System and method of templating specific human voices

• US 20020072900A1
• Filed: 10/04/2001
• Published: 06/13/2002
• Est. Priority Date: 11/23/1999
• Status: Abandoned Application

• Current Assignee: Voice Odyssey, Inc.
• Original Assignee: Voice Odyssey, Inc.
• Inventors: Keough, Steven J., Keough, Katherine Axia
• Application Number:
  US09/972,730
• Publication Number:
  US 20020072900A1
• Time in Patent Office: Days
• US Class Current: 704/220
• CPC Class Codes: G10L 13/033 Voice editing, e.g. manipulating the voice of the synthesiser
• CPC Class Codes: G10L 2021/0135 Voice conversion or morphing

Abstract

Systems and methods are disclosed to capture an enabling portion of a voice and then to create a voice template or profile signal which may be combined at a later time with noise of another origin to reconstitute the original voice. Such reconstituted voice may then be used to speak any form or content provided via digital input thereto, and to say content which was not spoken in an original form by the original voice. Products and processes for online use are disclosed, as are certain business methods and industry applications.

Claims

1. A system for capturing an enabling portion of a specific voice sufficient for using that portion as a template in further use of the voice, comprising:

a. means for capturing an enabling portion of a voice in a form useful for analysis as to voice characteristics;

b. analysis means for receiving and analyzing the captured voice and for characterizing elements of the captured voice as characterization data;

c. storage means for receiving characterization data from the analysis means for a specific voice; and

d. retrieval means for retrieving the analysis and characterization data for further use.

2. The system of claim 1 in which the means for capturing the voice comprises digital recording means.

3. The system of claim 1 in which the means for capturing the voice comprises a flash memory card.

4. The system of claim 1 in which the means for capturing the voice comprises analog recording means.

5. The system of claim 1 in which the means for capturing the voice comprises input means for receiving a live voice and for transmitting that live voice to the analysis means.

6. The system of claim 1 in which the analysis means comprises digital data storage means.

7. The system of claim 1 in which the analysis means comprises means for identifying specific patterns, syntax, frequency, pitch and tones of speech in the captured voice data.

8. The system of claim 1 in which the analysis means comprises means for identifying specific vocabulary, pronunciation, or accent unique to the captured voice.

9. The system of claim 1 in which the analysis means comprises means for identifying specific features unique to the captured voice deriving principally from specific anatomic structures of the originator of the voice.

10. The system of claim 1 in which the analysis means comprises means for determining the vocabulary of the originator of the captured voice.

11. The system of claim 10 in which the analysis means comprises means for setting the vocabulary as characterization data for use in forming a future templated voice.

12. The system of claim 1 in which the analysis means comprises digital processing apparatus for digitally processing input data in the form of a voice or digital representation of a recorded voice.

13. The system of claim 1 in which the analysis means comprises second input means for receiving additional data regarding the physiology of the voice originator.

14. The system of claim 13 in which the analysis means second input means comprises digital signal processor means suitable for selectively receiving audio or other data comprising visualization information on the morphology of the voice originator.

15. The system of claim 1 in which the analysis means comprises comparison means for comparing an input voice data set with stored data comprising age data, language data, educational data, gender data, occupation data, accent data, nationality data, ethnic data, voice type data, custom data and setting data.

16. The system of claim 1 in which the analysis means comprises third input means for receiving data regarding the voice originator comprising age data, educational data, gender data, occupation data, accent data, nationality data, ethnic data, voice type data, custom data, language data and setting data.

18. The method of claim 17 in which the analyzing step comprises the steps of identifying the components in the captured enabling portion of the specific human'"'"'s voice relating to at least one of the components including frequency, tone, pitch, volume, accent, gender, harmonic structure, acoustic power, phonetic or timing accent, power and periodicity.

19. The method of claim 18 in which the step of capturing an enabling portion of a specific human'"'"'s voice for storage and use includes capturing either larynx generated noise or turbulence generated noise of the specific human'"'"'s voice.

