Systems and methods for providing network-based voice authentication

• US 8,014,496 B2
• Filed: 07/28/2004
• Issued: 09/06/2011
• Est. Priority Date: 07/28/2004
• Status: Expired due to Fees

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Verizon Business Global LLC (Verizon Communications Inc.)
• Inventors: Schultz, Paul T.
• Primary Examiner(s): Elahee; Md S
• Application Number:
  US10/900,335
• Publication Number:
  US 20060029190A1
• Time in Patent Office: 2,596 Days
• Field of Search: 379 8801- 8804, 379/88.21, 379/93.03, 379/88.19, 379/88.2, 379/142.05, 379/142.06, 704246-250, 704/233, 704/251, 704E17001-E17003, 704/E17.01, 704/E17.011, 704/E15.001, 704/E15.002, 704/231, 370/392, 726/7, 726/5
• US Class Current: 379/88.02
• CPC Class Codes: H04M 3/385 using speech signals

Abstract

A system enables voice authentication via a network. The system may include an intelligent voice response engine operatively coupled to the network for receiving transaction or access requests from a plurality of telecommunications devices over the network. A speech recognition and verification services engine may be operatively coupled to the network and a database may be operatively coupled to the speech recognition and verification services engine for storing user voice print profiles. The speech recognition and verification services engine may receive a speaker verification call from the intelligent voice response engine and perform speaker verification on the received speaker verification call based on the stored user voice print profiles. The speech recognition and verification services engine may generate a verification score based upon results of the speaker verification.

Claims

1. A system for providing voice authentication via a network, comprising:

a voice response system operatively coupled to the network for receiving transaction or access requests from a plurality of telecommunications devices over the network;

a speech recognition and verification services engine operatively coupled to the network; and

a memory operatively coupled to the speech recognition and verification services engine for storing user voice print profiles,

where the voice response system is to;

receive a request from a telecommunications device of the plurality of telecommunications devices,

determine, based on the received request, an authentication security level, out of a plurality of candidate authentication security levels, for the request, and

place a speaker verification call, after receiving the request and determining the authentication security level, to the speech recognition and verification services engine using a particular access number, the particular access number being selected based on the determined authentication security level,

where the speech recognition and verification services engine is to;

receive the speaker verification call from the voice response system,

perform speaker verification on the received speaker verification call based on the stored user voice print profiles, and

generate a verification score based upon results of the speaker verification.

2. The system of claim 1, where the speech recognition and verification services engine passes the verification score to the voice response system.

3. The system of claim 2, where the voice response system determines whether the transaction or access should be allowed based upon the received verification score and the determined authentication security level.

4. The system of claim 2, where the speech recognition and verification services engine passes the speaker verification call to the voice response system following generating of the verification score.

5. The system of claim 1, where the network is a Public Switched Telephone Network (PSTN).

6. The system of claim 5, where the voice response system remains in a call path following placing of the speaker verification call to the speech recognition and verification services engine.

7. The system of claim 5, where the speech recognition and verification services engine passes the verification score to the voice response system via discrete multitone frequency (DTMF) tones.

8. The system of claim 1, where the speech recognition and verification services engine is associated with a party not affiliated with the voice response system.

9. The system of claim 1, where the voice response system determines an authentication security level for each transaction or access request received.

10. The system of claim 1, where the speech recognition and verification services engine is associated with three distinct access numbers associated with three authentication security levels.

11. The system of claim 9, where the network is a packet switched network (PSN), and calls between the plurality of telecommunication devices and the voice response system are voice over Internet Protocol (VoIP) calls.

12. The system of claim 11, where

the speech recognition and verification services engine receives the speaker verification call and the authentication security level from the voice response system;

and

the speech recognition and verification services engine performs speaker verification based upon the determined authentication security level.

13. The system of claim 12, where

the speaker verification call is passed to the speech recognition and verification services engine using real time protocol (RTP);

and

the authentication security level is passed to the speech recognition and verification services engine via Session Initiation Protocol (SIP).

14. The system of claim 9, where the verification score is passed from the speech recognition and verification services engine to the voice response system via Session Initiation Protocol (SIP).

