JPH05204390A - Accent giving device and voice synthesis device - Google Patents

Abstract

PURPOSE:To provide a very natural synthetic voice by adding, and emphasis using a voice synthetic speech process and an accent giving method. CONSTITUTION:The device has a text inputting means which inputs a text, a morpbens processing means 1a which separates the inputted text into morphems a dictionary means 1b which stores a dictionary information which is referred by the means 1a, an accent giving means 1d which gives accents and an intonation selecting means 1e which selects intonation. The dictionary information stored in the means 1b holds two accent informations, i.e., a general accent information and an accent information when an emphasis is given. The means 1d gives an accent using these accent informations and based on the intonation selected by the means 1e.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a language processing method used in a speech synthesizer, and more particularly to an accent imparting device.

[0002]

2. Description of the Related Art As a conventional speech rule synthesizer, for example, Furui: Digital Speech Processing p.146 (Tokai University Press)
The device shown in 1985) is known. FIG. 6 is a block diagram showing the configuration of this conventional speech synthesizer.
Kanji and kana mixed sentence data is input to the character string input terminal 0. In the morpheme processing unit 1a, the input kanji / kana mixed sentence is divided into morphemes using the dictionary 1b, and the reading / accent type / part of speech of each morpheme is added. The syntactic analysis unit 1c uses the information of each morpheme obtained by the morpheme processing unit 1a to determine a phrase and performs a dependency analysis between the phrases.
The accent processing unit 1d determines accent phrases, accent positions, and poses and intonation upright positions. These morpheme processing unit 1a and dictionary 1
b, the syntax analysis unit 1c, and the accent processing unit 1d constitute the language processing unit 1. The sound processing unit 2 creates a synthesis parameter based on the reading and accent information obtained by the language processing unit 1. This synthesis parameter includes an amplitude that determines the volume of the voice, a vocal tract description parameter that determines the state of the vocal tract (PARCOR coefficient, formant frequency, etc.),
There is a voiced / unvoiced determination flag that determines the state of the vocal cords, and a fundamental frequency that determines the pitch of the voice. The synthesis processing unit 3 converts the synthesis parameter sequence obtained by the acoustic processing unit 2 into a voice waveform,
A voice waveform is obtained at the synthetic sound output terminal 4.

FIG. 7 is a diagram showing a configuration and an example of the dictionary 1b. For each word, it has information of notation, reading, part-of-speech number, pre-connection number, post-connection number, accent, and combined accent. The pre-connection number and post-connection number are given to check the connection of morphemes. For example, when "I think" is morphologically analyzed, the noun /
Although there are two candidates for the 5-dan verb continuous form, since "masu" is not connected to a noun, the 5-dan verb is selected. The accent is
It shows the accent type when the morpheme is uttered alone, and often does not have an adjunct. The combined accent type indicates how the front part influences the accent when composing a compound word or phrase.

[0004]

However, such a conventional speech rule synthesizing device is being used in various fields such as reading WP sentences and public guide broadcasting.
From the viewpoint of clarity that individual syllables of synthesized speech can be understood, it is approaching a practical level.
If the position of the accent and the number of vocal morae are the same, the same intonation is not given, and the standup and fluctuation of the vocal speed are not added, and from the viewpoint of naturalness, it is said to be very monotonous and mechanical. I have no choice.

The present invention has been made in view of the problems of the conventional voice rule synthesizing device, and provides an accent imparting device capable of outputting a high quality rule synthesizing voice without monotonicity of the rule synthesizing voice. Has a purpose.

[0006]

According to the present invention, there are provided text input means for inputting text, morpheme processing means for dividing the input text into morphemes, and dictionary means for storing dictionary information referred to by the morpheme processing means. , An accent imparting device having an accent imparting means for imparting an accent, having a tone selecting means for selecting a tone and a storing means for holding a plurality of types of accent information, and selecting by the tone selecting means. This is an accent imparting device in which the accent imparting means imparts an accent in accordance with a tone by using accent information.

