JP2586243Y2 - Voice message editing device - Google Patents

Description

[Detailed description of the invention]

[0001]

The present invention is applied to, for example, the creation of a voice message used for a voice response service. The Japanese text including the input kanji is read for the kanji by text analysis means, the accent form of the phrase, the interval between phrases. A voice message editing device that sets the degree of connection of the voice message, converts the data into voice synthesis data in a predetermined format, displays the data, and corrects the voice message by keyboard input or the like to edit the voice message. It is related to the modification of the coupling degree.

[0002]

2. Description of the Related Art Conventionally, the degree of inter-phrase connection at the time of editing a voice message of this type has been modified by changing and modifying inter-phrase connection symbols different from a predetermined degree of connection. For example, "Autumn trip was decided in Shimoda, Izu. ] And a Japanese sentence, the next data for speech synthesis is automatically created.

[0003] Aquino [* 01] Ryokowa [00] Izuno [/ 00] Scarlet mite [/ 00] Chimarimasita [. 0
4] Here, the kana sequence indicates the reading, and the two-digit number in [] indicates the position from the beginning of the accented syllable,
“*”, “/”, “,”, “.” Indicate the degree of connection of clauses. The person who performs the editing checks the meaning of each of the phrase connection symbols indicating the degree of the phrase connection in advance by looking at the display of the data for speech synthesis, or from time to time, and determines whether or not the degree of the phrase connection is appropriate. Judgment is made, and if inappropriate, the symbol is changed to another appropriate phrase combining symbol.

[0004]

[Problems to be Solved by the Invention] As described above, conventionally, the editor needs to know the meaning of each phrase connecting symbol in advance or at any time, that is, the editor needs specialized knowledge and experience. Was disadvantageous. It is an object of the present invention to provide a voice message editing device that enables even a person with little technical knowledge or an inexperienced person to easily create a high-quality voice message without a clause connection error in a voice message. .

[0005]

According to the present invention, each phrase combining symbol in speech synthesis data of a predetermined format using a different phrase combining symbol determined in advance for each degree of phrase combining is used. By the connecting symbol conversion means, it is converted into an array of the same specific symbols of the number corresponding to the degree of the phrase connection, so that the editor can immediately know the individual degree of the phrase connecting from the number of the specific symbols and correct it. Can also be easily performed. Also, the speech synthesis data including the modified phrase connection degree of the same specific symbol sequence is inversely converted into conventional speech synthesis data using a different phrase connection symbol for each phrase connection degree.

[0006]

FIG. 1A shows an embodiment of a voice message editing apparatus according to the present invention. When a Japanese text including a kanji is input to the text analysis unit 2 from the input terminal 1, the text analysis unit 2 sets the reading for the kanji, the accent form of the phrase, and the degree of connection between the phrases, and a predetermined conventional format. Create voice synthesis data. The data for voice synthesis is converted into a display data format by the data synthesis unit for voice synthesis 3 and sent to the data display / operation unit 4. The data display operation unit 4 displays the input data on a display, and performs insertion, deletion, and change processing of kana characters and symbols in accordance with instructions from the operation unit of the operator. The voice synthesis data editing unit 3 converts the correction data received from the data display / operation unit 4 from the display data format to the voice synthesis data format as needed, and based on an instruction from a data display / operation unit 4 based on a command or the like. , And sends the corrected data for speech synthesis to the speech synthesis unit 5. The voice synthesizer 5 synthesizes the voice message 6.

[0007] The voice message editing apparatus according to the present invention displays voice synthesizing data for facilitating voice message editing by an operator in a visually understandable form.
That is, the phrase connection degree symbol is a predetermined specific symbol, for example, an edit screen, irrespective of the type of symbol such as "*", "/", ",", "." Upper underbar “ ”Alone, and the strength of the bond is indicated by the number.

That is, as shown in FIG. 1B, the strong coupling "*"
One underbar [ ], The weak link "/" is two en
Durbar [   ], Short pose "" (space) is 3
Underbars [     ], Medium pose “,” is 4
Underbars [ ], The long pose "." Has 5 unders
Bar [         ]. Further in this example
Under bar " If there are three or more, pause (pause)
Is inserted between the phrases to distinguish it from the case without pause.
Therefore, the underbar is displayed in green, and when there are two or less
Displays underbars in white color and indicates the meaning of phrase connection
Is easily grasped visually.

In the text analysis unit 2, an input text is created as speech synthesis data in a conventional format. In the present invention, each of the clause connection symbols in the speech synthesis data is used as a connection symbol in the speech synthesis data editing unit 3. By the conversion means 11,
With the relationship shown in FIG. The data for speech synthesis including the converted symbol string is converted to display format data, and the data display / operation unit 4
Sent to and displayed. This display is corrected for reading errors, accent errors, and the degree of phrase connection in the same manner as before, and the corrected speech synthesis data is sent to the speech synthesis data editing unit 3 and the display format is changed to the speech synthesis data. Then, the underbar symbol string indicating the degree of each phrase connection is inversely converted into a different phrase connection symbol in accordance with the number thereof by the connection symbol reverse conversion means 12 according to the relationship shown in FIG. 1B. In this way, the data is returned to the voice synthesizing unit 6 in the same format as the conventional one, and sent to the voice synthesizing unit 6.

