Enhanced interface for use with speech recognition

• US 8,583,439 B1
• Filed: 01/12/2004
• Issued: 11/12/2013
• Est. Priority Date: 01/12/2004
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Verizon Services Corporation (Verizon Communications Inc.)
• Inventors: Kondziela, James Mark
• Primary Examiner(s): Colucci, Michael
• Application Number:
  US10/755,502
• Time in Patent Office: 3,592 Days
• Field of Search: 704/258, 704/260, 704/270, 704/275, 704/200, 704/257, 704/241, 704/271, 704/268, 704/267, 704/256, 704/253, 381/94.3, 379/88.23, 379/189
• US Class Current: 704/267
• CPC Class Codes: G10L 13/033 Voice editing, e.g. manipulating the voice of the synthesiser
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 21/16 Transforming into a non-visible representation devices or methods enabling ear patients to replace direct auditory perception by another kind of perception A61F11/04

Abstract

Improved methods of presenting speech prompts to a user as part of an automated system that employs speech recognition or other voice input are described. The invention improves the user interface by providing in combination with at least one user prompt seeking a voice response, an enhanced user keyword prompt intended to facilitate the user selecting a keyword to speak in response to the user prompt. The enhanced keyword prompts may be the same words as those a user can speak as a reply to the user prompt but presented using a different audio presentation method, e.g., speech rate, audio level, or speaker voice, than used for the user prompt. In some cases, the user keyword prompts are different words from the expected user response keywords, or portions of words, e.g., truncated versions of keywords.

Claims

1. A voice interface method comprising:

operating a device including a processor to generate an audible user prompt to solicit input from a user, the audible user prompt including words presented using a first rate of speech, and

prior to receiving a user response to said user prompt, generating an audible keyword prompt indicating at least some keywords which may be spoken by a user in response to said user prompt, the audible keyword prompt including speech that has at least one feature, including overlapping words, wherein said overlapping words comprise two or more voice keyword prompts that are mixed together to be at least partially overlapped.

2. The method of claim 1, wherein said step of generating an audible keyword prompt includes generating speech that has truncated words.

3. The method of claim 1, wherein said step of generating an audible keyword prompt includes the step of:

processing recordings of different keywords to produce processed versions of said keywords having said at least one feature.

4. The method of claim 3, wherein the step of generating an audible keyword prompt includes:

combining speech corresponding to different speakers to form said keyword prompt.

5. The method of claim 4, wherein said step of combining speech corresponding to different speakers includes overlapping the speech of different people, the overlapped speech of at least two different people corresponding to different keywords.

6. The method of claim 1, wherein the step of generating an audible keyword prompt includes synthesizing speech having said at least one feature.

7. The method of claim 1,

wherein operating a device including a processor to generate an audible user prompt includes presenting speech at a first rate;

and wherein generating an audible keyword prompt includes generating speech at a second rate which is faster than said first rate.

8. The method of claim 1,

wherein generating an audible user prompt includes presenting speech at a first volume level;

and wherein generating an audible keyword prompt includes generating speech at a second volume level which is lower than said first volume level.

9. The method of claim 1,

wherein generating an audible keyword prompt includes generating speech which is a truncated version of at least one keyword.

10. The method of claim 1,

wherein generating an audible keyword prompt includes generating speech corresponding to first and second keywords, at least a portion of the generated speech corresponding to the first and second keywords being presented to the user at the same time resulting in overlapping speech.

11. The method of claim 10, wherein generating speech corresponding to first and second keywords includes:

accessing a first speech recording corresponding to a first keyword, said first speech recording being a recording of speech from a first speaker;

accessing a second speech recording to a second keyword, said second speech recording being a recording of speech from a second speaker, the second speaker being different from said first speaker; and

playing said first and second speech recordings so that at least some of the generated audio corresponding to the first and second speech recordings overlaps in time.

12. The voice interface method of claim 1, wherein said step of operating a device including a processor to generate an audible keyword prompt is performed for at least some portion of time in parallel with said step of generating a first audible user prompt thereby resulting in an overlapping first audible user prompt and audible keyword prompt.

