JPH0628900U - Voice message editing device - Google Patents

Abstract

(57) [Summary] [Purpose] To enable data correction work efficiently
It [Structure] Create a Japanese sentence of the message to be edited
(S₁), Text-analyze the Japanese sentence (S₂), So
Create data for speech synthesis based on the result of
(S ₃). Next, start the voice message creation program
(S_Ten), Utters the voice synthesis data by voice synthesis
Then, check if this is satisfied (S₁₁),Dissatisfaction
, The key for the displayed voice synthesis data
Due to the operation of the board, reading errors, accent errors,
By changing, deleting, joining, etc.
Correct (S₁₂). Immediately after this step S₁₁Back to
And the modified data for voice synthesis is voice-synthesized.
Be uttered. At that time, specify the clause, sentence, etc.
You can utter only minutes.

Description

[Detailed description of the device]

[0001]

[Industrial applications]

 The present invention is applied, for example, to the creation of a voice message used in a voice response service, reads the input Japanese text, sets the accent and inter-phrase coupling degree, converts it into voice synthesis data, and displays it. Relates to a voice message editing device for correcting voice synthesis data and editing a voice message.

[0002]

[Prior art]

 In the conventional voice message editing device, a voice synthesis routine is not included in the message editing program. Therefore, in order to confirm the edited result by voice, the message edit program is terminated once, and then the voice synthesis program is started, and the voice synthesis data of the edited result is voice-synthesized and uttered.

That is, conventional editing of voice messages has been performed as shown in FIG. First, a message sentence is created as a Japanese sentence mixed with kanji (S ₁ ), the Japanese message sentence is text-analyzed (S ₂ ), and voice synthesis data is created based on the analysis result (S _3). ). Next, a voice synthesizing program is started (S ₄ ), the voice synthesizing data is voice synthesized and uttered, the synthesized voice is auditioned, and it is confirmed whether or not this is satisfactory (S ₅ ). . If it is insufficient, the voice synthesis program is terminated (S ₆ ) and the document editing program is started (S ₀ ). By looking at the displayed voice synthesis data, the reading error, the accent position error, the phrase connection degree error, etc. are corrected by operating the keyboard to correct the voice synthesis data (S). ₇ ). After that, the document editing program is terminated and the process returns to step S ₄ (S ₈ ).

As described above, each time the speech synthesis data is modified, the speech synthesis program is started, and the modified speech synthesis data is subjected to speech synthesis and auditioned. Was being repeated.

[0005]

[Problems to be solved by the device]

 As described above, in the past, the voice synthesis program was started to synthesize the voice synthesis data by voice, the voice synthesis program was terminated, and then the document editing program was started to correct the voice synthesis data. Then, the document editing program is terminated, the speech synthesis program is started to audition the corrected speech synthesis data, and the same procedure is repeated. In other words, the activation and termination of the voice synthesis program and the activation and termination of the text editing program were repeated, and the switching operation of both software was relatively complicated. In particular, even if the number of switching operations is relatively small, it is troublesome to modify the voice synthesis data many times.

Further, in the speech synthesis by the speech synthesis program, the entire speech synthesis data to be edited is synthesized by one speech at a time, and the utterance is made. I couldn't do that, and it was hard to judge whether the correction I made earlier was good or not.

[0007]

[Means for Solving the Problems]

 According to this invention, in the document editing program, there is provided a means for designating a phrase, a sentence, and a sentence of the voice synthesis data, and performing a voice synthesis of the designated phrase, sentence, and a sentence and outputting them.

[0008]

【Example】

 An embodiment of the voice message editing device according to the present invention is shown in FIG. When Japanese text including Kanji is input from the input terminal 1 to the text analysis unit 2, the text analysis unit 2 sets readings for Kanji, accents of phrases, and degree of connection between phrases, and outputs data for voice synthesis. Form. This voice synthesis data is sent to the data display operation unit 4 which is converted into display data by the voice synthesis data editing unit 3. The data display operation unit 4 displays data and performs insertion, deletion, and change processing of kana characters and symbols according to the operator's instructions. The voice synthesizing data editing unit 3 converts the data received from the data display operating unit 4 from the display data format to the voice synthesizing data format at any time, and when the data display operating unit 4 receives an instruction from the data display operating unit 4, the data is edited. The voice synthesis data is sent to the voice synthesizer 5. The voice synthesis unit 5 synthesizes the voice message 6.

An operation method at the time of editing the voice synthesizing function will be described. After editing, move the editing cursor to any position in the target phrase with the arrow keys, and then perform the following operations. * Phrase synthesis: Press the "F6" key. Process outline: Speech synthesis is performed for the phrase in which the cursor currently exists. * Partial composition: Press the "F7" key.

Outline of processing: Speech synthesis is performed from the phrase where the cursor is currently located to the end of the sentence. * Sentence synthesis: Press the "F8" key. Process overview: Speech synthesis is performed from the beginning of the sentence currently being edited to the end of the sentence. * Region composition: Press the "F9" key. Process overview: Performs voice synthesis for a specified area. Area is specified in units of sentences.

The flow of processing in this invention is shown in FIG. Steps corresponding to those in FIG. 2 are denoted by the same symbols. In this invention, when the voice synthesis data is obtained in step S ₃ , the voice message creation program is started (S ₁₀ ), and then the voice synthesis data at that time is voice synthesized and uttered. Confirm whether or not it is (S ₁₁ ). If not satisfied, the voice synthesis data displayed on the data display operation unit 4 is corrected and the process returns to step S ₁₁ (S ₁₂ ). In other words, the corrected voice synthesis data is immediately voice-synthesized, and in that case, as described above, a phrase or sentence can be designated and only that portion can be uttered. When editing, the editing cursor is located at the editing location, so the voice synthesis of that part can be performed with one key operation without moving the editing cursor, which is very easy to operate. Moreover, since only the part of interest is uttered, the listening confirmation can be performed well.

In this voice message editing apparatus, in the voice synthesizing data editing process, V-RAM is used for the editing data in order to speed up the editing speed such as character input, insertion, and deletion. There is. In addition, the voice synthesis data corrected in the voice synthesis process, the paging process, and the like are syntax-checked to see if they are in the correct format as the voice synthesis data. Only normal V-RAM data should be stored in the buffer. It has become. This plays the role of preventing rudimentary editing mistakes during editing.

[0013]

[Effect of device]

 As described above, according to the present invention, the voice synthesis function is added to the message editing software, so the correction contents can be immediately confirmed by the synthesized voice with a simpler operation than before, and the correction portion Only the synthesized voice can be uttered, and the audition confirmation can be performed well and the data correction work efficiency is improved.

[Brief description of drawings]

FIG. 1 is a flowchart showing a processing flow in an embodiment of the present invention.

FIG. 2 is a flowchart showing a conventional voice message editing procedure.

FIG. 3 is a block diagram showing a general configuration of a voice message editing device.

Claims (1)

[Scope of utility model registration request] 1. A voice message in which input Japanese text is read, accents and degree of inter-phrase coupling are set, converted into data for voice synthesis and displayed, and the data for voice synthesis is corrected by observing the display. A voice message editing device for editing a voice message editing device, comprising: a phrase of the voice synthesizing data, and a unit for voice-synthesizing the designated phrase and sentence when the sentence is designated.