[0007]

In the present invention, a plurality of types of accent information, for example, a general accent and two types of accent types when emphasized are held, are determined, and which accent type is used according to the selected tone, By adding a general accent or an emphasized accent, it is possible to reduce the monotonicity of the regular synthetic sound and provide a synthetic sound with high naturalness.

Further, a random number is generated and, for example, emphasized reading or general reading is selected to provide a synthetic voice rich in variation.

Further, according to the part-of-speech information of the morpheme or the information written in the dictionary, for example, the emphasized reading or the general reading is selected to provide a synthetic sound with high naturalness.

Further, when a plurality of voice qualities are present, the difference between voice qualities is made clearer by selecting, for example, emphasized reading or general reading according to the selected voice qualities, and a high-quality synthesized voice is provided.

[0011]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing the construction of a speech synthesizer using an embodiment of the accent imparting apparatus of the present invention according to claim 1, and the construction will be described below together with its operation. A kanji-kana mixed sentence is input to the character string input terminal 0. In the morpheme processing unit 1a, the input kanji / kana mixed sentence is divided into morphemes using the dictionary 1b, and the reading / accent type / part of speech of each morpheme is added. The syntactic analysis unit 1c uses the information of each morpheme obtained by the morpheme processing unit 1a to determine a phrase and performs a dependency analysis between the phrases. The accent processing unit 1d determines accent phrases, accent positions, and poses and intonation upright positions. The tone selection section 1e is a means for selecting whether to synthesize by a emphasized voicing with sharpness or general voicing, and 5 is an input terminal of the selection information. The selection information is input to the accent processing unit 1d. The morpheme processing section 1a, the dictionary 1b, the syntax analysis section 1c, the accent processing section 1d, and the tone selection section 1e constitute the language processing section 1.

The sound processing unit 2 creates a synthesis parameter based on the reading and accent information obtained by the language processing unit 1. This synthesis parameter is an amplitude that determines the volume of the voice, a vocal tract description parameter that determines the state of the vocal tract (PARCOR coefficient, formant frequency, etc.), a voiced / unvoiced determination flag that determines the state of the vocal cord, and a basic that determines the pitch of the voice. Frequency etc. The synthesis processing unit 3 converts the synthesis parameter sequence obtained by the acoustic processing unit 2 into a voice waveform, and obtains a voice waveform at the synthetic voice output terminal 4. In this embodiment, the methods of the sound processing unit 2 and the synthesis processing unit 3 are not particularly limited.

FIG. 2 is a block diagram showing the structure and an example of the auxiliary word dictionary of the dictionary 1c in this embodiment. In addition to the dictionary items of the conventional example, it has an accent accent type. The emphasized accent type is information about the possibility of excellence and the formation of unique accent phrases, and if the number is 1 or more, the possibility exists.

Details of each process will be described using a specific example. Consider the case where the sentence "This is a speech synthesizer." Is entered. The input sentence is morpheme-divided by the morpheme processing unit 1a as follows to obtain accent and reading information. Here, the combined accent types A and b given to "ha" and "desu" are those listed in the NHK Accent Dictionary / Explanatory Appendix (Japan Broadcasting Corporation 1985), and combined with independent words. It shows the combined accent type when constructing a phrase. The accent type of each word indicates a syllable position with an accent. The combined accent type of an independent word indicates an accent type when it becomes a subsequent word of a compound word. In this example, the emphasized accent type is set only to the case particle “ha”. Further, the syntactic analysis unit 1c determines the phrase boundary,
Dependence between phrases (distance between phrases) is determined. In this example, the bunsetsu “korewa” is applied to the bunsetsu immediately after, so that the bunsetsu distance is 1. The accent processing unit 1d determines the accent phrase and the accent position. In this example,
"This is" part is flat type, "Speech synthesizer" part is 9 type, 9 syllables, that is, "device" with "accent" (language processing unit / process result "so" The "1" added to the back indicates accented syllables). (Input sentence) "This is voice synthesis." (Morpheme division) This is /, /, voice / synthesis / device. (Reading) Korewa Onsei Kosei Sauce Dise (Accent) 0-1 10 0- (Joined Accent) 1 A 1 1 1 1 1 b 1 1 1 1 1 b (Emphasis Accent) -1 ------ (Part of speech) (Phrase) This is a / voice synthesizer. (Phrase distance) 1 (Accent phrase) This is a / voice synthesizer. (Phrase accent) 09 (Language processing unit output 1) ko re wa poz o nn se e go o se e so1 o ti de su (Language processing unit output 2) ko re wa ' poz o nn se e go o se e so1 o ti de su Here, poz indicates that a pose can be inserted at this position to restore the intonation. The language processing unit output 1 indicates the case of general reading, and the output 2 indicates the case of emphasized reading. In the case of emphasized reading, ['] added to [wa] indicates superiority, and when the intonation is given by the sound processing unit 2, the fundamental frequency of [wa] is increased as excellence as compared with the case of general reading. It is set higher.

