Synthesis-based pre-selection of suitable units for concatenative speech

• US 7,233,901 B2
• Filed: 12/30/2005
• Issued: 06/19/2007
• Est. Priority Date: 07/05/2000
• Status: Expired due to Term

• Current Assignee: Cerence Operating Company (Cerence Inc.)
• Original Assignee: AT&T Corporation (AT&T, Inc.)
• Inventors: Conkie, Alistair D.
• Primary Examiner(s): MCFADDEN, SUSAN IRIS
• Application Number:
  US11/275,432
• Publication Number:
  US 20060100878A1
• Time in Patent Office: 536 Days
• Field of Search: 704/260, 704/266
• US Class Current: 704/260
• CPC Class Codes: G10L 13/07 Concatenation rules

Abstract

A system and computer-readable medium synthesize speech from text using a triphone unit selection database. The instructions on the computer-readable medium control a computing device to perform the steps: receiving input text, selecting a plurality of N phoneme units from the triphone unit selection database as candidate phonemes for synthesized speech based on the input text, applying a cost process to select a set of phonemes from the candidate phonemes and synthesizing speech using the selected set of phonemes.

Claims

1. A system for synthesizing speech from text using a triphone unit selection database, the system comprising:

a module configured to receive input text;

a module configured to select a plurality of N phoneme units from the triphone unit selection database as candidate phonemes for synthesized speech based on the input text;

a module configured to apply a cost process to select a set of phonemes from the candidate phonemes; and

a module configured to synthesize speech using the selected set of phonemes.

2. The system of claim 1, wherein a Viterbi search is applied as the cost process.

3. The system of claim 1, further comprising:

a module configured to parse the received input text into recognizable units.

4. The system of claim 3, wherein the module configured to parse the received input text further:

applies a text normalization process to parse the received text into known words and convert abbreviations into known words; and

applies a syntactic process to perform a grammatical analysis of the known words and identify their associated part of speech.

5. A system for synthesizing speech from text using a triphone unit selection database, the system comprising:

means for receiving input text;

means for selecting a plurality of N phoneme units from the triphone unit selection database as candidate phonemes for synthesized speech based on the input text;

means for applying a cost process to select a set of phonemes from the candidate phonemes; and

means for synthesizing speech using the selected set of phonemes.

6. The system of claim 5, wherein a Viterbi search is applied as the cost process.

7. The system of claim 5, further comprising:

means for parsing the received input text into recognizable units.

8. The system of claim 7, wherein the means for parsing the received input text further:

applies a text normalization process to parse the received text into known words and convert abbreviations into known words; and

applies a syntactic process to perform a grammatical analysis of the known words and identify their associated part of speech.

9. A computer-readable medium storing instructions for controlling a computing device to synthesize speech from text using a triphone unit selection database, the instructions comprising:

receiving input text;

selecting a plurality of N phoneme units from the triphone unit selection database as candidate phonemes for synthesized speech based on the input text;

applying a cost process to select a set of phonemes from the candidate phonemes; and

synthesizing speech using the selected set of phonemes.

10. The computer-readable medium of claim 9, wherein a Viterbi search is applied as the cost process.

11. The computer-readable medium of claim 9, wherein subsequent to the step of receiving the input text the following step is performed:

parsing the received text into recognizable units.

12. The computer-readable medium of claim 11, wherein the parsing further comprises the steps:

applying a text normalization process to parse the received text into known words and convert abbreviations into known words; and

applying a syntactic process to perform a grammatical analysis of the known words and identify their associated part of speech.