Speech synthesis from acoustic units with default values of concatenation cost

• US 8,788,268 B2
• Filed: 11/19/2012
• Issued: 07/22/2014
• Est. Priority Date: 04/30/1999
• Status: Expired due to Fees

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: AT&T Intellectual Property II LP (AT&T, Inc.)
• Inventors: Beutnagel, Mark Charles, Mohri, Mehryar, Riley, Michael Dennis
• Primary Examiner(s): Lerner, Martin
• Application Number:
  US13/680,622
• Publication Number:
  US 20130080176A1
• Time in Patent Office: 610 Days
• Field of Search: 704/258, 704/263, 704/266, 704/260, 704/262
• US Class Current: 704/258
• CPC Class Codes: G10L 13/00 Speech synthesis; Text to speech systems
• CPC Class Codes: G10L 13/027 Concept to speech synthesisers; Generation of natural phrases from machine-based concepts generation of parameters for speech synthesis out of text G10L13/08
• CPC Class Codes: G10L 13/07 Concatenation rules
• CPC Class Codes: G10L 13/08 Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination

Abstract

A speech synthesis system can select recorded speech fragments, or acoustic units, from a very large database of acoustic units to produce artificial speech. When a pair of acoustic units in the database does not have an associated concatenation cost, the system assigns a default concatenation cost. The system then synthesizes speech, identifies the acoustic unit sequential pairs generated and their respective concatenation costs, and stores those concatenation costs likely to occur.

Claims

1. A method comprising:

when, while synthesizing speech via a processor, an acoustic unit sequential pair does not have an associated concatenation cost in a concatenation cost database;

assigning a default value as the associated concatenation cost; and

updating the concatenation cost database by synthesizing, via the processor, a body of speech, identifying the acoustic unit sequential pair in the body of speech, and recording a respective concatenation cost in the concatenation cost database.

2. The method of claim 1, further comprising synthesizing the speech using the respective concatenation cost.

3. The method of claim 1, wherein recording the respective concatenation cost comprises deriving an actual concatenation cost.

4. The method of claim 1, wherein the concatenation cost database contains a portion of all possible concatenation costs associated with a list of acoustic units.

5. The method of claim 1, wherein the respective concatenation cost comprises a weighted sum of subcosts across phones.

6. The method of claim 1, wherein the concatenation cost database is generated using statistical techniques which predict which of the acoustic unit sequential pairs are most likely to occur in common speech.

7. The method of claim 1, wherein the concatenation cost database stores acoustic units in linear predictive coding parameters.

8. A system comprising:

a processor; and

a computer-readable storage medium having instructions stored which, when executed by the processor, cause the processor to perform operations comprising;

when, while synthesizing speech, an acoustic unit sequential pair does not have an associated concatenation cost in a concatenation cost database;

assigning a default value as the associated concatenation cost; and

updating the concatenation cost database by synthesizing a body of speech, identifying the acoustic unit sequential pair in the body of speech, and recording a respective concatenation cost in the concatenation cost database.

9. The system of claim 8, the computer-readable storage medium having additional instructions stored which result in the operations comprising synthesizing the speech using the respective concatenation cost.

10. The system of claim 8, wherein recording the respective concatenation cost comprises deriving an actual concatenation cost.

11. The system of claim 8, wherein the concatenation cost database contains a portion of all possible concatenation costs associated with a list of acoustic units.

12. The system of claim 8, wherein the respective concatenation cost comprises a weighted sum of subcosts across phones.

13. The system of claim 8, wherein the concatenation cost database is generated using statistical techniques which predict which of the acoustic unit sequential pairs are most likely to occur in common speech.

14. The system of claim 8, wherein the concatenation cost database stores acoustic units in linear predictive coding parameters.

15. A computer-readable storage device having instructions stored which, when executed by a computing device, cause the computing device to perform operations comprising:

when, while synthesizing speech, an acoustic unit sequential pair does not have an associated concatenation cost in a concatenation cost database;

assigning a default value as the associated concatenation cost; and

updating the concatenation cost database by synthesizing a body of speech, identifying the acoustic unit sequential pair in the body of speech, and recording a respective concatenation cost in the concatenation cost database.

16. The computer-readable storage device of claim 15, the computer-readable storage device having additional instructions stored which result in operations comprising synthesizing the speech using the respective concatenation cost.

17. The computer-readable storage device of claim 15, wherein recording the respective concatenation cost comprises deriving an actual concatenation cost.

18. The computer-readable storage device of claim 15, wherein the concatenation cost database contains a portion of all possible concatenation costs associated with a list of acoustic units.

19. The computer-readable storage device of claim 15, wherein the respective concatenation cost comprises a weighted sum of subcosts across phones.

20. The computer-readable storage device of claim 15, wherein the concatenation cost database is generated using statistical techniques which predict which of the acoustic unit sequential pairs are most likely to occur in common speech.