15. The system of claim 1, where the particular access number is selected from a plurality of candidate access numbers.

16. A system for performing speaker verification, comprising:

a speech recognition and verification services engine operatively coupled to a first network; and

a memory, operatively coupled to the speech recognition and verification services engine, to store user voice print profiles,

where the speech recognition and verification services engine is to;

receive a speaker verification call from a voice response engine over the first network;

receive an authentication security level and a call indicator key associated with the speaker verification call from the voice response engine, the authentication security level being determined by the speech recognition engine before the speech recognition and verification services engine receives the authentication security level, where the authentication security level is one of a plurality of candidate authentication security levels;

perform speaker verification on the received speaker verification call based on the stored user voice print profiles and the authentication security level associated with the speaker verification call;

generate a verification score based upon results of the speaker verification; and

match the verification score to the call using the call indicator key.

17. The system of claim 16, where the first network is a Public Switched Telephone Network (PSTN).

18. The system of claim 16, where the speech recognition and verification services engine receives the authentication security level over a packet switched network.

19. The system of claim 16, where the speech recognition and verification services engine passes the verification score to the voice response engine via a second network.

20. The system of claim 16, where

the voice response engine passes the authentication security level and the call indicator key associated with the speaker verification call to the speech recognition and verification services engine via a second network;

and

the speech recognition and verification services engine matches the received authentication security level with the speaker verification call using the call indicator key.

21. The system of claim 20, where

the speech recognition and verification services engine passes the verification score and the call indicator key to the voice response engine via the second network.

22. A method comprising:

receiving, at a network device, a call that includes a transaction or access request from one of a plurality of telecommunications devices over a first network;

passing the call to a speech recognition and verification services engine over the first network;

transmitting a call indicator key and an authentication level to the speech recognition and verification services engine over a second network, the second network being different than the first network;

receiving the call, over the first network, and the call indicator key, over the second network, by the speech recognition and verification services engine;

performing speaker verification on the received call based on user voice print profiles and the authentication level;

generating a verification score based upon results of the speaker verification;

transmitting, by the speech recognition and verification services engine, the call indicator key and the verification score to the network device;

matching, by the network device, the verification score to the call using the call indicator key; and

one of permitting the request or denying the request based on the verification score.

23. The method of claim 22, where the first network is the Public Switched Telephone Network (PSTN).

24. The method of claim 22, where transmitting the verification score includes:

transmitting the verification score via discrete multitone frequency (DTMF) tones.

25. The method of claim 22, comprising:

determining, by the network device, that the authentication security level is associated with the call, where the authentication security level is one of a plurality of authentication security levels.

26. The method of claim 25, comprising:

calling, by the network device, the speech recognition and verification services engine at an access number associated with the determined authentication security level.

27. The method of claim 25, where at least one of the first and second networks is a packet switched network (PSN), and

where the call between the one of the plurality of telecommunication devices and the entity receiving the call is a voice over Internet Protocol (VoIP) call.

28. The method of claim 22, further comprising:

passing the call to the speech recognition and verification services engine using real time protocol (RTP); and

passing the authentication security level to the speech recognition and verification services engine via Session Initiation Protocol (SIP).

29. A method for performing speaker verification, comprising:

receiving a call from a voice response system over a first network;

receiving an authentication security level and a call indicator key associated with the call from the voice response system, where the received authentication security level is one of a plurality of candidate authentication security levels, and is determined by the voice response system before the authentication security level is received from the voice response system;

performing speaker verification on the received call based on stored user voice print profiles and the authentication security level;

generating a verification score based upon results of the speaker verification; and

matching the verification score to the call using the call indicator key.

30. The method of claim 29, where the first network is a Public Switched Telephone Network (PSTN) and the authentication security level is received via a packet switched network.

31. The method of claim 29, further comprising:

passing the verification score to the voice response system via a second network that is different from the first network.

32. The method of claim 29, further comprising:

passing the verification score and the call indicator key to the voice response system via a second network, the second network being different than the first network.

33. A non-transitory computer-readable medium comprising instructions for controlling one or more processors to perform a method for performing speaker verification over a network, the method comprising:

receiving a call, an authentication security level associated with the call, and a call indicator key associated with the call from a voice response system via the network, where the authentication security level is one of a plurality of candidate authentication security levels;

performing speaker verification on the received call based on user voice print profiles and based on the received authentication security level;

generating a verification score based upon results of the speaker verification;

passing the call indicator key and the verification score to the voice response system for access determination after performing the speaker verification and generating the verification score; and

matching the verification score to the call using the call indicator key.