17. A method of creating a voice-like noise which is identical in sound to an actual specific human'"'"'s voice, comprising the steps of:

a. capturing an enabling portion of a specific human'"'"'s voice for storage and use;

b. storing the enabling portion of the specific human'"'"'s voice;

c. analyzing the enabling portion to identify essential components or characteristics of the captured voice; and

d. utilizing the identified essential components or characteristics to create a new voice which, when assigned data from one or more database means and when heard, sounds identical in all respects to the voice of the specific human'"'"'s voice to a listener having normal aural discretion abilities.

20. A method of accurately replicating a human voice comprising the steps of:

a. identifying a minimum size data set comprising a combination of words, sounds or phrases which must be emitted by the originator of a voice to be replicated;

b. capturing the emission of the combination of words, sounds or phrases by the originator of the voice to be replicated in a medium;

c. analyzing the captured emission to identify voice characteristics of the originator of the voice sufficient to allow artificial generation of the voice, using the identified characteristics, so that the artificially generated voice is substantially identical in all respects to a listener having normal aural discretion abilities when the listener hears the generated voice utilizing some language components not contained in the captured emission of the originator'"'"'s actual voice.

21. An article of manufacture comprising:

a. a computer usable medium having computer readable program code means embodied therein for causing replication of a human voice, the computer readable program code means in said article of manufacture comprising;

b. computer readable program code means for causing a computer to effect an analysis of a captured enabling portion of an originator'"'"'s voice to identify voice characteristics data sufficient to allow artificial generation of the voice; and

c. computer readable program code means for causing use of the identified voice characteristics data to artificially generate a voice, so that the artificially generated voice is substantially identical in sound and usage to a listener when the listener hears the generated voice utilizing some language components not contained in the captured emission of the originator'"'"'s actual voice.

22. The article of manufacture of claim 21 further comprising computer readable program code means for storing the generated voice for later use.

23. The article of manufacture of claim 21 further comprising computer readable program code means for using the voice characteristics data to create a voice profile of the originator of the voice.

24. The article of manufacture of claim 21 further comprising computer readable program code means for accessing data base means for storing data comprising age data, educational data, gender data, occupation data, accent data, language, nationality data, ethnic data, voice type data, custom data, general data and setting data.

25. A computer program product for use with an aural output device, said computer program product comprising:

a. a computer usable medium having computer readable program code means embodied therein for causing replication of a human voice via an output aural device, the computer program product comprising;

b. computer readable program code means for causing a computer to effect an analysis of a captured enabling portion of an originator'"'"'s voice to identify voice characteristics data sufficient to allow artificial generation of the voice; and

c. computer readable program code means for causing use of the identified voice characteristics data to artificially generate and output a voice via an aural output device, so that the artificially generated voice is substantially identical in sound and usage to a listener when the listener hears the generated voice utilizing some language components not contained in the captured emission of the originator'"'"'s actual voice.

26. A computer program product for use with a display device, said computer program product comprising:

a. a computer usable medium having computer readable program code means embodied therein for causing replication of a human voice and verification of the accuracy of the replicated voice displayed on the display device, the computer program product comprising;

d. computer readable program code means for causing a computer to effect an analysis of a captured enabling portion of an originator'"'"'s voice to identify voice characteristics data sufficient to allow artificial generation of the voice; and

e. computer readable program code means for causing use of the identified voice characteristics data to artificially generate a voice and to compare the characteristics of the generated voice to the originator'"'"'s voice on a display device, so that the artificially generated voice is substantially identical in sound to a listener when the display device so indicates and when a listener actually hears the generated voice utilizing some language components not contained in the captured emission of the originator'"'"'s actual voice.

27. A computer program product for use with an aural output device, said computer program product comprising:

a. a computer usable medium having computer readable program code means embodied therein for initiating replication of a human voice via an output aural device, the computer program product comprising;

b. computer readable program code means for causing a computer to receive and activate a voice characteristics data file unique to a specific voice sufficient to allow artificial generation of the voice; and

c. computer readable program code means for causing use of the identified voice characteristics data to artificially generate and output a voice via an aural output device, so that the artificially generated voice is substantially identical in sound to a listener when the listener hears the generated voice and a captured emission of the originator'"'"'s actual voice.

