JPH08328575A - Voice synthesizer - Google Patents

Abstract

PURPOSE: To provide a voice synthesizer in which the voices generated as synthesized voices are converted to the voices having the voice quality of, for example, the character of an animation or the blood relative of the user who is using the synthesizer. CONSTITUTION: The synthesizer is provided with a voice data keeping section 8 which keeps the voice data to be reproduced as synthesized voices, a text data keeping section 9 which keeps the text data, a voice synthesizing section 11 which reproduces the data, that are kept in these sections, as synthesized voices, a voice quality feature extracting section 3 which generates the voice quality data in that the features of the voice quality of the inputted voice signals are extracted and a voice generation mode setting section 7 which sets the generation mode during the generation of the synthesized voices. The data kept in the sections 8 or 9 are voice synthesized by the section 11 using the voice quality data extracted by the section 3 and are reproduced in accordance with the generation mode specified by the section 7.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice synthesizing device, and is particularly suitable for use in a variety of devices that need to produce a synthetic voice for children, such as a personal computer for children. Speech synthesizer.

[0002]

2. Description of the Related Art With the recent development and popularization of personal computers, personal computers mainly for learning purposes have been commercialized for children before and after reaching school age. By the way, in such a personal computer, the operation itself of the personal computer, or the learning procedure, question, answer, how to obtain the answer, and further, reading out the story, singing a song, etc. are performed by uttering a synthetic voice. In many cases

[0003]

As described above, the conventional personal computer for children as described above is designed to give various instructions to the child by means of the synthesized voice, but the synthesized voice so far is inflectioned. , Accents, etc. are unnatural, the rate of speech is constant, and the voice quality is one type or at most several types. Moreover, those voice types are preset in the personal computer and are familiar to children. Often has no voice. As a result, children using such personal computers are less likely to learn.

The present invention has been made in view of the above circumstances, and can convert a voice uttered as a synthetic voice into various voice qualities, for example, an animated character, a voice timbre of a relative of a person who uses the apparatus, or the like. An object is to provide a speech synthesizer.

[0005]

SUMMARY OF THE INVENTION A speech synthesizer according to the present invention comprises a data storage section for storing voice data and text data to be reproduced as synthetic speech, and a data storage section for storing data stored in the data storage section. And a voice quality feature extraction unit that generates voice quality data by extracting voice quality features of the input voice signal.

Further, the voice synthesizing apparatus according to the present invention includes a data storage unit for storing voice data and text data to be reproduced as a synthetic voice, and a voice for reproducing the data stored in the data storage unit as a synthetic voice. A synthesizing unit and a voicing style setting unit for designating a voicing style when the voice synthesizing unit utters a synthetic voice.

Further, the voice synthesizing apparatus according to the present invention includes a data storage unit for storing voice data and text data to be reproduced as synthetic voice, and a voice for reproducing data stored in the data storage unit as synthetic voice. A synthesizing unit, a voice quality feature extracting unit that generates voice quality data obtained by extracting a voice quality feature of the input voice signal, and a voicing style setting unit that specifies a voicing style when the voice synthesizing unit utters a synthetic voice. .

[0008]

In the voice synthesizing apparatus according to the present invention, the data stored in the data storage unit is converted into a synthetic voice by using the voice quality data extracted by the voice quality feature extraction unit.

Further, in the voice synthesizing apparatus according to the present invention, the data stored in the data storage unit is reproduced according to the utterance style specified by the utterance style setting unit.

Further, in the voice synthesizing apparatus according to the present invention, the data stored in the data storage unit is synthesized into voice by using the voice quality data extracted by the voice quality feature extraction unit,
Reproduction is performed according to the utterance style specified by the utterance style setting unit.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below in detail with reference to the drawings showing the embodiments thereof. FIG. 1 is a block diagram showing the basic configuration of a speech synthesizer according to the present invention.

In FIG. 1, reference numeral 3 indicates a voice quality feature extraction unit, which directly inputs an analog voice signal from a microphone 1 for inputting a raw voice, a video tape recorder, a television, a radio, or the like. A line input terminal 2 for inputting a voice signal already recorded in the recorder is provided. The voice quality feature extraction unit 3 extracts voice quality features of the voice signal input from the microphone 1 or the line input terminal 2 described above, for example, by analyzing its frequency component to generate voice quality data.
Hereinafter, a voice signal from which voice quality features are extracted by the voice quality feature extraction unit 3 to generate voice quality data is referred to as a sample voice.

Reference numeral 4 indicates a voice quality data storage unit, which stores voice quality feature data of the sample voice extracted by the voice quality feature extraction unit 3, that is, voice quality data. The plurality of voice quality data stored in the voice quality data storage unit 4 is stored in the voice quality selection unit 5 using, for example, a keyboard.
It is possible to select by the operation of.

