Voice print identification for identifying speakers

• US 8,606,579 B2
• Filed: 05/24/2010
• Issued: 12/10/2013
• Est. Priority Date: 05/24/2010
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Rasmussen, David John
• Primary Examiner(s): Opsasnick, Michael N
• Application Number:
  US12/786,259
• Publication Number:
  US 20110288866A1
• Time in Patent Office: 1,296 Days
• Field of Search: 704/246
• US Class Current: 704/246
• CPC Class Codes: G06Q 10/063114 Status monitoring or status determination for a person or group
• CPC Class Codes: G06Q 10/109 Time management, e.g. calendars, reminders, meetings or time accounting
• CPC Class Codes: G06Q 50/01 Social networking
• CPC Class Codes: G10L 17/04 Training, enrolment or model building
• CPC Class Codes: G10L 17/06 Decision making techniques; Pattern matching strategies
• CPC Class Codes: G10L 17/10 Multimodal systems, i.e. based on the integration of multiple recognition engines or fusion of expert systems
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces
• CPC Class Codes: H04L 12/1831 Tracking arrangements for later retrieval, e.g. recording contents, participants activities or behavior, network status

Abstract

Voice print identification for identifying speakers may be provided. A plurality of speakers are recorded and associated with identity indicators. Voice prints for each speaker are associated with the plurality of recorded speakers. If the voice print for at least one speaker corresponds to a known user according to the identity indicators, a database entry associating the user with the voice print may be created. Additional information associated with the user may also be displayed.

Claims

1. A method for providing voice print identification, the method comprising:

recording a plurality of speakers at an event;

associating each of the plurality of speakers with at least one identity indicator;

creating a plurality of voice prints, each associated with at least one of the plurality of speakers;

determining whether at least one of the plurality of voice prints corresponds to a user according to the at least one identity indicator;

in response to determining that the at least one of the plurality of voice prints corresponds to the user, creating a database entry associated with the user comprising the corresponding at least one of the plurality of voice prints;

in response to determining that the at least one of the plurality of voice prints does not correspond to the user, determining whether the at least one of the plurality of voice prints is associated with less than a threshold number of possible users; and

in response to determining that the at least one of the plurality of voice prints is associated with less than a threshold number of a plurality of possible users, displaying a request to a host user to associate the at least one of the plurality of voice prints with a new known user from the plurality of possible users.

2. The method of claim 1, wherein the at least one identity indicator comprises at least one of the following:

a location of the recording, a time of the recording, a caller ID, a device identifier, the host user, a calendar entry associated with the event, a recorded image, and a voice to text segment.

3. The method of claim 1, further comprising:

in response to determining that the at least one of the plurality of voice prints corresponds to the user, displaying at least one information element associated with the user.

4. The method of claim 3, wherein the at least one information element comprises at least one of the following:

a list of previous meetings with the corresponding user, contact information associated with the corresponding user, an email associated with the corresponding user, a document associated with the corresponding user, and social network information associated with the corresponding user.

5. The method of claim 1, further comprising:

creating an event presence database entry for each of the plurality of speakers, wherein each database entry comprises a voice print associated with each of the plurality of speakers and an event identifier associated with the event.

6. The method of claim 5, further comprising:

creating an event attendee database entry comprising the event identifier associated with the event and a list of users at the event.

7. The method of claim 6, wherein the event identifier associated with the event and the list of users at the event are identified according to a calendar entry associated with the event.

8. The method of claim 6, further comprising:

creating an identity indicator database entry comprising the event identifier and the at least one identity indicator associated with each of the plurality of speakers.

9. The method of claim 8, wherein determining whether at least one of the plurality of voice prints corresponds to the user according to the at least one identity indicator comprises:

correlating a plurality of speaker database entries, a plurality of event database entries, and a plurality of identifier database entries; and

determining whether the at least one of the plurality of voice prints is associated with a unique user according to the correlated plurality of speaker database entries, plurality of event database entries, and the plurality of identifier database entries.

