System and methods for providing voice transcription

• US 9,871,916 B2
• Filed: 03/05/2009
• Issued: 01/16/2018
• Est. Priority Date: 03/05/2009
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Dingler, John R., Ramanathan, Sri, Terry, Matthew A., Trevathan, Matthew B.
• Primary Examiner(s): Dorvil, Richemond
• Assistant Examiner(s): Le, Thuykhanh
• Application Number:
  US12/398,271
• Publication Number:
  US 20100228546A1
• Time in Patent Office: 3,239 Days
• Field of Search: None
• CPC Class Codes: G09B 21/009 Teaching or communicating with deaf persons G09B21/04 takes precedence; deaf and dumb persons G09B21/00
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 17/00 Speaker identification or verification techniques
• CPC Class Codes: G10L 2021/065 Aids for the handicapped in understanding
• CPC Class Codes: H04L 65/1096 Supplementary features, e.g. call forwarding or call holding systems providing special services or facilities to telephony subscribers H04M3/42
• CPC Class Codes: H04L 65/1104 Session initiation protocol [SIP]
• CPC Class Codes: H04M 2201/40 using speech recognition speech recognition per se G10L15/00
• CPC Class Codes: H04M 2201/60 Medium conversion
• CPC Class Codes: H04M 3/42391 where the subscribers are hearing-impaired persons, e.g. telephone devices for the deaf

Abstract

A system and methods is provided for providing SIP based voice transcription services. A computer implemented method includes: transcribing a Session Initiation Protocol (SIP) based conversation between one or more users from voice to text transcription; identifying each of the one or more users that are speaking using a device SIP_ID of the one or more users; marking the identity of the one or more users that are speaking in the text transcription; and providing the text transcription of the speaking user to non-speaking users.

Claims

1. A computer implemented method comprising:

transcribing a Session Initiation Protocol (SIP) based conversation between one or more users from voice to text transcription;

identifying each of the one or more users that are speaking using a device SIP_ID of the one or more users;

marking the identity of the one or more users that are speaking in the text transcription;

receiving a request to record the conversation, the request comprising a confirmation to use voice recognition algorithms for recording the conversation;

receiving authorization from a non-requesting user to record the conversation;

determining whether the SIP based conversation is accurately transcribed from the voice to the text transcription;

providing an error notification in response to the SIP based conversation not being accurately transcribed from the voice to the text transcription;

providing a notification that the recorded conversation has been accurately transcribed from the voice to the text transcription in response to the SIP based conversation being accurately transcribed from the voice to the text transcription, the text transcription being saved in a database, and access to the text transcription being granted through a hyperlinked address;

providing the text transcription of the speaking user to non-speaking users; and

authenticating a media server with the one or more users that are speaking using the device SIP_ID of the one or more users using a trusted token, and after authenticating the media server, a caller'"'"'s device will send call-in information to the media server.

2. The method of claim 1, further comprising correlating the device SIP_ID with the speaking user.

3. The method of claim 2, wherein the correlating includes identifying a waveform of speech of the speaking user to the device SIP_ID.

4. The method of claim 2, further comprising storing the device SIP_ID in a central repository.

5. The method of claim 1, further comprising storing profile information of the one or more users in a repository in order to distinguish between the speaking and non-speaking users.

6. The method of claim 1, further comprising storing voiceprints of the one or more users in a repository in order distinguish between the speaking and non-speaking users.

7. The method of claim 1, further comprising billing the one or more users for the text transcription using diameter-based protocol.

8. The method of claim 1, wherein the text transcription is provided in real time by sending an instant message.

9. The method of claim 1, further comprising recording the conversation and placing the recorded conversation in a queue for off-line transcription with a message ID and caller SIP_ID.

10. The method of claim 1, further comprising dialing into a call between the one or more users, creating a call record to a start of recording, beginning the recording, and receiving a callee'"'"'s SIP, callee'"'"'s IP, caller'"'"'s SIP and caller SIP_ID.

11. The method of claim 1, wherein the steps are provided on a hardware component or a combination of the software component and the hardware component.

12. The method of claim 1, wherein the error notification is a vibration, beep, or flashing screen.

13. The method of claim 12, further comprising providing the text transcription to one set of the one or more users in real-time and providing the text transcription to another set of the one or more users off-line.

14. The method of claim 1, wherein the error notification informs the one or more users that the SIP based conversation is not accurately transcribed due to background noise.

15. The method of claim 1, wherein the text transcription of the speaking user is provided to the non-speaking users based on a Boolean send value in an XML message.