When the morpheme to be emphasized is one syllable, it is processed as outstanding, but when it is more than two syllables, an original accent phrase is formed. For example, if you process the sentence "I don't understand," it will be "bo1 ku de wa poz wa kara1 na i" in general reading, and "bo1 ku / de1 wa poz wa ka ra1 na i" in emphasized reading. Here, "/" means a break in the accent phrase.

As described above, according to the present embodiment, it is possible to select a general accent (intonation) and an accent in which standup and emphasis are added, and to reduce the mechanical monotonicity of the rule-synthesized sound and to make it natural. It is possible to provide a synthetic sound with high property.

Embodiment 2 FIG. 3 is a block diagram showing the configuration of an embodiment of a voice synthesizing apparatus using the accent imparting apparatus of the present invention as defined in claim 2. The same elements as those in the embodiment of FIG. 1 are designated by the same reference numerals and the description thereof will be omitted. Tone selection terminal 5
There is a random number generating means 1f instead. Random number generator 1
The tone selection section 1e selects a tone according to the random number generated in f. By generating a random number and selecting emphasized reading or general reading, it is possible to provide a synthetic sound that changes.

Embodiment 3 FIG. 4 is a block diagram showing the construction of an embodiment of a voice synthesizing apparatus using the accent imparting apparatus of the present invention as defined in claim 3. The same elements as those in the embodiment of FIG. 1 are designated by the same reference numerals and the description thereof will be omitted. Tone selection terminal 5
Instead of this, the tones are determined according to the information obtained from the dictionary 1b. For example, using part-of-speech information, when following a noun / proper noun, it is emphasized,
No emphasis is placed on other parts of speech. In the case of "This is a speech synthesizer", "ha" is a general utterance,
In the sentence "This is the voice synthesizer.", "Ha"
Is uttered outstandingly. In the present embodiment, the tone is determined using the part-of-speech information, but this does not restrict the present invention. For example, it is possible to emphasize all the words in the user dictionary, or to newly have information that each word is emphasized. In this way, a synthetic sound with high naturalness can be provided by highlighting reading according to morpheme part-of-speech information or information written in a dictionary and selecting general reading.

Embodiment 4 FIG. 5 is a block diagram showing the configuration of an embodiment of the speech synthesizer of the present invention according to claim 4. Instead of the tone selection terminal 5 in the case of FIG. 1, it has a voice quality selection terminal 6 and a voice quality selection unit 7 connected to the terminal 6. Voice quality selector 7
Is connected to the tone selection unit 1e and the sound processing unit 2. When a plurality of voice qualities of the synthetic sound are prepared, when the voice quality is selected by the voice quality selection unit 7, the tones selection unit 1e can automatically select the tone according to the selection. The accent processing unit 1d uses it to perform accent processing. As described above, by selecting the emphasized reading and the general reading, the difference between voice qualities can be made clearer and a high-quality synthesized speech can be provided.

Each means of the present invention may be realized by software using a computer, or may be realized by using a dedicated hardware circuit having each of these functions.

