JPH06195094A - Phonogram string display method and speech synthesizing device - Google Patents

Abstract

(57) [Summary] [Purpose] When displaying phonetic character strings in a speech synthesizer, by arranging the notation of each syllable unit in a display space at equal intervals, the temporal positional relationship of each syllable is displayed. It is possible. [Structure] The input sentence "red lantern" is divided into reading information and tonal component information by the language processing unit. The reading information is written in Roman letters, for example, and is called phonetic text. The phonetic text is composed of one or more alphabets, and the syllable containing the most character strings is extracted from the phonetic texts of each syllable. In (b) of the figure, "CH
The three characters "O" are the longest character string, and a display space for three characters is secured for each syllable in the display of phonetic text. Then, the phonetic character string is displayed left-justified for each syllable ((c) in the figure). As described above, the phonetic texts are stored one syllable in the equally spaced display space.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a phonetic character string display method and a voice synthesizer. More specifically, the present invention relates to a phonetic character string display method and a voice synthesizing device for displaying reading information obtained from a kanji / kana mixed sentence in voice synthesis for converting character information into voice.

[0002]

2. Description of the Related Art Conventionally, there is a voice rule synthesizer for converting character information into voice and outputting the voice. When converting a kanji / kana mixed sentence into speech by this voice rule synthesizing device, first, a process of converting a kanji / kana mixed sentence into “reading” and “tone component (intonation, accent component)” (hereinafter referred to as language processing )I do. Next, it is necessary to adjust the voice parameter from the "reading" and the "tone component" generated by the language processing and convert it into voice (hereinafter referred to as acoustic processing).

As data to be passed from the language processing to the acoustic processing, editable text format data (hereinafter referred to as phonetic text) is generally used. This phonetic text can be easily corrected when the result output by the language processing is incorrect. In this phonetic text, the reading information is written in Roman letters or kana (Hiragana, Katakana), etc., and the tone information (accent position, pause position, etc.) is included in the reading information. Texts are commonly used.

[0004]

However, in the above-mentioned conventional voice rule synthesizer, the number of characters in the phonetic character string for each syllable in the phonetic text changes depending on the kind of the sound represented by the phonetic text. For example, the phonetic character string of the syllable "cho" is "CHO" and is represented by three characters. Also, the phonetic character string of the syllable "A" becomes "A" and is expressed by one character. In this way, since the phonetic character strings that represent the readings of each syllable are unequal in length, it is difficult to obtain the timing of occurrence (temporal position of generating a certain syllable) from the phonetic character string. There is a drawback that. That is,
In Japanese vocalization, rhythms are created in syllable units, so
Each syllable is uttered at almost regular intervals in time. However, in the phonetic text, the reading string corresponding to each syllable is
It is not written at equal intervals for each syllable.

For example, when a kanji-kana-mixed sentence "red lantern" is converted into phonetic text reading, it is written in Roman letters as "AKAICHO-TIN". When this notation is broken down into syllable units, "A", "KA", "
I "," CHO ","-"," TI "," N ",
It can be seen that the reading length of each syllable varies from syllable to syllable. Also, when the reading of "red lantern" is expressed in Hiragana, it becomes "Akaichichochin", but even in this case, only the reading of the syllable "Cho" is two characters, and each syllable is equally spaced. Not lined up.

The present invention has been made in view of the above problems, and in displaying a phonetic character string, by arranging the notation of each syllable in a display space at equal intervals, the time of each syllable is increased. An object of the present invention is to provide a phonetic character string display method and a voice synthesizer capable of displaying a positional relationship.

[0007]

A speech synthesizer of the present invention which achieves the above object has the following configuration. That is, a speech synthesizing device that analyzes a sentence to generate reading information and tonal component information, and generates synthetic speech based on these information, and is a phonetic character with the largest number of characters included in one syllable. Space determining means for determining a display space of a phonetic character string of each syllable based on the number of characters in a string; and display means for displaying a phonetic character string of one syllable in the display space determined by the space determining means. , Is provided.

The phonetic character string display method according to the present invention for achieving the above object includes the following steps. That is, it is a phonetic character string display method of displaying reading information obtained by analyzing a sentence in a phonetic character string, and the phonetic character string display method is based on the number of characters of the phonetic character string including the largest number of characters in one syllable. A space determination step of determining a display space of a syllabic phonetic character string and a display step of displaying a phonetic character string of one syllable in the display space determined by the space determining means.

In the present invention, the phonetic character string is composed of one or more characters or symbols for representing reading information.

[0010]

With the above configuration, in the phonetic character string representing the reading information obtained by analyzing the kanji / kana mixed sentence, etc., each phoneme character string containing the largest number of characters in one syllable Determines syllable-based display space. Then, the phonetic character strings are arranged and displayed for each syllable in the determined display space. By arranging the phonetic character strings in the display space determined for each syllable in this way, the correspondence between the time interval of utterance and the notation position of the phonetic can be easily grasped.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described below with reference to the accompanying drawings.