28. A computer program product for use with an electronic device, said computer program product comprising:

a. a computer usable medium having computer readable program code means embodied therein for initiating replication of a human voice, the computer program product comprising;

b. computer readable program code means for causing receipt and activation of a voice characteristics data file unique to a specific voice sufficient to allow artificial generation of the voice; and

c. computer readable program code means for causing use of the identified voice characteristics data file and a noise generation means sound output to artificially generate a voice, so that the artificially generated voice is substantially identical in sound to the originator'"'"'s actual voice.

29. A memory for storing data for access by an application program being executed on a data processing sub-system, comprising:

a. a data structure stored in said memory, said data structure including information resident in a database used by said application program and including;

b. at least one voice enabling portion data file stored in said memory, each of said voice enabling portion data file set containing information substantially different from any other voice enabling portion data file set;

c. a plurality of voice characteristics data files containing different reference information for a plurality of voice characteristics; and

d. a plurality of voice profile sets each having at least one voice profile data file having data unique to that data file only;

wherein the data structure allows access to the voice characteristics data files and the voice profile data files to conduct comparison operations with at least one voice enabling portion data file.

30. A data processing system executing an application program and containing a database used by said application program, said data processing system comprising:

a. CPU means for processing said application program; and

b. memory means for holding a data structure for access by said application program, said data structure being composed of information resident in a database used by said application program and including;

at least one voice enabling portion data file stored in said memory, each of said voice enabling portion data file set containing information substantially different from any other voice enabling portion data file set;

a plurality of voice characteristics data files containing different reference information for a plurality of voice characteristics;

a plurality of voice profile sets each having at least one voice profile data file having data unique to that data file only; and

c. wherein the data processing system allows access to the voice characteristics data files and the voice profile data files to conduct comparison operations with at least one voice enabling portion data file.

31. A computer data signal embodied in a transmission medium comprising:

a. an encryption source code for a unique voice profile template useful for keying additional electronic noise to create a specific generated voice; and

b. a carrier medium suitable for carrying the encryption source code to a location and configured so that the encryption source code is removable from the carrier medium to be applied as a key to create a generated voice.

32. A method for using a selected voice as a personal voice assistant with an electronic device, comprising the steps of:

a. activating electronic means for accessing a remote database;

b. transmitting a signal portion to a remote database having a voice database containing a plurality of voice profile sets each having at least one voice profile data file having data unique to that data file only and identifiable by a unique identifier;

C. transmitting a signal portion to the remote database to uniquely identify a desired data file and then to effect transfer of the data file content to the user'"'"'s designated electronic device location; and

d. implementing use of the selected and transferred data file as a voice template, in combination with appropriate noise generated either by the electronic device or other means for generating such noise, so that as desired the user may receive noise from the electronic device in the sound of the selected voice as determined by the identified voice.

33. The method of claim 32 in which the data file includes data characteristics of the selected voice arranged as computer readable program code means for causing use of the identified voice characteristics data to artificially generate a voice template.

34. The method of claim 32 in which the implementing step comprises application of authorization means to only allow authorized users to access and use the voice template technology and data.

35. The method of claim 32 in which the implementing step comprises application of selectively accessible verification means for verifying that voices heard are either real or template generated.

37. The method of claim 36 in which the means for capturing the voice comprises digital input means.

38. The method of claim 36 in which the enabling portion of the voice is received electronically.

39. The method of claim 36 in which the characterization data is bundled to form a voice template signal useful for combining with generated noise to create a templated voice which sounds like the original specific voice.

40. The method of claim 36 in which the templated voice is controlled so that the templated voice may receive speech input commands to elicit new words in the templated voice but which were not inputted by the specific voice.

36. A method of doing business in which a system is used for capturing an enabling portion of a specific voice sufficient for using that portion as a template in further use of the voice, comprising the steps of:

a. capturing an enabling portion of a voice in a form useful for analysis as to voice characteristics;

b. inputting the enabling portion into an analysis module for characterizing elements of the captured voice as characterization data;

c. receiving the characterization data from the analysis module for a specific voice; and

d. storing the characterization data for further use.