Reference numeral 6 indicates a voice quality data processing unit, which processes the voice quality data of the sample voice selected by the voice quality selection unit 5 and output from the voice quality data storage unit 4. The processing of the voice quality data includes, for example, synthesis of a plurality of voice quality data, feminization of a male voice, conversely, maleization of a female voice,
It is possible to adultize the voice of a child and to make the voice of an adult a child.
The processing of the voice quality data by the voice quality data processing unit 6 can also be selected by the voice quality selection unit 5. Such processing of voice quality data is easy to mutually convert, for example, if a general difference tendency between the voice quality data of male voice and the voice quality data of female voice is known. .

Reference numeral 7 indicates a utterance style setting section, which can give various prosody information depending on the purpose, situation, sentence content, etc. when uttering a synthetic voice. For example, when a person asks a question and when a story is read aloud, the way of speaking is different, but conventional synthetic speech cannot make such a distinction and speak synthetic speech. However, in the speech synthesizing apparatus of the present invention, the prosody information can be changed by the utterance style setting unit 7 according to the purpose and situation when uttering the synthesized voice, and the intonation, accent, or utterance speed of the synthesized voice can be changed. It will be possible. The voice data storage unit 8 can also use, for example, a keyboard or the like, and of course, one keyboard can be used together with the voice quality selection unit 5.

Reference numeral 8 indicates a voice data storage unit, which stores various information already existing as voice data. Reference numeral 9 indicates a text data storage unit, in which various text data are stored. The voice data storage unit 8 and the text data storage unit 9 may use one storage device.

Reference numeral 10 indicates an output data selection unit, which selects either the voice data or the text data stored in the voice data storage unit 8 and the text data storage unit 9 to select the voice synthesis unit 11. Output to. As for the output data selecting section 10, it is of course possible to use the voice quality selecting section 5 and the utterance style setting section 7 together with one keyboard.

Reference numeral 11 indicates a voice synthesizing unit, and when the data stored in the voice data storing unit 8 or the text data storing unit 9 is selected by the output data selecting unit 10, the voice quality data processing unit 6 The voice quality data of the sample voice given by the above is converted into a synthetic voice and reproduced from the speaker 12. It is of course possible to use not only the speaker 12 but also an earphone, a headphone, or the like, and it is also possible to output from the line output terminal to various recorders for recording.

The operation of the speech synthesizer of the present invention as described above will be described below with reference to the flowchart of FIG.

By the way, the voice quality (timbre) of a human voice is one of the three elements (volume, pitch, and timbre) of a voice (sound), and mainly the frequency of each partial sound constituting the sound. Stipulated by Therefore, it becomes possible to extract an analog voice signal as a voice quality characteristic coefficient by analyzing its frequency component. The voice quality feature extraction unit 3 inputs the sample voice from the microphone 1 or the line input terminal 2 by such a method (step S1), frequency-analyzes the input analog voice signal of the sample voice, and the voice quality feature coefficient. That is, voice quality data is extracted (step S2).

The voice quality data of the sample voice thus extracted by the voice quality feature extraction unit 3 is stored in the voice quality data storage unit 4 (step S3). Therefore, by inputting voices of various people as sample voices from the microphone 1 by the voice quality feature extraction unit 3 described above,
It is possible to extract voice quality data of these sample voices and store them in advance. For example, when the device is mainly used by a child, it is desirable that the voice of the child's parents, grandparents, or the like, or the voice of a teacher in a nursery school, kindergarten, or the like is used as the sample voice. Further, from the line input terminal 2, voices are directly recorded from a video tape recorder, a television, a radio, or the like, or by inputting voices already recorded in a recorder, various voice quality data are stored in advance as sample voices. It is possible to set. In this case, for example, it is possible to store voice data of animation characters, voice data of talent voices, and the like.

On the other hand, regarding the text data, various text data can be stored in the text data storage unit 9 in advance. For example, if a flexible disk driver is used as the text data storage unit 9, a flexible disk as a storage medium in which text data is already stored may be mounted, or if a function as a word processor is provided, a keyboard is provided. It is also possible to directly input the text data by an operation such as.

Regarding voice data, it is possible to store various voice data in the voice data storage unit 8 in advance. For example, as in the case of the text data storage unit 9, if a flexible disk driver is used as the audio data storage unit 8, a flexible disk as a storage medium in which audio data is already stored may be mounted, or It is also possible to directly input using a microphone and a line input terminal not shown. Of course, as described above, the voice data storage unit 8
The text data storage unit 9 and the text data storage unit 9 may be combined into a single storage device such as a flexible disk driver or a hard disk driver to be used in common.

Next, a procedure when the user actually speaks a synthetic voice will be described. First, the user operates the voice quality selection unit 5 to select which voice quality should be used for synthesizing the synthesized voice, and whether the voice quality should be further processed, and if so, what kind of processing should be performed. The selection unit 5 is operated to set (step S4).

