Apparatus and method for editing speech synthesis, and computer readable medium

Patent metadata

Field	Value
Publication number	US-9020821-B2
Application number	US-201113235656-A
Country	US
Kind code	B2
Filing date	Sep 19, 2011
Priority date	Mar 17, 2011
Publication date	Apr 28, 2015
Grant date	Apr 28, 2015

Abstract

An acquisition unit analyzes a text, and acquires phonemic and prosodic information. An editing unit edits a part of the phonemic and prosodic information. A speech synthesis unit converts the phonemic and prosodic information before editing the part to a first speech waveform, and converts the phonemic and prosodic information after editing the part to a second speech waveform. A period calculation unit calculates a contrast period corresponding to the part in the first speech waveform and the second speech waveform. A speech generation unit generates an output waveform by connecting a first partial waveform and a second partial waveform. The first partial waveform contains the contrast period of the first speech waveform. The second partial waveform contains the contrast period of the second speech waveform.

First claim

What is claimed is:

1. An apparatus for editing speech synthesis, comprising:
an acquisition unit, executed by a computer using a program stored in a memory device, configured to analyze a text, and to acquire a phonemic and prosodic information to synthesize a speech corresponding to the text;
a display that displays the phonemic and prosodic information;
an editing unit, executed by the computer, configured to edit at least a part of the phonemic and prosodic information displayed on the display;
a speech synthesis unit, executed by the computer, configured to convert the phonemic and prosodic information in which the part is not edited to a first speech waveform, and to convert the phonemic and prosodic information in which the part is edited to a second speech waveform;
a period calculation unit, executed by the computer, configured to specify a partial sequence corresponding to the part not edited in the phonemic and prosodic information, and the part edited in the phonemic and prosodic information respectively, and to calculate a contrast period corresponding to the partial sequence in the first speech waveform and the second speech waveform respectively;
a speech generation unit, executed by the computer, configured to generate an output waveform by connecting a first partial waveform and a second partial waveform, the first partial waveform being the contrast period of the first speech waveform, the second partial waveform being the contrast period of the second speech waveform; and
a speaker that reproduces the output waveform.

2. The apparatus according to claim 1 , wherein
the speech generation unit inserts a silent period having a predetermined length between the first partial waveform and the second partial waveform in the output waveform.

3. The apparatus according to claim 1 , wherein
the acquisition unit comprises
a reading/prosodic sign generation unit configured to generate a reading sign and a prosodic sign by analyzing the text, and
a synthesized speech control information generation unit configured to generate a synthesized speech control information by analyzing the reading sign and the prosodic sign, and

the editing unit edits at least one of the reading sign, the prosodic sign and the synthesized speech control information, or a combination thereof.

4. The apparatus according to claim 3 , wherein
the period calculation unit calculates the contrast period by using a duration included in the synthesized speech control information.

5. The apparatus according to claim 4 , wherein
the period calculation unit comprises
a partial sequence editing unit configured to edit the partial sequence, and

calculates the contrast period corresponding to the partial sequence edited by the partial sequence editing unit.

6. The apparatus according to claim 1 , further comprising: wherein
the display displays an information representing which of the first partial waveform and the second partial waveform is being outputted by the speaker.

7. A method for editing speech synthesis, comprising:
analyzing, by a computer using a program stored in a memory device, a text;
acquiring, by the computer, a phonemic and prosodic information to synthesize a speech corresponding to the text;
displaying, by the computer, the phonemic and prosodic information via a display;
editing, by the computer, at least a part of the phonemic and prosodic information displayed on the display;
converting, by the computer, the phonemic and prosodic information in which the part is not edited to a first speech waveform;
converting, by the computer, the phonemic and prosodic information in which the part is edited to a second speech waveform;
specifying, by the computer, a partial sequence corresponding to the part not edited in the phonemic and prosodic information, and the part edited in the phonemic and prosodic information respectively;
calculating, by the computer, a contrast period corresponding to the partial sequence in the first speech waveform and the second speech waveform respectively;
generating, by the computer, an output waveform by connecting a first partial waveform and a second partial waveform, the first partial waveform being the contrast period of the first speech waveform, the second partial waveform being the contrast period of the second speech waveform; and
reproducing, by the computer, the output waveform via a speaker.

8. A non-transitory computer readable medium for causing a computer to perform a method for editing speech synthesis, the method comprising:
analyzing, by the computer using a program stored in a memory device, a text;
acquiring, by the computer, a phonemic and prosodic information to synthesize a speech corresponding to the text;
displaying, by the computer, the phonemic and prosodic information via a display;
editing, by the computer, at least a part of the phonemic and prosodic information displayed on the display;
converting, by the computer, the phonemic and prosodic information in which the part is not edited to a first speech waveform;
converting, by the computer, the phonemic and prosodic information in which the part is edited to a second speech waveform;
specifying, by the computer, a partial sequence corresponding to the part not edited in the phonemic and prosodic information, and the part edited in the phonemic and prosodic information respectively;
calculating, by the computer, a contrast period corresponding to the partial sequence in the first speech waveform and the second speech waveform respectively;
generating, by the computer, an output waveform by connecting a first partial waveform and a second partial waveform, the first partial waveform being the contrast period of the first speech waveform, the second partial waveform being the contrast period of the second speech waveform; and
reproducing, by the computer, the output waveform via a speaker.

Assignees

• Nishiyama Osamu
• Toshiba Kk

Inventors

• Nishiyama Osamu

Classifications

• G10L13/033 Primary

  Voice editing, e.g. manipulating the voice of the synthesiser · CPC title

• G10L13/08

  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 · CPC title