As the accent information of the present invention,
The accent reading is not limited to the above-described emphasized reading, and any type of accent information may be used.

The accent information need not always exist in the dictionary and may be stored in another storage means.

Further, the accent imparting device and the voice synthesizing device of the present invention do not necessarily have to undergo the syntax analysis.

[0025]

As described above, according to the present invention, it is possible to add an accent, which is the basis of intonation, in consideration of prominence and emphasis, and reduce the mechanical monotony of the rule-synthesized sound. It is possible to provide an effective synthetic sound that is highly natural and natural.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an accent imparting device according to an embodiment of the present invention according to claim 1.

FIG. 2 is a block diagram showing a configuration of an auxiliary word dictionary of the embodiment.

FIG. 3 is a block diagram showing a configuration of an accent imparting device according to an embodiment of the present invention according to claim 2.

FIG. 4 is a block diagram showing a configuration of an accent imparting device according to an embodiment of the present invention according to claim 3;

FIG. 5 is a block diagram showing a configuration of a voice synthesizing apparatus according to an embodiment of the present invention according to claim 4.

FIG. 6 is a block diagram showing a configuration of a conventional speech synthesizer.

FIG. 7 is a block diagram showing the structure of a conventional auxiliary word dictionary.

[Explanation of symbols]

 0 text input terminal 1 language processing unit (means) 1a morphological processing unit (means) 1b dictionary 1c syntactic analysis unit (means) 1d accent processing unit (means) 1e tone selection unit (means) 1f random number vocalization unit (means) 2 sound Processing unit (means) 3 Synthesis processing unit (means) 4 Synthetic sound output terminal 5 Tonal input terminal 6 Voice quality input terminal 7 Voice quality selection section (means)

Claims (6)

[Claims] 1. A text input unit for inputting text, a morpheme processing unit for dividing the input text into morphemes, a dictionary unit for storing dictionary information referred to by the morpheme processing unit, and an accent imparting for imparting an accent. An accent imparting device including means, comprising: a tone selection means for selecting a tone and a storage means for holding a plurality of types of accent information, and the accent according to the tone selected by the tone selection means. An accent imparting device, wherein the accent imparting means imparts an accent using information. 2. A text input means for inputting text, a morpheme processing means for dividing the input text into morphemes, a dictionary means for storing dictionary information referred to by the morpheme processing means, and an accent attachment for giving an accent. An accent applying device including means, a random number generating means for generating a random number, a tone selecting means for selecting a tone,
Storage means for storing a plurality of types of accent information, selecting a tone by the tone selection means according to a random number generated by the random number generation means, and the accent according to the tone selected by the tone selection means. An accent imparting device, wherein the accent imparting means imparts an accent using information. 3. Text input means for inputting text, morpheme processing means for dividing the input text into morphemes, dictionary means for storing dictionary information referred to by the morpheme processing means, and accent addition for giving accent In the accent imparting device including means, there is provided a tone selection means for selecting a tone and a storage means for holding accent information relating to the dictionary information, and the tone selection is performed by the tone selection means from the accent information. An accent imparting device, characterized in that the accent imparting means imparts an accent in accordance with the tone selected by the tone selecting means. 4. The accent imparting device according to claim 1, wherein the dictionary means also serves as the storage means. 5. A text input means for inputting text, a morpheme processing means for dividing the input text into morphemes, a dictionary means for storing dictionary information referred to by the morpheme processing means, and an accent attachment for giving an accent. Means and a voice synthesizing means for synthesizing a voice, in an accent imparting device, a tone selecting means for selecting a tone, a voice quality selecting means for selecting a voice quality of a synthesized voice, and a storage means for holding a plurality of types of accent information. The tone quality selecting means selects a tone according to the voice quality selected by the tone quality selecting means, and the accent assigning means using the accent information according to the tone selected by the tone quality selecting means. A voice synthesizer characterized by adding an accent. 6. The speech synthesizer according to claim 5, wherein the dictionary means also serves as the storage means.