For example, when the following Japanese sentence is input,
"The autumn trip has been decided in Shimoda, Izu. ] It is automatically converted to the following speech synthesis data. Aquino [* 01] Ryokowa [00] Izuno [/ 0
0] Dust mite [/ 00] Kimarimasita [. 04] In the data for speech synthesis, the phrase connection symbol is converted into a column of underbars, and becomes the following display data.

Aquino Ryokowa     Izno   Spider mite
   Kimalita  Listen to this data with synthesized speech using the speech synthesis function
Make corrections while doing so. For example, after "Akino" 』
It is better to remove and use "Akinoryo Kowa"
And between "Shimadani" and "Kimarimasita"
" ] Sounds more natural. Also, "Ryoko
The segment break after "Uwa" is a middle pose, that is, " ]
It is better to add one. The corrected voice synthesis data is
It looks like this:

<Speech synthesis data format> Akinoryo Kowa [, 00] Izno [/ 00]
Kimarimasita [. 08] <Display data format> Akinoryo Kowa        Izno   Simoda
Niki Marita  User refers to data displayed in <display data format>
To edit.

The key operation in the editing operation can be easily performed by using the shift key and the cursor key. When the target voice message is confirmed, the data for voice synthesis is saved, and the editing operation ends. An operation method at the time of editing the phrase connection degree symbol will be described. At the time of editing, after moving the cursor to an arbitrary position of the target phrase with a key, work is performed by the following operations.

(1) Underbar " Insert; SH
Press the "↑" key while holding down the "IFT" key. (2) Underbar " Delete; press the “↓” key while holding down the “SHIFT” key. FIG. 2 shows an example of a screen at the time of editing by this device, in which the upper half of the screen displays the original Japanese sentence input from the input terminal 1 of FIG. 1A and the lower half displays display data for speech synthesis for editing. . Since the original text is displayed in full-width and the display data is displayed in half-width, the correspondence between the upper and lower character positions is almost the same, and editing work such as reading error detection is facilitated.

In this voice message editing apparatus, in the voice synthesizing data editing processing, the editing data is V-RAM in order to speed up editing such as inputting, inserting and deleting characters.
It takes the form to use. That is, the contents normally displayed on the editing screen are the same as the contents of the buffer being input. In addition, a syntax check process is performed to determine whether or not the data for speech synthesis corrected in the speech synthesis process, the paging process, etc. has a correct format as the speech synthesis data, and only normal V-RAM data is stored in the buffer. Like that. As a result, a rudimentary editing error can be prevented.

[0016]

As described above, according to the present invention, in the voice message editing device for editing the data for speech synthesis, the inter-segment connection symbol indicating the degree of inter-segment connection is not displayed by an esoteric symbol. Since it is shown in a visually comprehensible expression, a user who does not have specialized knowledge of voice can easily create a desired voice message.

[Brief description of the drawings]

FIG. 1A is a block diagram showing a voice message editing apparatus according to the present invention, and FIG. 1B is a diagram showing the correspondence between inter-phrase connection symbols and inter-phrase connection symbol strings.

FIG. 2 is a diagram showing an example of a screen layout at the time of editing in the present invention.

────────────────────────────────────────────────── ─── Continuing on the front page (72) Kazuo Hakoda Nippon Telegraph and Telephone Corporation, 1-1-6 Uchisaiwai-cho, Chiyoda-ku, Tokyo (72) Creator Atsushi Yamazaki 41-3 Kubo-cho, Nishi-ku, Yokohama-shi, Kanagawa Japan Inside Technical Industry Co., Ltd. (72) Kunihiko Matsumoto 41-3 Kubo-cho, Nishi-ku, Yokohama-shi, Kanagawa Prefecture Japan Technical Industry Co., Ltd. (56) References JP-A-4-166899 (JP, A) JP-A-4- 51099 (JP, A) JP-A-2-281298 (JP, A) JP-A-63-311299 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) G10L 3 / 00-9 / 18 G06F 3/16

Claims (1)

(57) [Scope of request for utility model registration] 1. An input method for a Japanese text including a kanji character set by a tesquito analyzing means for reading a kanji character, an accent form of a phrase, and a degree of connection between phrases, and converting the data into speech data in a predetermined format. In the voice message editing device that converts the data into a voice message and edits the voice message by correcting the voice synthesis data, the degree of connection between successive phrases in the converted voice synthesis data is determined by Means for converting and displaying an array of a predetermined number of identical symbols corresponding to the degree of coupling; means for correcting and editing the data obtained by converting the display of the degree of coupling; Means for obtaining the voice message by inversely converting the data into the predetermined format, and a voice message editing apparatus.