Method and apparatus of selecting segments for speech synthesis by way of speech segment recognition

• US 7,054,814 B2
• Filed: 03/29/2001
• Issued: 05/30/2006
• Est. Priority Date: 03/31/2000
• Status: Expired due to Fees

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Komori, Yasuhiro, Fukada, Toshiaki, Okutani, Yasuo
• Primary Examiner(s): ARMSTRONG, ANGELA A
• Application Number:
  US09/819,613
• Publication Number:
  US 20020051955A1
• Time in Patent Office: 1,888 Days
• Field of Search: 704/256, 704/10, 704/231, 704/260, 704/268, 704/258, 704/9, 704/254, 704/256.2, 704/256.4, 704/256.5, 358/1.9, 712/11
• US Class Current: 704/256.4
• CPC Class Codes: G09B 19/04 Speaking with audible presentation of the material to be studied G09B5/04
• CPC Class Codes: G09B 7/02 of the type wherein the student is expected to construct an answer to the question which is presented or wherein the machine gives an answer to the question presented by a student

Abstract

A speech segment search unit searches a speech database for speech segments that satisfy a phonetic environment, and a HMM learning unit computes the HMMs of phonemes on the basis of the search result. A segment recognition unit performs segment recognition of speech segments on the basis of the computed HMMs of the phonemes, and when the phoneme of the segment recognition result is equal to a phoneme of the source speech segment, that speech segment is registered in a segment dictionary.

Claims

1. An apparatus for forming a speech segment dictionary for use in speech synthesis, comprising:

speech segment search means for searching a speech database for speech segments that satisfy a phonetic environment;

HMM learning means for computing HIMMs of phonemes on the basis of a search result of said speech segment search means;

segment recognition means for performing segment recognition of the speech segments on the basis of the HMMs of the phonemes; and

registration segment means for determining and registering a speech segment in a segment dictionary when the number of speech segments recognized by said segment recognition means is not less than a predetermined value.

2. The apparatus according to claim 1, wherein said registration segment means registers a speech segment in the segment dictionary if at least a vowel part of the speech segment is correctly recognized, even when the number of speech segments recognized by said segment recognition means is not more than a predetermined value.

3. A method for forming a speech segment dictionary for use in speech synthesis, comprising:

a speech segment search step of searching a speech database for speech segments that satisfy a phonetic environment;

an HMM learning step of computing HMMs of phonemes on the basis of a search result in said speech segment search step;

a segment recognition step of performing segment recognition of the speech segments on the basis of the HMMs of the phonemes; and

a registration segment step of determining and registering a speech segment in a segment dictionary when the number of speech segments recognized in said segment recognition step is not less than a predetermined value.

4. The method according to claim 3, wherein said registration segment step includes a step of registering a speech segment in the segment dictionary if at least a vowel part of the speech segment is correctly recognized, even when the number of speech segments recognized in said segment recognition step is not more than a predetermined value.

5. A speech synthesis apparatus comprising a unit for synthesizing speech and a unit for forming a speech segment dictionary for use in speech synthesis, the unit for forming a speech segment dictionary for use in speech synthesis comprising:

speech segment search means for searching a speech database for speech segments that satisfy a phonetic environment;

HMM learning means for computing HMMs of phonemes on the basis of a search result of said speech segment search means;

segment recognition means for performing segment recognition of the speech segments on the basis of the HMMs of the phonemes; and

registration segment means for determining and registering a speech segment in a segment dictionary when the number of speech segments recognized by said segment recognition means is not less than a predetermined value.

6. The apparatus according to claim 5, wherein said registration segment means registers a speech segment in the segment dictionary if at least a vowel part of the speech segment is correctly recognized, even when the number of speech segments recognized by said segment recognition means is not more than a predetermined value.

7. A speech synthesis apparatus comprising (i) a unit for forming a speech segment dictionary for use in speech synthesis, and (ii) a unit for synthesizing speech, wherein said unit for forming a speech segment dictionary for use in speech synthesis comprises:

speech segment search means for searching a speech database for speech segments that satisfy a phonetic environment;

HMM learning means for computing HMMs of phonemes on the basis of a search result of said speech segment search means;

segment recognition means for performing segment recognition of the speech segments on the basis of the HMMs of the phonemes; and

registration segment means for determining and registering a speech segment in a segment dictionary when the number of speech segments recognized by said segment recognition means is not less than a predetermined value,

and wherein said unit for synthesizing speech comprises;

a unit for inputting text data; and

a unit for synthesizing speech corresponding to the inputted text data, using the unit for forming a speech segment dictionary for use in speech synthesis.

8. The apparatus according to claim 7, wherein said registration segment means registers a speech segment in the segment dictionary if at least a vowel part of the speech segment is correctly recognized, even when the number of speech segments recognized by said segment recognition means is not more than a predetermined value.

9. A speech synthesis method comprising (i) a method for forming a speech segment dictionary for use in speech synthesis, and (ii) a method for synthesizing speech, wherein said method for forming a speech segment dictionary for use in speech synthesis comprises:

a speech segment search step of searching a speech database for speech segments that satisfy a phonetic environment;

an HMM learning step of computing HMMs of phonemes on the basis of a search result in said speech segment search step;

a segment recognition step of performing segment recognition of the speech segments on the basis of the HMMs of the phonemes; and

a registration segment step of determining and registering a speech segment in a segment dictionary when the number of speech segments recognized in said segment recognition step is not less than a predetermined value,

and wherein said method for synthesizing speech comprises;

an input step of inputting text data; and

a speech synthesis step of synthesizing speech corresponding to the inputted text data, using the method for forming a speech segment dictionary for use in speech synthesis.

10. The speech signal processing method according to claim 9, wherein said registration segment step includes a step of registering a speech segment in the segment dictionary if at least a vowel part of the speech segment is correctly recognized, even when the number of speech segments recognized in said segment recognition step is not more than a predetermined value.