FIG. 1 is a block diagram showing the schematic arrangement of a speech rule synthesizing apparatus according to this embodiment. In the figure, reference numeral 11 denotes a CPU, which executes various controls in the speech synthesis rule device. A ROM 12 stores various control programs for the CPU 11 to execute processing. Incidentally, R
The OM 12 also stores a control program represented by the flowchart of FIG. 3 described later. Reference numeral 13 denotes a RAM, which temporarily stores data and the like required when the CPU 11 executes various controls. An input unit 14 inputs various data, control commands, and the like. It should be noted that the input unit 14 also performs edit input when editing the phonetic text obtained by the language processing of the present speech rule synthesizing device. Reference numeral 15 is a dictionary in which readings and accent information of kanji and the like are registered, and is referred to in a language process for analyzing the input kanji / kana mixed sentence to obtain reading information. 4 is a display unit,
Various displays such as display of phonetic text indicating reading of sentences are made. Reference numeral 6 denotes a sound processing unit, which performs voice synthesis from the obtained phonetic text and converts it into a voice signal. Then, the speaker 16 outputs the converted audio signal.

FIG. 2 is a block diagram showing the functional arrangement of the speech rule synthesizing apparatus of this embodiment. In the figure, reference numeral 1 is a language processing unit, which analyzes an input sentence containing kanji and kana and generates "reading" and "tone component". Reference numeral 2 denotes a character string length counter unit, which obtains the character string length of the reading notation corresponding to each syllable. Reference numeral 3 denotes a phonetic text generation unit, which generates phonetic text based on the "reading" obtained by the language processing unit 1. A display unit 4 displays phonetic text and the like.
An editing unit 5 edits the phonetic text displayed on the display unit 4. A sound management unit 6 generates a sound from phonetic text.

The operation of the speech rule synthesizing device of the present embodiment having the above-mentioned structure will be described below with reference to the flowchart of FIG.

FIG. 3 is a flow chart showing the control procedure in the speech rule synthesizing device of this embodiment. First, in step S11, the language processing unit 1 analyzes the target kanji / kana-mixed sentence and generates information of “reading” and “tone component”. At this time, regarding the "reading" information, the boundary of each syllable should be known. And step S1
In 2, the character string length counter unit 2 counts the character string length of each phonetic character string representing “reading” to obtain the number of characters of the phonetic character string having the maximum character string length. For example, "Cho"
“CHO”, which is the reading notation of, has a character string length of 3 characters.

In step S13, 1 is added to the maximum character string length.
Determines the space for writing syllabic phonetic strings. Next, in step S14, the phonetic text generation unit 3 arranges the reading and writing character strings corresponding to each syllable at equal intervals in syllable units. That is, the longest one of the character string lengths obtained by the character string length counter unit 2 is set as the character string length used for reading and writing all syllables in the phonetic text, and the reading of each syllable is included in the character string length. Place the notation.

FIG. 4 shows the processing contents in each of the above steps.
Will be further described with reference to. FIG. 4 is a diagram showing a state in which a phonetic text is obtained from an input sentence. The input sentence is “red lantern” as shown in FIG. When this is written in Roman alphabet, as shown in FIG. 4 (b),
The reading "CHO" consists of three characters and is the longest. Therefore, this length (3 characters) is used as a space for reading and writing each syllable. In this way, a space of three-character sentences is given to the reading character strings of all syllables, and the reading of each syllable is arranged in the space of the three-character sentences ((c) of FIG. 4).
In FIG. 4 (c), phonetic character strings in which readings are written in Roman letters are arranged left-justified in a space for three characters. On the other hand, FIG. 5 shows an example in which phonetic character strings are arranged right-justified in a space for three characters.

Further, FIG. 6 shows a state in which the phonetic text is obtained from the input sentence when the reading is written in hiragana. When hiragana is used for phonetic text, the reading "cho" has the longest character string, as is apparent from FIG. 6 (b). Therefore, the phonetic character string is arranged in the space for two characters ((c) of FIG. 6).

In the following step S15, the phonetic text generated as described above is displayed on the display unit 4.
Then, if there is an input for correcting the phonetic text in step S16, the process proceeds to step S17, and the editing unit 5 performs the correction. Here, a display device of a character terminal, a console such as a workstation, or the like is used for the display unit 4, and a text editor or the like is used for the editing unit 5. After the correction is executed in step S17, step S
Returning to 15, the correction content is displayed.

On the other hand, if the text correction is unnecessary in step S16, the process proceeds to step S18, and it is determined whether or not voice output is executed. Then, when the execution of voice output is instructed, the phonetic text completed as described above is converted into a voice signal in the acoustic processing unit 6 and reproduced and output from the speaker 16.

As described above, by arranging the phonetic notation corresponding to each syllable in the phonetic text at equal intervals in syllable units, it is possible to clearly indicate the timing of occurrence of each syllable.

