Method and apparatus for performing speaker verification based on speaker independent recognition of commands

• US 6,973,426 B1
• Filed: 12/29/2000
• Issued: 12/06/2005
• Est. Priority Date: 12/29/2000
• Status: Active Grant

• Current Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Original Assignee: Cisco Technology, Inc. (Cisco Systems, Inc.)
• Inventors: Jackson, Patrick E., Schier, John E.
• Primary Examiner(s): ABEBE, DANIEL DEMELASH
• Application Number:
  US09/752,043
• Time in Patent Office: 1,803 Days
• Field of Search: 704/197, 704/275, 704/270, 704/273, 704/246
• US Class Current: 704/246
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces

Abstract

A method and apparatus for performing speaker verification with speaker verification templates are disclosed. A verification unit receives an utterance from a speaker. The verification unit identifies a command associated with the utterance by performing speaker independent recognition. If a speaker verification template associated with the identified command includes adequate verification data, the verification unit eliminates a prompt for a password and verifies the speaker identity by comparing the utterance with a speaker verification template associated with the identified command.

Claims

1. A method for performing speaker verification based on speaker independent recognition of commands, the method comprising:

receiving, from a client, an utterance spoken by a speaker;

identifying a user associated with the client;

identifying a command associated with the utterance by performing speaker independent recognition (SIR);

determining whether a speaker verification template associated with the identified command includes adequate verification data to verify utterances of the identified command by the identified user;

eliminating a prompt for a password if the speaker verification template associated with the identified command includes adequate verification data; and

verifying a speaker identity by comparing the utterance with the speaker verification template associated with the identified command.

2. The method of claim 1, further comprising recording speaker dependent voice patterns associated with the utterance to create the speaker verification template.

3. The method of claim 1, further comprising:

recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

executing the identified command if the utterance matches the speaker verification template.

4. The method of claim 1, further comprising:

prompting the speaker to enter a password if the utterance does not match the speaker verification template; and

verifying the speaker identity based on the password.

5. The method of claim 1, further comprising:

prompting the speaker to enter a password if the utterance does not match the speaker verification template;

verifying the speaker identity based on the password;

recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

executing the identified command.

6. The method of claim 1, further comprising:

prompting the speaker for a password if the speaker verification template associated with the identified command does not comprise adequate verification data; and

verifying the speaker identity based on the password.

7. The method of claim 1, further comprising:

prompting the speaker for a password if the speaker verification template associated with the identified command does not comprise adequate verification data;

verifying the speaker identity based on the password;

recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

repeating the prompting, verifying and recording steps until the speaker verification template associated with the identified command comprises adequate verification data.

8. A speaker verification unit, comprising:

a network interface operable to couple to a communication network;

a database coupled to the network interface, the database comprising a plurality of speaker verification templates associated with a plurality of commands; and

a processing module coupled to the network interface, the processing module operable to;

receive, from a client, an utterance spoken by a speaker;

identify a user associated with the client;

identify a command associated with the utterance by performing speaker independent recognition (SIR);

determine whether a speaker verification template associated with the identified command includes adequate verification data to verify utterances of the identified command by the identified user;

eliminate a prompt for a password if the speaker verification template associated with the identified command includes adequate verification data; and

verify a speaker identity by comparing the utterance with the speaker verification template associated with the identified command.

9. The speaker verification unit of claim 8, wherein the processing module is further operable to:

record speaker dependant voice patterns associated with the utterance; and

store the speaker dependent voice patterns in the database to create the speaker verification template associated with the identified command.

10. The speaker verification unit of claim 8, further comprising a service interface operable to couple to a server comprising a plurality of services associated with the commands.

11. The speaker verification unit of claim 10, wherein the processing module is further operable to:

record speaker dependent voice patterns associated with the utterance;

store the speaker dependent voice patterns in the database to create the speaker verification template associated with the identified command; and

access a service associated with the identified command from the server.

12. The speaker verification unit of claim 8, wherein:

the database further comprises a password associated with the user; and

the processing module is further operable to;

prompt the speaker to enter the password if the utterance does not match the speaker verification template;

compare the entered password with the password stored in the database; and

verify the speaker identity if the entered password matches the password stored in the database.

13. The speaker verification unit of claim 8, wherein:

the database further comprises a password associated with the user; and

the processing module is further operable to;

prompt the speaker to enter the password if the utterance does not match the speaker verification template;

compare the entered password with the password stored in the database;

verify the speaker identity if the entered password matches the password stored in the database;

record speaker dependent voice patterns associated with the utterance; and

store the speaker dependent voice patterns in the database to create the speaker verification template associated with the identified command.

14. The speaker verification unit of claim 8, wherein:

the database further comprises a password associated with the user; and

the processing module is further operable to;

prompt the speaker to enter the password if the speaker verification template does not comprise adequate verification data;

compare the entered password with the password stored in the database; and

verify the speaker identity if the entered password matches the password stored in the database.

15. The speaker verification unit of claim 8, wherein:

the database further comprises a password associated with the user; and

the processing module is further operable to;

prompt the speaker to enter the password if the speaker verification template does not comprise adequate verification data;

compare the entered password with the password stored in the database;

verify the speaker identity if the entered password matches the password stored in the database;

record speaker dependant voice patterns associated with the utterance; and

store the speaker dependent voice patterns in the database to create the speaker verification template associated with the identified command.

16. The speaker verification unit of claim 8, wherein the command is selected from the group consisting essentially of call, store, access, forward, or redial.

17. Logic encoded in media for performing speaker verification based on speaker independent recognition of commands and operable to perform the following steps:

receiving, from a client, an utterance spoken by a speaker;

identifying a user associated with the client,

identifying a command associated with the utterance by performing speaker independent recognition (SIR);

determining whether a speaker verification template associated with the identified command includes adequate verification data to verify utterances of the identified command by the identified user;

eliminating a prompt for a password if the speaker verification template associated with the identified command includes adequate verification data; and

verifying a speaker identity by comparing the utterance with the speaker verification template associated with the identified command.

18. The logic encoded in media of claim 17, further comprising:

recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

executing the identified command if the utterance matches the speaker verification template.

19. The logic encoded in media of claim 17, further comprising:

prompting the speaker to enter a password if the utterance does not match the speaker verification template;

verifying the speaker identity based on the password;

recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

executing the identified command.

20. The logic encoded in media of claim 17, further comprising:

prompting the speaker for a password if the speaker verification template associated with the identified command does not comprise adequate verification data;

verifying the speaker identity based on the password;

recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

repeating the prompting, verifying and recording steps until the speaker verification template associated with the identified command comprises adequate verification data.

21. An apparatus for performing speaker verification based on speaker independent recognition of commands:

means for receiving, from a client, an utterance spoken by a speaker;

means for identifying a user associated with the client;

means for identifying a command associated with the utterance by performing speaker independent recognition (SIR);

means for determining whether a speaker verification template associated with the identified command includes adequate verification data to verify utterances of the identified command by the identified user;

means for eliminating a prompt for a password if the speaker verification template associated with the identified command includes adequate verification data; and

means for verifying a speaker identity by comparing the utterance with the speaker verification template associated with the identified command.

22. The apparatus of claim 21, further comprising:

means for recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

means for executing the identified command if the utterance matches the speaker verification template.

23. The apparatus of claim 21, further comprising:

means for prompting the speaker for a password if the speaker verification template associated with the identified command does not comprise adequate verification data;

means for verifying the speaker identity based on the password;

means for recording speaker dependent voice patterns associated with the utterance to create the speaker verification template; and

means for repeating the prompting, verifying and recording steps until the speaker verification template associated with the identified command comprises adequate verification data.