Voice profile management and speech signal generation

• US 9,875,752 B2
• Filed: 05/23/2017
• Issued: 01/23/2018
• Est. Priority Date: 04/30/2014
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Sinder, Daniel Jared, Manjunath, Sharath
• Primary Examiner(s): Baker, Charlotte M
• Application Number:
  US15/603,270
• Publication Number:
  US 20170256268A1
• Time in Patent Office: 245 Days
• Field of Search: 704270
• CPC Class Codes: G10L 13/033 Voice editing, e.g. manipulating the voice of the synthesiser
• CPC Class Codes: G10L 17/00 Speaker identification or verification techniques
• CPC Class Codes: G10L 21/003 Changing voice quality, e.g. pitch or formants
• CPC Class Codes: G10L 21/0208 Noise filtering
• CPC Class Codes: G10L 25/48 specially adapted for particular use

Abstract

A device includes a receiver, a memory, and a processor. The receiver is configured to receive a remote voice profile. The memory is electrically coupled to the receiver. The memory is configured to store a local voice profile associated with a person. The processor is electrically coupled to the memory and the receiver. The processor is configured to determine that the remote voice profile is associated with the person based on speech content associated with the remote voice profile or an identifier associated with the remote voice profile. The processor is also configured to select the local voice profile for profile management based on the determination.

Claims

1. A device comprising:

a memory configured to store substitute speech data corresponding to substitute speech signals; and

a processor configured to;

generate search result data via a search of the memory based on input speech data corresponding to an input speech signal;

select particular search result data based on a constraint, wherein the particular search result data is associated with particular substitute speech data; and

generate processed speech data by replacing a portion of the input speech data with replacement speech data, wherein the replacement speech data is determined based on the particular search result data.

2. The device of claim 1, wherein the input speech signal comprises a distorted signal, and wherein the processor is further configured to determine that the input speech data includes distortion.

3. The device of claim 2, wherein the processor is further configured to generate search criteria data based on the input speech data, wherein the search criteria data comprises speech parameter data associated with the input speech data.

4. The device of claim 3, wherein the search result data is generated using the input speech data, the search criteria data, or both, wherein the search result data comprise a set of substitute speech data, and wherein the set of substitute speech data substantially matches the input speech data.

5. The device of claim 1, wherein the substitute speech data comprises pulse code modulation (PCM) waveforms.

6. The device of claim 1, wherein the memory is further configured to store speech parameter data associated with the substitute speech data, and wherein the speech parameter data includes a pitch parameter, an energy parameter, a linear predictive coding (LPC) parameter, mel-frequency cepstral coefficients (MFCC), line spectral pairs (LSP), line spectral frequencies (LSF), a cepstral, line spectral information (LSI), discrete cosine transform (DCT) parameters, pulse code modulation (PCM), discrete fourier transform (DFT) parameters, fast fourier transform (FFT) parameters, formant frequencies, or a combination thereof.

7. The device of claim 1, wherein the processor is further configured to generate the search result data based on a comparison of the input speech data and speech parameter data.

8. The device of claim 1, wherein the processor is further configured to:

copy the input speech data to generate copied speech data;

remove the portion of the input speech data from a particular location of the copied speech data; and

concatenate the replacement speech data and the copied speech data at the particular location to generate the replacement speech data.

9. The device of claim 1, wherein the constraint comprises a cost constraint, an error constraint, or both.

10. A device comprising:

a memory configured to store a plurality of substitute speech data corresponding to a plurality of substitute speech signals; and

a processor electrically coupled to the memory and configured to;

identify at least a portion of speech data, the speech data corresponding to speech signals;

select, based on the substitute speech data, a portion of the substitute speech data that matches the portion of the speech data; and

replace at least the portion of the speech data with the selected portion of the substitute speech data to generate output speech data.

11. The device of claim 10, wherein the processor is further configured to encode the output speech data to generate encoded output speech data, and further comprising a transmitter configured to transmit the encoded output speech data.

12. The device of claim 11, wherein the encoded output speech data comprises packetized speech data, and further comprising an output device configured to project the output speech data, wherein the output device comprises a speaker device.

13. The device of claim 10, further comprising a receiver configured to receive encoded speech data, wherein the processor is further configured to decode the encoded speech data to generate reconstructed speech data, the reconstructed speech data corresponding to the speech data.