In the above embodiment, the character string length counter unit 2 obtains the character string length of the reading notation of each syllable. However, if the reading notation of each syllable is determined in advance,
It is not necessary to calculate the character string length every time. In that case, the character string length counter unit 2 becomes unnecessary, and the maximum character string length of the phonetic reading of each syllable that is known in advance can be used as the character string length used for the phonetic reading of each syllable.

Further, in the above embodiment, a certain character string length is given as a space for writing and writing each syllable in the phonetic text, and the reading and writing of each syllable is described in the character string length. I have to. Therefore, when the phonetic transcription of a certain syllable is shorter than the given character string length, a blank character is inserted. However, by changing the character width, it is possible to display the phonetic notation in a given space. Display examples in that case are shown in FIGS.
Shown in. Figure 7 is an example of a phonetic character string in Roman letters,
FIG. 8 is an example of a phonetic character string in hiragana. In these cases, in the display unit 4 in FIG. 2, the character width is calculated so that the phonetic character string of each syllable falls within the given display space, and the phonetic text is displayed with the character width.

As described above, in the voice synthesizing device of this embodiment, in displaying the phonetic character string, the phonetic character string corresponding to each syllable is arranged at an appropriate position in the display space,
Alternatively, by appropriately changing the character width and arranging the readings and notations at equal intervals in each syllable, the temporal positional relationship of each syllable can be clearly expressed.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of a single device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0026]

As described above, according to the phonetic character string display method and the voice synthesizing device of the present invention, in the display of the phonetic character string, the notation of each syllable unit is arranged in the display space at equal intervals. By doing so, it becomes possible to display the temporal positional relationship of each syllable, and it becomes easy to grasp the temporal relationship of each syllable.

[0027]

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a voice rule synthesizing device according to this embodiment.

FIG. 2 is a block diagram showing a functional configuration of a voice rule synthesizing device according to this embodiment.

FIG. 3 is a flowchart showing a control procedure in the voice rule synthesizing device of the present embodiment.

FIG. 4 is a diagram showing how a phonetic text is obtained from an input sentence.

FIG. 5 is a diagram showing an example in which phonetic character strings are arranged right-justified in a space for three characters.

FIG. 6 is a diagram showing a state from input sentence to obtaining phonetic text when reading is written in hiragana.

FIG. 7 is a diagram showing an example in which a phonetic character string in Roman characters is displayed with the character width changed according to the size of the display space.

FIG. 8 is a diagram showing an example in which a phonetic character string in hiragana is displayed with the character width changed according to the size of the display space.

[Explanation of symbols]

 1 language processing unit 2 character string length counter unit 3 phonetic text generation unit 4 display unit 5 editing unit 6 acoustic processing unit 11 CPU 12 ROM 13 RAM 14 input unit 15 dictionary 16 speaker

Claims (6)

[Claims] 1. A voice synthesizer that analyzes a sentence to generate reading information and tonal component information, and generates a synthesized voice based on these information, and has the largest number of characters included in one syllable. Space determining means for determining the display space of the phonetic character string of each syllable based on the number of characters of the phonetic character string, and displaying the phonetic character string of one syllable in the display space determined by the space determining means. A speech synthesis apparatus comprising: a display unit. 2. A voice synthesizing device for generating reading information and tone component information by analyzing a sentence and generating synthetic speech based on these information, wherein a table of each syllable unit representing the reading information. Counting means for counting the number of characters included in the phonetic character string, as a result of counting by the counting means, extraction means for extracting the phonetic character string having the largest number of characters, and the phonetic character string extracted by the extracting means. Space determining means for determining a display space of a phonetic character string of each syllable based on the number of characters included therein; and display means for displaying a phonetic character string of one syllable in the display space determined by the space determining means. A voice synthesizer comprising: 3. The display means displays a phonetic character string of one syllable for each of the display spaces determined by the space determining means, and a phonetic character string corresponding to each syllable is displayed in the display space. The speech synthesizer according to claim 1 or 2, characterized in that if there is an extra space when placed, it is left blank. 4. The display unit for each syllable so that the display unit displays the phonetic character string of one syllable for each of the display spaces determined by the space determination unit in a given display space. The voice synthesizer according to claim 1 or 2, wherein the characters of the phonetic character string are displayed with different widths. 5. A phonetic character string display method for displaying reading information for voice synthesis in a phonetic character string, which is based on the number of characters in a phonetic character string including the largest number of characters in one syllable. A space determination step of determining a display space of a phonetic character string of each syllable, and a display step of displaying a phonetic character string of one syllable in the display space determined by the space determining means. Characteristic phonetic character string display method. 6. A phonetic character string display method for displaying reading information for voice synthesis in a phonetic character string, wherein the number of characters included in the phonetic character string of each syllable unit representing the reading information is A counting step of counting, an extraction step of extracting a phonetic character string having the largest number of characters as a result of the counting by the counting step, and each syllable based on the number of characters included in the phonetic character string extracted by the extracting step. And a display step of displaying a phonetic character string of one syllable in the display space determined by the space determining step. Phonetic string display method.