint Systems Incorporated)
• Inventors: Rajakumar, Anthony
• Application Number:
  US11/754,975
• Publication Number:
  US 20070280436A1
• Time in Patent Office: Days
• US Class Current: 379/88.20
• CPC Class Codes: G06Q 20/24 Credit schemes, i.e. "pay after"
• CPC Class Codes: G06Q 20/4014 Identity check for transactions
• CPC Class Codes: G10L 17/00 Speaker identification or verification techniques

Abstract

In one embodiment the invention provides a method, comprising: (a) obtaining an audio recording corresponding to a legal proceeding of interest; (b) extracting from the audio recording audio belonging to of the defendant of the legal proceeding; and (c) enrolling the extracted audio into a voiceprint database.

Claims

1. A method, comprising:

(a) obtaining an audio recording corresponding to a legal proceeding of interest;

(b) extracting from the audio recording audio belonging to a person prosecuted for or convicted as a result of the legal proceeding; and

(c) enrolling the extracted audio into a voiceprint database.

2. The method of claim 1, wherein the legal proceeding of interest comprises a legal proceeding for a crime of moral turpitude.

3. The method of claim 1, further comprising performing a screening operation to determine if a voice sample for a candidate matches a voiceprint in the voiceprint database.

4. The method of claim 3, wherein the screening operation is used to disqualify the candidate if the voice sample matches a voiceprint in the voiceprint database.

5. The method of claim 3, further comprising, for each voiceprint in the voiceprint database associating metadata therewith.

6. The method of claim 5, wherein the metadata comprises location information.

7. The method of claim 5, further comprising segmenting the voiceprint database according to the metadata.

8. The method of claim 6, wherein a confidence of a match between the voice sample of the candidate and a voiceprint in the database is increased based on a match between location information for the candidate and the location information associated with the voiceprint.

9. The method of claim 7, wherein each voice signature is assigned a home segmented database within which it enjoys a threshold rate accept rate.

10. The method of claim 9, wherein for each voiceprint that has gone inactive within its home segmented database, a search is carried out to determine if the voiceprint is active within another segmented database.

11. The method of claim 10, wherein if the voiceprint is determined to be active in the other segmented database, the other segmented database is set as the home segmented database for the voiceprint.

12. The method of claim 10, wherein determining if the voiceprint is active in the other segmented database comprises confirming the matching using human intervention.

13. A system, comprising:

a processor; and

a memory coupled to the processor, the memory storing instructions which when executed by the processor causes the system to perform a method comprising;

(a) receiving an audio recording corresponding to a legal proceeding of interest;

(b) extracting from the audio recording audio belonging to a person convicted as a result of the legal proceeding; and

(c) enrolling the extracted audio into a voiceprint database.

14. The system of claim 13, wherein the legal proceeding of interest comprises a legal proceeding for a crime of moral turpitude.

15. The system of claim 14, wherein the method further comprises performing a screening operation to determine if a voice sample for a candidate matches a voiceprint in the voiceprint database.

16. The system of claim 15, wherein the screening operation is used to disqualify the candidate if the voice sample matches a voiceprint in the voiceprint database.

17. The system of claim 13, wherein the method further comprises, for each voiceprint in the voiceprint database associating metadata therewith.

18. The system of claim 17, wherein the metadata comprises location information.

19. The system of claim 17, wherein the method further comprises segmenting the voiceprint database according to the metadata.

20. The system of claim 19, wherein a confidence of a match between the voice sample of the candidate and a voiceprint in the database is increased based on a match between location information for the candidate and the location information associated with the voiceprint.