Next, the user operates the utterance style setting unit 7 to set the utterance style by designating the prosody information. This is specified, for example, according to the content of the sentence to be output as synthesized speech, the listener of the sentence, and the like. After this, the user selects the output data selection unit 10
Is operated to select which one of the text data and the voice data stored in advance in the voice data storage unit 8 and the text data storage unit 9 is to be output as the synthesized voice (step S6).

As described above, the voice synthesizing unit 11 is provided with voice quality data, utterance style data (prosodic data), and sentence data to be output (text data or voice data). When the voice data stored in the voice data storage unit 8 is selected, the voice synthesis unit 11 converts the voice quality data of the voice data into the voice quality data of the sample voice selected by the voice quality selection unit 5, Furthermore, according to the data of the vocalization style selected by the vocalization style setting unit 7,
Speak as synthesized voice from the speaker 12 (step S7)
. When the text data stored in the text data storage unit 9 is selected, the voice synthesis unit 11 adds the voice quality data of the sample voice selected by the voice quality selection unit 5 to the text data, and According to the vocalization style data set by the vocalization style setting unit 7, the speaker 12 speaks as a synthetic voice (step
S7).

As described above, in the embodiment of the voice synthesizing apparatus of the present invention, when the already stored text data or voice data is uttered by the synthetic voice, the voice quality of the pre-registered sample voice is uttered. Therefore, when applied to a personal computer for children, for example, the operation of the personal computer itself,
Or to speak a synthetic voice with the voices of parents, grandparents, or animated characters who are familiar to children such as learning procedures, problem setting, answers, how to obtain answers, reading stories, singing songs, etc. Will be able to do it. Therefore, compared to the conventional dry and dry synthetic speech, the child is more familiar and can use the personal computer with interest.

The voice synthesizing apparatus of the present invention can be applied not only to the above-mentioned personal computer for children but also to various synthesizing voice synthesizing apparatuses such as a karaoke apparatus, an answering machine and a voice guidance apparatus. Needless to say.

[0029]

As described above in detail, according to the speech synthesizer of the present invention, various voice qualities, for example, animation characters and apparatuses can be used in place of the unnatural and uniform synthesized speech that has been used so far. Since it becomes possible to utter a synthetic voice with the voice quality of the person who uses it or the acquaintance of an acquaintance, a device that is familiar to the user is realized. It is also possible to change the utterance style in accordance with the content uttered by the synthetic voice, the person listening to it, and the like.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration example of a speech synthesizer of the present invention.

FIG. 2 is a flowchart showing an operation procedure of the speech synthesizer of the present invention.

[Explanation of symbols]

 3 Voice quality feature extraction unit 7 Vocal style setting unit 8 Voice data storage unit 9 Text data storage unit 11 Voice synthesis unit

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Masashi Ochiiwa 2-5-5 Keihan Hondori, Moriguchi City, Osaka Prefecture Sanyo Electric Co., Ltd. (72) Inventor Kiji Izumi 2 Keihan Hondori, Moriguchi City, Osaka Prefecture 5-5 Sanyo Electric Co., Ltd. (72) Inventor Akashige Sawada 2-5-5 Keihan Hondori, Moriguchi City, Osaka Sanyo Electric Co., Ltd.

Claims (3)

[Claims] 1. A voice synthesizer comprising: a data storage unit for storing voice data and text data to be reproduced as synthetic voice; and a voice synthesis unit for reproducing the data stored in the data storage unit as synthetic voice. In the device, a voice quality feature extraction unit that generates voice quality data by extracting voice quality features of an input voice signal is provided, and the voice synthesis unit uses the voice quality data extracted by the voice quality feature extraction unit to perform the data conversion. A voice synthesizing apparatus, characterized in that the data stored in the storage unit is converted into a synthetic voice for reproduction. 2. A voice synthesizer comprising a data storage unit for storing voice data and text data to be reproduced as synthetic voice, and a voice synthesis unit for reproducing the data stored in the data storage unit as synthetic voice. In the device, the speech synthesis unit includes a vocalization style setting unit for designating a vocalization style at the time of producing a synthetic voice, and the speech synthesis unit uses the vocalization style setting unit to store the data stored in the data storage unit. A voice synthesizer characterized by being reproduced in accordance with a designated utterance mode. 3. A voice synthesizer comprising a data storage unit for storing voice data and text data to be reproduced as synthetic voice, and a voice synthesizer unit for reproducing the data stored in the data storage unit as synthetic voice. The apparatus is provided with a voice quality feature extraction unit that generates voice quality data by extracting voice quality features of an input voice signal, and a voice style setting unit that specifies a voice style when the voice synthesis unit utters a synthetic voice. The voice synthesizing unit synthesizes the data stored in the data storage unit by using the voice quality data extracted by the voice quality feature extracting unit, and according to the utterance style specified by the utterance style setting unit. A speech synthesizer characterized by being played back.