14. The device of claim 10, wherein the speech data comprises sequential speech data frames, wherein a first speech data frame is sequentially prior to a second speech data frame that corresponds to the portion of the speech data, and wherein the processor is further configured to:

generate a first match value indicative of a first quality of a first match between the portion of the substitute speech data and the second speech data frame; and

in response to a determination that the first match value is less than a threshold value, generate a second match value indicative of a second quality of a second match among the portion of the substitute speech data, the first speech data frame and the second speech data frame.

15. The device of claim 14, wherein the processor is further configured to compare the second match value to a second threshold, and wherein generation of the output speech data corresponds to a concatenation of the portion of the substitute speech data to the first speech data frame.

16. A device comprising:

a receiver configured to receive an indicator representative of a locator value associated with substitute speech data;

a memory configured to store the substitute speech data corresponding to a plurality of substitute speech signals; and

a processor electrically coupled to the memory and configured to;

select a portion of the substitute speech data based on the locator value, the portion associated with a particular substitute speech signal; and

generate output speech data based on the portion of the substitute speech data.

17. The device of claim 16, wherein the substitute speech data includes a plurality of substitute speech data portions substantially devoid of distortion.

18. The device of claim 16, wherein the receiver is further configured to receive a representation of speech data corresponding to a speech signal, and wherein the output speech data comprises an overlap of components of the portion of the substitute speech data with a portion of reconstructed speech data, the reconstructed speech data based on the representation of the speech data.

19. The device of claim 16, wherein the output speech data comprises a concatenation of the portion of the substitute speech data and a first speech frame corresponding to speech data.

20. The device of claim 19, wherein the concatenation comprises a substitution of a second speech frame of the speech data with the portion of the substitute speech data and a joining of the portion of the substitute speech data with the first speech frame, and wherein the second speech frame comprises distorted speech data.

21. A device comprising:

a memory configured to store substitute speech data corresponding to a plurality of substitute speech signals and to store a plurality of locator values, each locator value associated with a corresponding portion of the substitute speech data;

a microphone configured to receive a speech signal and to convert the speech signal to speech data; and

a processor electrically coupled to the memory and configured to;

receive the speech data from the microphone;

identify, based on at least a portion of the speech data, a locator value corresponding to a particular portion of the substitute speech data that matches the portion of the speech data; and

generate output data that includes an indication of the locator value.

22. The device of claim 21, wherein the locator value corresponds to an index value indicative of a memory location of the memory at which the particular portion of the substitute speech data is stored.

23. The device of claim 21, wherein the locator value corresponds to a pointer indicative of a memory location of the memory at which the particular portion of the substitute speech data is stored.

24. The device of claim 21, further comprising a transmitter configured to transmit the output data including the indication of the locator value.

25. The device of claim 21, wherein the processor is further configured to select the particular portion of the substitute speech data that matches the portion of the speech data based on a comparison of first audio parameters of the substitute speech data and second audio parameters of the portion of the speech data.

26. The device of claim 21, wherein the processor is further configured to:

determine a first match value indicative of a quality of a match between the particular portion of the substitute speech data and the portion of the speech data; and

compare the first match value to a threshold value.

27. The device of claim 26, wherein the processor is further configured to select the particular portion of the substitute speech data in response to a determination that the first match value equals or exceeds the threshold value.

28. A device comprising:

a memory configured to store substitute speech data corresponding to a plurality of substitute speech signals;

a processor configured to generate a graphical user interface (GUI) that indicates a playback duration of the substitute speech data; and

an interface configured to display the GUI.

29. The device of claim 28, wherein the playback duration is associated with a quantity of frames of substitute speech data, wherein the memory is further configured to store an association between the substitute speech data and a plurality of user identifiers, each user identifier corresponding to a different person, and wherein the substitute speech data comprises first substitute speech data corresponding to a first user identifier, second substitute speech data corresponding to a second user identifier, and third substitute speech data corresponding to a third user identifier.

30. The device of claim 29, wherein the processor is further configured to transmit a request for additional substitute speech data associated with a particular user identifier of the plurality of user identifiers in response to a determination that a first playback duration of first substitute speech data satisfies an update threshold.

31. The device of claim 29, wherein the GUI is configured to receive a first command indicating deletion of the first substitute speech data.

32. The device of claim 29, wherein the GUI indicates a first age of first substitute speech data, a second age of second substitute speech data, and a third age of third substitute speech data, and wherein the substitute speech data corresponds to substantially undistorted speech signals.

33. The device of claim 31, wherein an age of substitute speech data corresponds to a date of acquisition of the substitute speech data.

34. The device of claim 29, wherein the GUI is configured to receive a second command indicating addition of fourth substitute speech data associated with a fourth user identifier of the plurality of user identifiers.