Text structure for voice synthesis, voice synthesis method, voice synthesis apparatus, and computer program thereof

• US 20050065795A1
• Filed: 08/10/2004
• Published: 03/24/2005
• Est. Priority Date: 04/02/2002
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Fukada, Toshiaki, Mutsuno, Masahiro
• Granted Patent: US 7,487,093 B2
• Time in Patent Office: Days
• US Class Current: 704/260
• CPC Class Codes: G10L 13/033 Voice editing, e.g. manipulating the voice of the synthesiser
• CPC Class Codes: G10L 13/04 Details of speech synthesis systems, e.g. synthesiser structure or memory management
• CPC Class Codes: G10L 13/08 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

Abstract

In a voice synthesis apparatus, by bounding a desired range of input text to be output by, e.g., a start tag “<morphing type=“emotion” start=“happy” end=“angry”>” and end tag </morphing>, a feature of synthetic voice is continuously changed while gradually changing voice from a happy voice to an angry voice upon outputting synthetic voice.

Claims

1. A voice synthesis method for synthesizing a voice waveform to continuously change a feature of synthetic voice of a range assigned a predetermined identifier included in input text upon outputting synthetic voice corresponding to the text, comprising:

a setting step of setting a desired range of text to be output, in which the feature of synthetic voice is to be continuously changed, using a predetermined identifier including attribute information that represents a change mode of the feature of synthetic voice;

a recognition step of recognizing the predetermined identifier and a type of attribute information contained in the predetermined identifier from the text with the identifier, which is set in said setting step; and

a voice synthesis step of synthesizing a voice waveform, whose feature of synthetic voice continuously changes, in accordance with the attribute information contained in the predetermined identifier, by interpolating synthetic voice corresponding to text within the desired range of the text with the identifier in accordance with a recognition result in said recognition step.

2. A voice synthesis method according to claim 1, wherein the attribute information contained in the predetermined identifier represents a change mode of the feature of synthetic voice at a start position of the range set by the identifier, and a change mode of the feature of synthetic voice at an end position.

3. A voice synthesis method according to claim 1, wherein the change mode of the feature of synthetic voice represented by the attribute information is at least one of a change in volume, a change in speaker, a change in output device, a change in number of speakers, a change in emotion, a change in uttering speed, and a change in fundamental frequency.

4. A voice synthesis method according to claim 1, wherein said voice synthesis step includes a step of:

generating synthetic voice corresponding to the text within the desired range on the basis of attribute information associated with start and end positions of the range set by identifiers contained in the predetermined identifier, and a mode of the feature of synthetic voice before the start position.

5. A voice synthesis method according to claim 4, wherein said voice synthesis step includes a step of:

generating synthetic voice corresponding to the text within the desired range on the basis of a ratio between values that represent uttering speeds set as the attribute information associated with the start and end positions, and a value that represents an uttering speed before the start position.

6. A voice synthesis method according to claim 4, wherein said voice synthesis step includes a step of:

generating synthetic voice corresponding to the text within the desired range on the basis of a ratio between values that represent volumes set as the attribute information associated with the start and end positions, and a value that represents a volume before the start position.

9. A computer program constituting operating commands for implementing the voice synthesis method described in claim 1.

7. A voice synthesis apparatus for synthesizing a voice waveform to continuously change a feature of synthetic voice of a range assigned a predetermined identifier included in input text upon outputting synthetic voice corresponding to the text, comprising:

recognition means for recognizing, from text with an identifier, in which a predetermined identifier that represents a desired range, in which the feature of synthetic voice is to be continuously changed, and which contains attribute information representing a change mode of the feature of synthetic voice, the predetermined identifier and a type of attribute information contained in the predetermined identifier from the text with the identifier; and

voice synthesis means for synthesizing a voice waveform, whose feature of synthetic voice continuously changes, in accordance with the attribute information contained in the predetermined identifier, by interpolating synthetic voice corresponding to text within the desired range of the text with the identifier in accordance with a recognition result of said recognition means.

8. A text structure for voice synthesis, in which a predetermined identifier is assigned to change a feature of synthetic voice of a desired range of text to be output by voice synthesis,

wherein the predetermined identifier contains attribute information that represents a change mode upon continuously changing the feature of synthetic voice.

10. A computer program that instructs a computer to serve as a voice synthesis apparatus for synthesizing a voice waveform to change a feature of synthetic voice of a range assigned a predetermined identifier included in input text upon outputting synthetic voice corresponding to the text, comprising:

a program code for a recognition function of recognizing, from text with an identifier, in which a predetermined identifier that represents a desired range, in which the feature of synthetic voice is to be continuously changed, and which contains attribute information representing a change mode of the feature of synthetic voice, the predetermined identifier and a type of attribute information contained in the predetermined identifier from the text with the identifier; and

a program code of a voice synthesis function of synthesizing a voice waveform, whose feature of synthetic voice continuously changes, in accordance with the attribute information contained in the predetermined identifier, by interpolating synthetic voice corresponding to text within the desired range of the text with the identifier in accordance with a recognition result of the recognition function.