JPH04120850A - Voice synthesizer - Google Patents

Abstract

PURPOSE:To efficiently receive voice service only concerning a desired part for a user by detecting the voice synthesization interrupting request of the user and returning a service menu to a top menu even when a rule concerning all the contents of an electronic mail is not completely synthesized. CONSTITUTION:Under the control of a main control part 4, a voice rule synthesizing part 1 synthesizes the voice of a character string from a document file 3, and this synthesized voice is transmitted through an NCU(Network Control Unit) part 2, telephone line 7 and network 5 to a telephone set 6 of the electronic mail user. In this case, even when the rule concerning all the contents of the electronic mail is not completely synthesized, the voice synthesization interrupting request of the user is detected and the service menu is returned to the top menu. Thus, the user can efficiently receive the voice service only concerning the desired part.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention (Field of Industrial Application) The present invention relates to a device in which a speech synthesis device that generates speech from arbitrary characters according to predetermined rules is applied to telephone services.

(Prior art) Speech rule synthesis analyzes an arbitrary input character string, obtains phonological information and prosody information, and based on predetermined rules,
It outputs synthesized speech corresponding to the input character string. For example, in a telephone service, the content of a user's email (text file) sent to a center via PC communication is converted into speech using speech rules synthesis.
It is conceivable to provide telephone services to users. With this service, users can listen to a desired email in synthesized voice by selecting a voice output email by inputting it via a button according to the guidance voice. However, once the content of the selected mail is selected, the user does not wait for PB input until all synthesized sounds have been heard. Therefore, there was a problem in that it was not possible to stop the audio output of the e-mail that was currently being read.

(Problem to be solved by the invention) The present invention has been made in consideration of these circumstances, and
The purpose of this is to listen to the e-mail contents that are being output in voice until the end (without listening to the end).
To provide a speech synthesizer for telephone services that can return the text to the top menu.

[Structure of the Invention] (Means for Solving the Problems) In order to achieve this object, the present invention provides a method for adding an accent to an input character string, synthesizing it into speech, and transmitting it via a telephone line. In the speech synthesis device, there is a means for detecting a predetermined detection time from a human-powered character string, a means for determining whether or not there is an interruption request from a telephone line when detected by the detection means, and a means for determining whether or not there is an interruption request from the telephone line. In this case, a means for interrupting the synthesis of the character strings being synthesized is provided, so that it is possible to respond to an interruption request from the telephone line during the synthesis of famous mountains.

(Function) According to the present invention, the speech synthesis device detects the user's speech synthesis interruption request even if the rule synthesis of all the contents of, for example, electric mail is not completed, and can be returned to the top menu, and compared to before,
Users can efficiently receive audio services for only the parts they want to hear. Additionally, the convenience of the device is improved, and the communication and advertising capabilities of the device can be greatly increased.

(Example) Hereinafter, one embodiment of the present invention will be described based on the drawings.

FIG. 1 is a block diagram showing the configuration of an apparatus according to an embodiment of the present invention. In FIG. 1, 1 is a speech rule synthesis unit, 2 is an N C
A U (Network Control Unit) section, 13 is a document file for e-mail, and 4 is a main control section. In FIG. 1, a voice rule synthesis unit 1 synthesizes a character string from a document file 3 into voice under the control of a main control unit 4, and this synthesized voice is sent to an NC8 unit 2, a telephone line 7
, and is transmitted to the e-mail user's telephone 6 via the network 5. Here, details of the speech rule synthesis section 1 will be explained using the block diagram of FIG. 2.

1] is a character string analysis unit that analyzes a character string, verifies the accent position with reference to the reading dictionary 12, and obtains a phonetic symbol string and prosody information. The phoneme symbol string is input to the phoneme parameter generation device 14, which generates a phoneme parameter string with reference to the phoneme no fill 15. On the other hand, the prosody information is given to the prosody parameter generation device 13, and a prosody parameter sequence is generated.

The synthesizer 16 generates and outputs a synthesized sound according to a predetermined synthesis rule in accordance with the phonetic parameter string and the prosody parameter string thus obtained.

Next, the detailed structure of the NC section 2 will be explained using the block diagram of FIG. NC8 part 2 is modular jack 27
is connected to a telephone line, and the synthesized sound output from the voice rule synthesis section 1 is inputted from the synthesized speech input section 21 and transmitted through the transformer 28 to the telephone line.

', ), 5 is a calling signal detection unit, which switches the telephone line to a connected state when there is a call from the other party. Further, 23 is a PB detection unit, which detects whether or not the other party presses "*" indicating an interruption request during voice synthesis, as will be described later. 22 is a 40 (IHz detection unit) that detects whether the line with the other party is connected or not, and whether the other party is busy. Note that 24 is used to make a call. The dial section 26 is a noise filter.

FIG. 4 is an example of an electrician's mail that is synthesized by voice using a telephone service. FIG. 5 shows an intermediate analysis result output when the e-mail shown in FIG. 4 is input to the character string analysis section 11.

"-" is an accent mark "," is a pause symbol, "/" is an accent phrase boundary symbol, "//" is a clause end symbol, and katakana represents a phoneme to be synthesized.

