Synthesis of speech from text

• US 5,463,713 A
• Filed: 04/21/1994
• Issued: 10/31/1995
• Est. Priority Date: 05/07/1991
• Status: Expired due to Fees

• Current Assignee: Kabushiki Kaisha Meidensha (Meidensha Corporation)
• Original Assignee: Kabushiki Kaisha Meidensha (Meidensha Corporation)
• Inventors: Hasegawa, Kazsuya
• Primary Examiner(s): MacDonald, Allen R.
• Assistant Examiner(s): Doerrler, Michelle
• Application Number:
  US08/232,438
• Time in Patent Office: 558 Days
• Field of Search: 395/2.67-2.78, 381/51-53
• US Class Current: 704/260
• CPC Class Codes: G10L 13/04 Details of speech synthesis systems, e.g. synthesiser structure or memory management
• CPC Class Codes: G10L 13/10 Prosody rules derived from text; Stress or intonation

Abstract

An apparatus for synthesizing speech from text includes a language processing section which determines an accent environment of each mora of the text. In a basic accent pattern table, a basic accent pattern is classified according to the accent environment of the mora. The basic accent pattern includes a pitch data which is edited from real voice data according to the accent environment. A basic accent pattern processing section selects the basic accent pattern of each more from the basic accent pattern table according to the accent environment and processes the basic accent pattern in pitch according to the accent environment. A correcting section receives the corrected pitch data in the basic accent patter processing section and corrects the corrected pitch data according to the number of mora in each phrase and the position of the mora in phrase so as to correct the data into the corrected accent component. A phrase pattern processing section determines a phrase component according to the number of mora in each phrase which is of the accent environment. A speech synthesizing section synthesizes speech according to an accent control pattern of the text which is obtained by adding the accent pattern and the phrase pattern.

Claims

1. An apparatus for synthesizing speech from text, comprising:

a language processing section determining an accent environment of each mora of each phrase of the text, said accent environment including a height of an accent of each mora;

a basic accent pattern table in which a basic accent pattern has been classified according to an accent environment of the mora, the basic accent pattern including pitch data which has been edited from real voice data according to the accent environment;

a basic accent pattern processing section selecting the basic accent pattern of each mora from said basic accent pattern table according to the accent environment and processing the basic accent pattern in a pitch according to the accent environment;

a correcting section receiving the basic access pattern in the pitch in said basic accent pattern processing section and correcting the pitch according to the number of moras in each phrase and the position of the moras in the phrase so as to correct the data in the corrected accent component;

a phrase pattern processing section determining a phrase component according to the number of moras in each phrase of the accent environment; and

a speech synthesizing section synthesizing speech according to an accent control pattern of the text which is obtained by adding the basic accent pattern and the basic phrase pattern.

2. An apparatus for synthesizing speech from text as claimed in claim 1, wherein said basic accent pattern table is classified in accordance with an accent environment and a position of an accent boundary.

3. An apparatus for synthesizing speech from text as claimed in claim 2, wherein the position of the accent is determined in accordance with whether the accent boundary is positioned at a forward portion of the mora or at a back portion of the mora.

4. An apparatus for synthesizing speech from text as claimed in claim 2, wherein the type of the mora is determined in accordance with whether the mora is a vowel, vocal consonant and vowel, voiceless consonant and vowel, long vowel, vocal consonant and long vowel, or voiceless consonant and long vowel.

5. An apparatus for synthesizing speech from text as claimed in claim 1, wherein said basic accent pattern table is classified in accordance with the accent environment of each mora and the type of each mora.

6. An apparatus for synthesizing speech from text as claimed in claim 1, wherein the maintenance of the apparatus is carried out by correcting the pitch data in said accent pattern table.

7. An apparatus as claimed in claim 1, further comprising a text input section at which the text is transmitted into signals and sent to said language processing section.

8. An apparatus for synthesizing speech from text as claimed in claim 1, wherein the accent environment includes the height of an accent of each mora and the accent height of forward and back moras of each mora.

9. An accent pattern calculating section in an accent control section of a speech synthesizer, the speech synthesizer having a text input section for inputting text data, the text input section being connected to a language processing section for analyzing the content of the text with morpheme analysis, an accent pattern component obtained from said accent pattern calculating section being combined with a phrase component formed in a phrase pattern calculating section, said accent pattern calculating section comprising:

a basic accent pattern table having a basic accent pattern classified according to an accent environment of the mora which includes a height of an accent of each mora, the basic accent pattern including pitch data which has been edited from a real voice data according to the accent environment;

a basic accent pattern processing section selecting the basic accent pattern of each mora from said basic accent pattern table according to the accent environment and processing the basic accent pattern in a pitch according to the accent environment; and

a correcting section receiving the basic accent pattern with the pitch from said basic accent pattern processing section and correcting the pitch according to the number of moras in each phrase and the position of the moras in the phrase, so as to correct the data in a corrected accent component.

10. A method for synthesizing speech from text, comprising the steps of:

a) inputting text data into a text input section;

b) analyzing the contents of the text in a language processing section with morpheme analysis;

c) obtaining an accent pattern component from an accent pattern calculating section;

d) obtaining a phrase component from a phrase pattern calculating section; and

e) combining said accent pattern component with said phrase component, wherein said step c) further comprises the steps of classifying a basic accent pattern in a basic accent pattern table according to an accent environment of each mora of the text data, the basic accent pattern including pitch data which has been edited from real voice data according to the accent environment;

selecting the basic accent pattern of each mora from said basic accent pattern table in a basic accent pattern processing section according to the accent environment and processing the basic accent pattern in a pitch according to the accent environment; and

receiving the basic accent pattern with the pitch from said basic accent pattern processing section in a correcting section and correcting the pitch according to the number of moras in each phrase and the position of the moras in the phrase so as to correct the data in a corrected accent component.