13. The voice interface method of claim 12, wherein said step of generating an audible keyword prompt includes:

generating speech indicating keywords at a lower audible level than speech corresponding to said audible user prompt.

14. The voice interface method of claim 13, wherein said audible user prompt is a voice prompt presenting at least a portion of a request for user input from a user of said voice interface.

15. The voice interface method of claim 1, wherein the audible keyword prompt has the feature of truncated words.

16. The voice interface method of claim 1, wherein the audible keyword prompt has the feature of a first word which is spoken by a first individual speaker and a second word which is spoken by an individual speaker different from the first individual speaker.

17. The method of claim 2, wherein said step of generating an audible keyword prompt includes generating speech that has words which are generated at an audible sound level which is different than an audible sound level at which words in said audible user prompt are generated.

18. The method of claim 1 wherein said two or more voice keyword prompts are mixed together so that they are fully overlapped.

19. The method of claim 1 wherein said two voice keyword prompts are mixed together such that the amount of overlap between the two voice keyword prompts is minimized.

20. The method of claim 1 wherein the keyword prompt starts playing before the user prompt has completed.

21. A method of presenting keyword information to a user of a voice interface, the method comprising:

operating a device including a processor to generate an audible keyword prompt indicating at least some keywords which may be spoken by a user in response to said user prompt, the audible keyword prompt including speech that has overlapping words, wherein said overlapping words comprise two or more voice keyword prompts that are mixed together to be at least partially overlapped.

22. The method of claim 21, wherein said generated audible keyword prompt also includes speech that includes truncated words.

23. The method of claim 21, wherein said generated audible keyword prompt includes overlapping words.

24. The method of claim 21, wherein said generated audible keyword prompt includes words spoken by different speakers.

25. The method of claim 24, wherein said step of generating an audible keyword prompt includes:

accessing a first speech recording corresponding to a first keyword, said first speech recording being a recording of speech from a first speaker;

accessing a second speech recording to a second keyword, said second speech recording being a recording of speech from a second speaker, the second speaker being different from said first speaker; and

playing said first and second speech recordings so that at least some of the generated audio corresponding to the first and second speech recordings overlaps in time.

26. A voice interface system comprising:

processor hardware for generating an audible user prompt to solicit input from a user, the audible user prompt including words presented using a first rate of speech,

means for generating, prior to receiving a user response to said audible user prompt, an audible keyword prompt indicating at least some keywords which may be spoken by a user in response to said user prompt, the audible keyword prompt including speech that has at least one feature, including overlapping words, wherein said overlapping words comprise two or more voice keyword prompts that are mixed together to be at least partially overlapped.

27. The voice interface system of claim 26, wherein the processor hardware includes:

a speech synthesizer capable of generating speech including truncated words.

28. The voice interface system of claim 26, wherein the processor hardware includes:

recordings of speech corresponding to at least two keywords.

29. The voice interface system of claim 28, wherein the processor hardware includes:

processing circuitry for adding said at least one feature to the recording of said at least two keywords as part of said step of generating an audible keyword prompt.

30. A voice interface method comprising:

operating a device including a processor to generate an audible user prompt to solicit input from a user, the audible user prompt including words presented using a first rate of speech, and

prior to receiving a user response to said user prompt, generating an audible keyword prompt indicating at least some keywords which may be spoken by a user in response to said user prompt, the audible keyword prompt including speech that has at least one word which is not fully pronounced.

31. A voice interface method comprising:

operating a device including a processor to generate an audible user prompt to solicit input from a user, the audible user prompt including words presented using a first rate of speech, and

prior to receiving a user response to said user prompt, generating an audible keyword prompt indicating at least one keyword which may be spoken by a user in response to said audible user prompt, the audible keyword prompt including speech that is generated at an audible sound level which is different than an audible sound level at which words in said audible user prompt are generated and at a rate which is different than the first rate of speech.