41. An automated machine for capturing an enabling portion of a specific voice and for using that portion as a template useful for further use of the templated voice, comprising:

a. an acquisition module for acquiring an enabling portion of a voice in a form useful for analysis as to voice characteristics;

b. an analysis module for receiving and analyzing the captured voice and for characterizing elements of the captured voice as characterization data; and

c. a template generator module for automatically generating a voice template signal as a unique identifier of the acquired specific voice.

42. The machine of claim 41 further comprising communication means for communicating with storage means for receiving characterization data from a database.

43. The machine of claim 41 further comprising communication means for communicating with storage means for storing the generated template until requested.

44. An online method for creating voice templates and generating revenue for such generation, comprising:

a. capturing an enabling portion of a specific voice;

b. analyzing the enabling portion of the specific voice to generate a data profile which defines the characteristics of the captured voice in a way that can be reconstituted for later use;

c. generating a voice template signal as a unique identifier of the acquired specific voice; and

d. providing at least one generated data profile for commercial use by another.

47. The method of doing business of claim 46 in which the step of providing is accomplished on an electronic data exchange.

50. The method of claim 49 further comprising identifying the voice to be replicated by genetic code.

51. The method of claim 49 further comprising a step of validating or adjusting the artificially generated voice by use of genetic code analysis of the originator of the voice being replicated.

45. A machine operated method for creating a voice template and generating revenue for such generation, comprising:

a. capturing an enabling portion of a specific voice;

b. analyzing the enabling portion of the specific voice to generate a data profile which defines the characteristics of the captured voice in a way that can be reconstituted for later use;

c. using the data profile, generating a voice template signal as a unique identifier of the captured specific voice; and

d. providing at least one voice template signal for commercial use.

46. A business method for creating a voice template, comprising:

a. capturing an enabling portion of a specific voice or templated voice;

b. using computer means, analyzing the enabling portion of the voice to generate a data profile which defines the characteristics of the captured voice in a way that can be reconstituted for later use;

c. electronically generating or retrieving a voice template signal as a unique identifier of the captured voice; and

d. providing at least one voice template signal for commercial use.

48. A method for creating a voice template from a plurality of voices, comprising:

a. capturing an enabling portion of a plurality of voices or templated voices;

b. using computer means, analyzing the enabling portions of the voices to generate a data profile which defines the characteristics of the captured voices in a way that can be bundled as a single voice signal suitable for reconstitution for later use; and

c. electronically generating a voice template signal as a unique identifier of the newly generated voice.

49. A method of accurately replicating a human voice of someone who lost the ability to speak in the desired normal voice, comprising the steps of:

a. identifying a minimum size data set comprising a combination of words, sounds or phrases which must be emitted by the originator of a voice to be replicated;

b. capturing the emission of the combination of words, sounds or phrases by the originator of the voice to be replicated in a medium;

c. analyzing the captured emission to identify voice characteristics of the originator of the voice sufficient to allow artificial generation of the voice, using the identified characteristics, so that the artificially generated voice is substantially identical in all respects to a listener having normal aural discretion abilities when the listener hears the generated voice utilizing some language components not contained in the captured emission of the originator'"'"'s actual voice.

52. A method of accurately replicating an actual human voice, comprising the steps of:

a. identifying a minimum size data set comprising a combination of fractions or segments of actual words, sounds or phrases which were emitted by the originator of the voice to be replicated;

b. capturing the emission of the combination of words, sounds or phrases by the originator of the voice to be replicated in a medium;

c. analyzing the captured emission to identify voice characteristics of the originator of the voice by analysis of fractions or segments of the words, sounds or phrases sufficient to allow artificial generation of the voice, using the identified characteristics, so that the artificially generated voice is substantially identical in all respects to a listener having normal aural discretion abilities when the listener hears the generated voice utilizing some language components not contained in the captured emission of the originator'"'"'s actual voice.

53. A method for creating a voice template from a plurality of voice fragments, comprising:

a. capturing an enabling portion of a plurality of voice fragments;

b. using computer means, analyzing the enabling portions of the voice fragments to generate a voice fragment data code which defines the characteristics of the captured voice fragments in a way that can be bundled as a single voice signal suitable for reconstitution for later use; and

c. electronically generating a voice template signal as a unique identifier of the newly generated voice.