JPS63157226A - Conversation type sentence reading device - Google Patents

Abstract

PURPOSE:To discriminate the sentence read by plural persons by providing a character discriminating part, a character information table and plural phoneme files. CONSTITUTION:A Japanese word dictionary analyzing part 1 refers to a Japanese word dictionary 2 to analyze the Japanese word sentences and divides these sentences into strings of words. Then the part 1 detects the names of characters and delivers them to a character discriminating part 5. While the part 1 divides a word reading string for each utterance unit for the read sentences shown in the brackets and delivers in terns the read words for each utterance unit to a voice parameter setting part 3. Then the part 5 informs the names of characters to a character information table 6 and the table 6 reads out the utterance feature information and also selects a phoneme file 7. The voice parameter, the utterance information, the voice range information, the uttering speed information, the accent information, etc., obtained from the part 3 are outputted to a voice synthesizing part 4 for output of a voice synthesizing signal.

Description

[Detailed Description of the Invention] [Summary] For example, this is an invention for causing a machine to read out lines of a script in which multiple characters appear. For this purpose, a character discrimination section for determining the characters written in the script, a character information table in which vocal characteristics of the characters are registered, and a plurality of phoneme files are provided. If it is written in the script that a character will read the sentence "XX Then, extract the phoneme parameters corresponding to "XX Output audio.

[Industrial application field]

The present invention relates to a conversational text reading device that automatically reads out lines in a script in which multiple characters appear, for example. The field of use of the conversational text reading device of the present invention is in the field of plays and movies, where each actor reads out the scenario while looking at the characters' sentences printed in advance in Japanese, or in the field of reading novels. Accordingly, such systems may be installed in facilities such as libraries for the blind and people with poor eyesight.

[Conventional technology]

FIG. 4 is a diagram showing the configuration of a conventional Japanese language reading device. In the figure, 1 is a Japanese text analysis unit, 2 is a Japanese dictionary, 3 is a voice parameter setting unit, 4 is a voice synthesis unit, 7 is a phoneme file, 8 is a speaker, and 9 is a manual operation panel. .

The Japanese sentence analysis unit 1 divides the Japanese sentence into a string of words while referring to the Japanese dictionary 2, divides the string of words into utterance units of appropriate size, and reads the utterance units in order. It is then sent to the audio parameter setting section 3. In the Japanese dictionary 2, words, pronunciations of words, accent information, etc. are registered. Although only one phoneme file 7 is shown, in reality, a plurality of phoneme files exist, such as a male phoneme file 13 and a female phoneme file. The manual operation panel 9 provides instructions on which phoneme file 7 to use, volume, tone, range,
This is for setting the speaking speed, etc. The audio parameter setting unit 7 retrieves phoneme parameters corresponding to the pronunciation of the unit of utterance from the phoneme file 7 instructed by the manual operation panel 9, and sets the range, timbre, speaking rate, volume, etc. according to the control information from the manual operation panel 9. settings, and send these to the speech synthesis section 4.

The speech synthesis section 4 performs speech synthesis based on the data from the speech parameter setting section 3. The electrical signal output from the voice synthesis section 4 is converted into voice by the speaker 8.

[Problem to be solved]

Conventional Japanese reading devices read out Japanese sentences based on setting information on a manual operation panel, or read out Japanese sentences with a certain tone, voice range, and speaking rate. However, recently, the field of application for reading out Japanese sentences has been expanding, and conventional technology has not been able to cope with fields that require reading aloud by multiple people.

The present invention was created in view of this point, and is a conversational text in which vocal characteristics of characters are registered in advance as data, and readings can be obtained by reading the text aloud by multiple people. The purpose is to provide a reading device.

[Means for solving problems]

FIG. 1 is a diagram showing the principle of the present invention. In the figure, 1 is a Japanese text analysis section, 2 is a Japanese dictionary, 3 is a speech parameter setting section, 4 is a speech synthesis section, 5 is a character discrimination section, 6 is a character information table, 7 is a phoneme file, 8 indicates each speaker.