Next, a series of operations of the e-mail telephone service in terms of such a speech synthesizer will be explained based on the flowchart of FIG. 6.

First, when a call arrives, the NC8 section 2 detects it with the ringing signal detection section 25, closes the telephone line, and then closes the telephone line.
(STI).

When the telephone line is connected, a guidance voice for the telephone service is output and the service is started (ST2).

If necessary, the user is prompted to enter the member number and password using the push button (PB). Read out the number of emails addressed to the user who made the call, the name given to the email, and enter the email number you want to hear (Sr1
).

The mail selected by PB is synthesized according to voice rules and sent to the telephone (Sr1).

Here, if the other party has not hung up, the process returns to Sr1 and the telephone service continues.

During this series of operations, the user who made the call may wish to cancel the e-mail voice synthesis midway through Sr1 and return to Sr1. In the past, once voice synthesis of an email started, it could not be stopped midway, and the user had to either listen to the end and wait for the call to return to Sr1, or hang up and then resume the call.

In the present invention, a request for interrupting speech synthesis from a user is received as PBr.
*J is predetermined. Furthermore, the speech rule synthesis section 1 has a phrase end symbol detection section for an intermediate analysis result that is output when an e-mail to be converted into speech is input to the character string analysis section 11. Therefore, when the phrase end symbol detecting section detects a phrase end symbol during speech synthesis, this fact is transmitted to the main control section 4. The main control unit 4 sends a signal to the PB detection unit 23 and detects PBr*J for about 2 seconds, thereby detecting a request to interrupt speech synthesis. At this time, the PB detection unit 2
When PB F*J is detected in step 3, the main control unit 4 interrupts the current e-mail composition and returns to the e-mail selection step. On the other hand, if pBr*J is not detected, the next phrase is synthesized into speech. These operations will be explained in more detail using flowchart 1 in FIG.

First, select the e-mail to be synthesized using voice rules, and open it to read it (STII). Next, the character string of the voice-converted email is stored in a buffer until the clause end symbol "//" of the voice-converted email is detected (STI 2., ST
I3).

The character strings stored in the buffer in this way are sent to the speech rule synthesis section, where the speech is synthesized and sent out to the telephone line (S T 1
4).

Next, detect if it is the end of the email (5T15)
, if it is not the end, the PB detection unit 23 of the NCU unit 2 detects the PB of “*” (STI6). The detection time here may be, for example, about 2 seconds. If the PB marked with “*” is not detected within this detection time, 5T12
Return to and repeat the speech synthesis process. If a PB with an asterisk (*) is detected, it is determined that the user making the call is requesting to interrupt the e-mail audio output, the file that was open for read is closed, and the speech synthesis reading process ends. (S T 17).

In the case of the e-mail text in Figure 5, "Higashi.../
/J Voice synthesize the part (a), then PB "*"
is detected, and if PBr*J is not detected, the part (b) of "Baira..." is synthesized into speech and PBr*J is detected.
If detected, the process of 5T17 is performed without performing speech synthesis on the part (b).

Note that in the above embodiment, P of the user's voice synthesis interruption request is
Although B is shown as "*", it is also possible to use "#" or a number. Furthermore, although the phrase end symbol is "//", it can also be any other symbol. In addition, it is said that "*" is detected by detecting the phrase end symbol, but in an e-mail,
A method may also be considered in which a predetermined symbol to be detected is inserted, and when that symbol is detected, PBr*j is detected. Here, the PB detection time is set to 2 seconds, but it is also possible to make it longer than this to ensure that the user's voice synthesis interruption request is detected.

[Effects of the Invention] As explained above, according to the present invention, the speech synthesis device detects a user's speech synthesis interruption request even if the rule synthesis of all the contents of an e-mail, for example, is not completed.
You can return the service menu to the top menu,
Compared to the past, users can more efficiently receive audio services for only the parts they want to hear.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the speech synthesis device of the present invention, FIG. 2 is a block diagram showing details of the speech rule synthesis section shown in FIG. 1, and FIG. 3 is a block diagram showing the details of the speech rule synthesis section shown in FIG. A block diagram showing the details of the NCU unit, FIG. 4 is a diagram showing an example of an e-mail, and FIG. 5 is a diagram showing an analysis result of the e-mail shown in FIG.
FIG. 6 is a diagram showing the processing flow of the telephone service, and FIG. 7 is a flowchart explaining details of the interruption request. ]...Voice rule synthesis unit, 2...NCU unit, 3...
Document file, 4... Main control unit, 23... PB detection unit. Agent Patent Attorney Noriyuki Ken Yudo Yamashita Diagram

Claims (1)

[Claims] (1) In a speech synthesis device that adds an accent to an input character string, synthesizes it, and transmits it via a telephone line, means for detecting a predetermined detection time from the input character string, and upon detection by the detection means, A speech synthesis device characterized by having means for determining whether or not there is an interruption request from a telephone line, and means for interrupting the synthesis of the character string being synthesized when the judgment means determines that there is an interruption request. .