16. A computer program product comprising a non-transitory computer usable storage medium having readable program code embodied in the storage medium, the computer program product being operable to:

listen into an established conversation between one or more users on an IMS platform;

transcribe the conversation from voice to text transcription;

identify each of the one or more users in the text transcription using a device SIP_ID of the each of the one or more users;

receive a request to record the conversation, the request comprising a confirmation to use voice recognition algorithms for recording the conversation;

receive authorization from a non-requesting user to record the conversation;

determine whether the established conversation is accurately transcribed from the voice to the text transcription;

provide an error notification in response to the established conversation not being accurately transcribed from the voice to the text transcription;

provide a notification that the recorded conversation has been accurately transcribed from the voice to the text transcription in response to the established conversation being accurately transcribed from the voice to the text transcription, the text transcription being saved in a database, and access to the text transcription being granted through a hyperlinked address; and

authenticate a media server with the one or more users that are speaking using the device SIP_ID of the one or more users using a trusted token, and after authenticating the media server, a caller'"'"'s device will send call-in information to the media server.

17. The computer program product of claim 16, wherein the computer program product is further operable to record the conversation between the one or more users and place the conversation in a queue for the transcribing with a message ID and caller SIP_ID.

18. The computer program product of claim 17, wherein the computer program product is further operable to notify the one or more users that the text transcription is available for review.

19. The computer program product of claim 16, wherein the computer program product is further operable to provide the text transcription of a speaking user to a non-speaking user by instant message.

20. The computer program product of claim 16, wherein the conversation is a Session Initiation Protocol (SIP) based conversation.

21. The computer program product of claim 16, wherein the computer program product is further operable:

correlate the device SIP_ID with a speaking user;

store the device SIP_ID and profile information of the one or more user in a repository in order to distinguish between the speaking user and non-speaking users; and

one of;

provide the text transcription in real time by sending a text message or instant message; and

record the conversation and place the recorded conversation in a queue for off-line transcription.

22. The computer program product of claim 16, wherein the computer program product is further operable:

dial into a call between the one or more users, create a call record to a start of the recording, begin the recording, and receive a callee'"'"'s SIP, callee'"'"'s IP, caller'"'"'s SIP and caller SIP_ID.

23. A system implemented on a computer infrastructure having at least hardware components or a combination of software components and the hardware components operable to:

listen into an established conversation between one or more users;

identify each of the one or more users in the conversation using a device SIP_ID of the each of the one or more users;

receive a request to record the conversation, the request comprising a confirmation to use voice recognition algorithms for recording the conversation;

record the conversation between the one or more users and place the conversation in a queue for the transcribing;

transcribe the conversation from voice to text transcription;

mark a speaking user in the text transcription using the device SIP_ID of the each of the one or more users;

determine whether the established conversation is accurately transcribed from the voice to the text transcription;

notify the one or more users that an error has occurred in response to the established conversation not being accurately transcribed from the voice to the text transcription;

notify the one or more users that the text transcription has been accurately transcribed from the voice to the text transcription in response to the established conversation being accurately transcribed from the voice to the text transcription, the text transcription being saved in a database, and access to the text transcription being granted through a hyperlinked address;

provide the text transcription of a speaking user to a non-speaking user by instant message or text message; and

authenticate a media server with the one or more users that are speaking using the device SIP_ID of the one or more users using a trusted token, and after authenticating the media server, a caller'"'"'s device will send call-in information to the media server.

24. The system of claim 23, wherein the at least hardware components or a combination of software components and the hardware components is further operable to:

correlate the device SIP_ID with the speaking user;

store the device SIP_ID in a repository in order to distinguish between the speaking user and non-speaking users;

receive authorization from a non-requesting user to record the conversation; and

provide notification that the recorded conversation has been transcribed from the voice to the text transcription.

25. The system of claim 23, further comprising dialing into a call between the one or more users, creating a call record to a start of the recording, beginning the recording, and receiving a callee'"'"'s SIP, callee'"'"'s IP, caller'"'"'s SIP and caller SIP_ID.

26. A method or transcribing a voice conversation using Session Initiation Protocol (SIP) from voice to text, comprising:

providing a computing infrastructure operable to;

call into an established conversation between one or more users;

identify each of the one or more users that are speaking using a device SIP_ID of the one or more users;

receive a request to record the conversation, the request comprising a confirmation to use voice recognition algorithms for recording the conversation;

receive authorization from a non-requesting user to record the conversation;

transcribe the conversation between one or more users from voice to text transcription;

mark the identity in the text transcription of the one or more users that are speaking;

determine whether the established conversation is accurately transcribed from the voice to the text transcription;

provide an error notification in response to the established conversation not being accurately transcribed from the voice to the text transcription;

provide a notification that the established conversation has been accurately transcribed from the voice to the text transcription in response to the established conversation being accurately transcribed from the voice to the text transcription, the text transcription being saved in a database, and access to the text transcription being granted through a hyperlinked address;

provide the text transcription to the one or more users using instant messaging or text messaging; and

authenticate a media server with the one or more users that are speaking using the device SIP_ID of the one or more users using a trusted token, and after authenticating the media server, a caller'"'"'s device will send call-in information to the media server.

27. The method of claim 26, wherein the computing infrastructure is further operable to:

record the conversation and place the conversation in a queue for the transcribing;

store the device SIP_ID in a repository in order to distinguish between the speaking and non-speaking users; and

provide notification that the recorded conversation has been transcribed from the voice to the text transcription.