10. A computer-readable storage device which stores a set of instructions which when executed performs a method for providing voice print identification, the method executed by the set of instructions comprising:

generating a plurality of voice prints from an event recording, wherein each of the plurality of voice prints is associated with an attendee user at the event;

creating a new event presence database entry for at least one of the plurality of voice prints, wherein the new event presence database entry comprises the at least one of the plurality of voice prints, an event identifier associated with the event, and an identity indicator associated with the event;

determining whether the at least one of the plurality of voice prints is associated with at least one prior event presence database entry; and

in response to determining that the at least one of the plurality of voice prints is associated with the at least one prior event presence database entry;

correlating the identity indicator associated with the at least one prior event presence database entry with the identify indicator associated with the new event presence database entry,

determining whether the correlated identity indicators uniquely identify a user associated with the at least one of the plurality of voice prints,

in response to determining whether the correlated identity indicators uniquely identify the user associated with the at least one of the plurality of voice prints, displaying at least one information element associated with the uniquely identified user;

in response to determining that the at least one of the plurality of voice prints does not correspond to the user, determining whether the at least one of the plurality of voice prints is associated with less than a threshold number of possible users; and

in response to determining that the at least one of the plurality of voice prints is associated with less than a threshold number of a plurality of possible users, displaying a request to a host user to associate the at least one of the plurality of voice prints with a particular user from the plurality of possible users.

11. The computer-readable storage device of claim 10, further comprising:

assigning a confidence value for the uniquely identified user.

12. The computer-readable storage device of claim 11, further comprising:

determining whether a voice print associated with a current speaker and the identity indicator associated with the current speaker match the voice print and the identity indicator associated with the uniquely identified user, and

in response to determining that the voice print associated with the current speaker and the identity indicator associated with the current speaker match the voice print and the identity indicator associated with the uniquely identified user, increasing the confidence value associated with the uniquely identified user.

13. The computer-readable storage device of claim 10, wherein the new event presence database entry and the at least one prior event presence database entry are each associated with a single host user.

14. The computer-readable storage device of claim 10, wherein the new event presence database entry and the at least one prior event presence database entry are each associated with a different host user.

15. The computer-readable storage device of claim 14, wherein the host user associated with the new event presence database entry and the host user associated with the at least one prior event presence database entry are both associated with a single organization.

16. The computer-readable storage device of claim 10, wherein the identity indicator comprises at least one of the following:

a location of the recording, a time of the recording, a caller ID, a device identifier, the host user, a calendar entry associated with the event, a recorded image, and a voice to text segment.

17. The computer-readable storage device of claim 16, further comprising:

determining whether the location of the recording is not associated with the uniquely identified user;

in response to determining that the location of the recording is not associated with the uniquely identified user, determining whether one of the plurality of voice prints is not associated with the uniquely identified user;

in response to determining that one of the plurality of voice prints is not associated with the uniquely identified user, determining whether the location of the recording is associated with a single user; and

in response to determining that the location of the recording is associated with the single user, associating the one of the plurality of voice prints with the single user.

18. The computer-readable storage device of claim 10, wherein the at least one information element comprises at least one of the following:

a list of previous meetings with the corresponding user, contact information associated with the corresponding user, an email associated with the corresponding user, a document associated with the corresponding user, and social network information associated with the corresponding user.

19. A system for providing voice print identification, the system comprising:

a memory storage; and

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

record a voice stream associated with an event via a microphone coupled to the system,

isolate a plurality of speakers associated with the voice stream,

generate a plurality of voice prints, each associated with one of the plurality of speakers associated with the voice stream,

determine whether at least one voice print is associated with a known user, wherein being operative to determine whether the at least one voice print is associated with the known user comprises being operative to determine whether a known user entry in a known user database comprising a plurality of users associated with an organization comprises the at least one voice print,

in response to determining that the at least one voice print is associated with a known user, display at least one informational element associated with the known user, wherein the at least one informational element comprises at least one of the following;

an e-mail associated with the known user, a document associated with at least one prior meeting with the known user, a speaker identification of the known user, and a contact data associated with the known user,

in response to determining that the at least one voice print is not associated with the known user;

create a presence entry in a presence database comprising the at least one voice print, a location associated with the voice stream, a time associated with the voice stream, a device identifier associated with the system, a host user identifier associated with the system, and at least one ambient indicator associated with the voice stream, and

create an attendee entry in an attendee database comprising the host user identifier associated with the system, an event identifier associated with a calendar entry associated with the event, a beginning time associated with the event, an end time associated with the event, and at least one expected user associated with the event,

periodically correlate a plurality of presence entries in the presence database with a plurality of attendee entries in the attendee database, wherein the plurality of presence entries and the plurality of attendee entries are associated with the plurality of users associated with the organization, and

create a new known user entry in the known user database comprising the at least one voice print according to a periodic correlation of the plurality of presence entries with the plurality of attendee entries,

wherein the periodic correlation includes determining whether the at least one voice print is associated with less than a threshold number of possible users, and

in response to determining that the at least one of the plurality of voice prints is associated with less than a threshold number of a plurality of possible users, displaying a request to a host user to associate the at least one of the plurality of voice prints with the known user from the one of the plurality of attendee entries.

20. The system of claim 19, further comprising:

assigning a confidence value for the known user.