Generally, multiple characters appear in the script, expressed as A: "XXXX" and Otsu: "XXXX". The Japanese dictionary analysis unit 1
The Japanese sentence is analyzed with reference to the Japanese dictionary 2 and divided into word sequences. When the character name is detected, it is passed to the character discrimination section 5. Regarding the read-aloud sentences in parentheses, the string of word pronunciations is divided into utterance units, and the pronunciations of the utterance units are sequentially passed to the audio parameter setting unit 3. At this time, word accent information is also passed. As shown in FIG. 2, the character information table 6 includes A.

Names of characters such as Otsu, . . . and utterance characteristic information that can identify the utterance characteristics of the characters are registered in advance.

The utterance characteristic information means age, gender, place of birth based on role, phoneme file number, etc.

The contents of the character information table 6 can be rewritten.

When the character identification unit 5 is given a character name (for example, A) by the Japanese text analysis unit 1, the character identification unit 5 inputs the character information table 6.
The contents of the character information table 6 are read out, and it is checked whether the character name A exists in the character information table 6. If it exists, the character determination unit 5 uses the character information table 6
Notify that the character is Party A. Then, the utterance characteristic information of the characters is read out from the character information table 6. Although only one phoneme file 7 is shown in FIG. 1, a plurality of phoneme files 7 exist, such as a phoneme file for men and a phoneme file for women. As shown in FIG. 3, the phoneme file 7 stores phonemes and their corresponding parameters. A phoneme file 7 is selected according to the phoneme file number read from the character information table 6. The age information and hometown information in the read vocalization feature information are passed to the voice parameter setting section 3. The audio parameter setting unit 3 extracts the parameters corresponding to the pronunciation passed from the Japanese sentence analysis unit 1 from the selected phoneme file 7, and the voice parameter setting unit 3 extracts the parameters corresponding to the pronunciation passed from the Japanese sentence analysis unit 1. It determines the speed, etc. and also modifies the accent. The audio parameter setting unit 3 then passes phoneme parameters, volume information, range information, speaking rate information, accent information, etc. to the audio synthesis unit 4. The speech synthesis section 4 synthesizes speech based on these data and outputs a speech synthesis signal.

The speech synthesis section 4 is, for example, of Pacor type.

The electrical speech synthesis signal output from the speech synthesis section 4 is
It is converted into audio by the speaker 8. Note that by analyzing the characters in "XXX" in the input Japanese text, the speaking order is automatically instructed.

When changing the overall sentence structure, editing functions such as moving, inserting, changing, and replacing the Japanese sentences input into the reading device are required. The connection form of the reading device can be considered.

〔Effect of the invention〕

Conventionally, there has been no reading device that has the function of reading out conversational texts for multiple people. It becomes possible to read out conversational texts, for example, automatically read out a scenario or read out aloud according to the characters in a novel.

[Brief explanation of the drawing]

Figure 1 is a diagram showing the principle of the present invention, Figure 2 is a diagram showing an example of the configuration of a character information table, Figure 3 is a diagram showing an example of the configuration of a phoneme file, and Figure 4 is a diagram showing the configuration of a conventional example. be. 1...Japanese text analysis department, 2...Japanese dictionary, 3.
...Audio parameter setting section, 4...Speech synthesis section, 5...
・Character identification unit, 6...Character information table, 7
... Phoneme file, 8 ... Speaker, 9 ... Manual operation panel.

Claims (1)

[Claims] A speech synthesis unit (4), a plurality of phoneme files (7), a Japanese dictionary (2), a Japanese text analysis unit (1), and a Japanese text to be read out. A character identification unit (5) that determines the character who is the entered speaker, and the character's gender, age, place of birth, etc., voice volume, timbre, vocal range, speaking speed, and accent conditions. Based on the character information table (6) that registers vocal feature information that affects , determines the volume, range, accent, etc., selects a phoneme file (7) based on the utterance feature information, and selects the phoneme parameters corresponding to the pronunciation sent from the Japanese language analysis unit (1). A conversational text reading device comprising: a voice parameter setting unit (3) that extracts phoneme parameters, volume, range, and accent from a phoneme file (7) and sends them to a voice synthesis unit (4).