Autonomous system and method for creating readable scripts for concatenative text-to-speech synthesis (TTS) corpora

• US 20070168193A1
• Filed: 01/17/2006
• Published: 07/19/2007
• Est. Priority Date: 01/17/2006
• Status: Active Grant

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: International Business Machines Corporation
• Inventors: Pitrelli, John, Aaron, Andrew, Ferrucci, David
• Granted Patent: US 8,155,963 B2
• Time in Patent Office: Days
• US Class Current: 704/260
• CPC Class Codes: G10L 13/04 Details of speech synthesis systems, e.g. synthesiser structure or memory management

Abstract

A method (and system) which autonomously generates a cohesive script from a text database for creating a speech corpus for concatenative text-to-speech, and more particularly, which generates cohesive scripts having fluency and natural prosody that can be used to generate compact text-to-speech recordings that cover a plurality of phonetic events.

Claims

1. A method of generating a speech corpus for concatenative text-to-speech, comprising:

autonomously generating a cohesive script based on a text database.

2. The method according to claim 1, wherein said autonomously generating comprises:

selecting at least one of a word and a word sequence from said text database based on an enumerated phoneme sequence; and

generating said coherent script including said selected at least one of said word and said word sequence.

3. The method according to claim 2, wherein said enumerated phoneme sequence comprises:

at least one of a diphone, a triphone, a quadphone, a syllable, and a bisyllable.

4. The method according to claim 1, wherein said autonomously generating said cohesive script, comprises:

extracting at least one predetermined sequence of phonemes from said text database;

associating said predetermined sequence of phonemes with a plurality of words included in said text database that include said predetermined sequence of phonemes;

selecting N words that include said predetermined sequence of phonemes; and

generating said cohesive script based on said N words.

5. The method according to claim 4, wherein said predetermined sequence of phonemes comprises:

at least one of a plurality of diphones, a plurality of triphones, a plurality of quadphones, a plurality of syllables defined in terms of phones, and a plurality of bisyllables defmed in terms of phones.

6. The method according to claim 1, wherein said text database comprises:

at least one of a vocabulary list, an unstructured vocabulary list, an inventory of occurrences of at least one phonemic unit, an inventory of occurrences of at least one phonemic sequence, a dictionary, and a word pronunciation guide.

7. The method according to claim 1, wherein said autonomously generating said cohesive script comprises:

generating said cohesive script based on at least one of a character template, a concept template, a location template, a story line template, and a script template.

8. The method according to claim 4, further comprising:

generating said speech corpus based on said cohesive script.

9. The method according to claim 4, further comprising:

controlling format mechanics of said cohesive script.

10. The method according to claim 9, wherein said format mechanics comprise:

at least one of a script size, a sentence structure, and a target sentence length of said cohesive script.

11. The method according to claim 1, wherein said cohesive script comprises:

a fluently-readable text document.

12. A method of deploying computing infrastructure in which computer-readable code is integrated into a computing system, and combines with said computing system to perform the method according to claim 1.

13. A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform the method according to claim 1.

14. A system for generating a speech corpus for concatenative text-to-speech, comprising:

an extracting unit that extracts at least one enumerated phoneme sequence from a text database;

an associating unit that associates each of said at least one enumerated phoneme sequence with a plurality of words included in said text database that include said each of said at least one enumerated phoneme sequence;

a selecting unit that selects N words that include said each of said at least one enumerated phoneme sequence; and

an autonomous language generating unit which receives the N selected words and generates a cohesive script.

15. The system according to claim 14, wherein said at least one enumerated phoneme sequence comprises:

at least one of a plurality of diphones, a plurality of triphones, a plurality of quadphones, a plurality of syllables defined in terms of phones, and a plurality of bisyllables defined in terms of phones.

16. The system according to claim 14, further comprising:

at least one of a character template unit, a concept template unit, a location template unit, a story line template unit, and a script template unit for providing input to said autonomous language generating unit.

17. The system according to claim 14, further comprising:

a control unit that controls format mechanics of said cohesive script.

18. The system according to claim 17, wherein said format mechanics comprise:

at least one of a script size, a sentence structure, and a target sentence length of said autonomous language generated by said autonomous language generating unit.

19. The system according to claim 14, further comprising:

a recording unit that generates said speech corpus from said cohesive script.

20. The system according to claim 14, wherein said text database comprises:

at least one of a vocabulary list, an unstructured vocabulary list, an inventory of occurrences of at least one phonemic unit, an inventory of occurrences of at least one phonemic sequence, a dictionary, and a word pronunciation guide.