JPS595919B2 - Method for detecting abnormality while voice response device is online - Google Patents

Abstract

PURPOSE:Online inspection of the mixer and trouble detection by previous setting of the voice data in the buffer.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Application of the Invention] The present invention relates to an abnormality detection method for a voice response device that makes it possible to detect the quality of a voice synthesizer while online.

[Background of the invention]

With the development of the information society, voice response devices have also progressed from analog type to digital edited recording type, and even larger capacity synthesized edited type voice response devices are being put into practical use.

Various types of synthesized and edited voice response devices have been invented, but among them, the PARCOR type voice response device has the best sound quality and has reached the stage of practical use. A PARCOR type voice response device records a compressed digital signal on, for example, a rotating recording medium, restores the original digital signal using a synthesizer, and then outputs audio using conventional digital technology. Therefore, the quality of the synthesizer influences the quality of the voice response device. However, all of the currently used methods of checking the synthesizer are performed with the line in the OFF state, and it has been impossible to check with the line in the ON state. [Object of the Invention] An object of the present invention is to eliminate the above-mentioned drawbacks of the prior art and to provide a method for detecting an abnormality in a voice response device, which can check the synthesizer of the voice response device while online.

[Summary of the invention]

In the present invention, audio information such as the most frequently used specific audio parameters, sound source synthesis data, and several samples of digital signals of the synthesis results are set in advance in individual buffers, and the specified audio information is stored online. When voice synthesis work is performed using this information, the preset specific information and synthesis information are compared to check the synthesizer and detect abnormalities in each part.

[Embodiments of the invention]

An embodiment of the present invention is shown in the accompanying drawings and will be described in detail below.

The attached figure is a block diagram of a PARCOR type voice response device which is an embodiment of the present invention, in which 1 is a magnetic drum device, 2 is a
is a word buffer, 3 is a word selector, 4 is a sound source synthesizer,
5 is a K data buffer, 6 is a synthesizer, 7 is a demodulator, 8 is a TKE switch, 9 is a digital buffer, 10 is a parameter buffer, 11 is a sound source buffer, 12 is a parameter comparator, 13 is a digital comparator, 14 is a talkie device,
15 is an audio output terminal, 16 is a digital abnormality output terminal,
Reference numeral 17 indicates a parameter abnormality output terminal, and 18 indicates a sound source synthesis abnormality output terminal.

Next, the operation of the PARCOR type voice response device constructed as shown in the attached diagram will be explained.

First, since the voice used in the voice response device changes on a business-by-job basis, when creating the voice, the most frequently used voice parameters are placed in the parameter buffer 10, the sound source synthesis data is placed in the sound source buffer 11, and the synthesis result is stored in the parameter buffer 10. Several samples of digital signals are set in the digital buffer 9 in advance. The parameters read from the magnetic drum device 1 are stored in the word buffer 2.

The parameters stored in the word buffer 2 are output to the sound source synthesizer 4 and the K data buffer 5 according to instructions from the word selector 3. If the selected parameter at this time is a check parameter set in advance, the digital comparator 13 is driven, and the sound source parameters from the word buffer 2 are used to synthesize the sound source in the sound source synthesizer 4, and the sound source is synthesized in the sound source synthesizer 4. Output. - The synthesizer 6 restores the digital signal using the data from the K data buffer 5 and the data from the sound source synthesizer 4, sends the digital signal to the demodulator 7, converts it into audio, and then
Send audio to the audio output terminal 15. When the digital signal from the synthesizer 6 is output to the demodulator 7, if the digital comparator 13 is driven, this digital signal is compared with the contents of the digital buffer 9, and if an abnormality is detected. The TKE switch 8 is operated to output a talkie voice to the user and a drive signal to the parameter comparator 12.

The driven parameter comparator 12 first compares the parameter buffer 10 and the sound source parameters of the sound source synthesizer 4 with K
A comparison is made with the K parameter of the data buffer 5, and if an abnormality is detected, a signal is sent to the parameter abnormality output terminal 17. Next, the sound source data output from the sound source synthesizer 4 is compared with the sound source buffer 11, and if an abnormality is detected, a signal is sent to the sound source synthesis abnormality output terminal 18. [Effect of the invention] As described above, by using the method of the present invention, the synthesizer 6
can be checked while online, and abnormalities in each part can be displayed.

As described above, the present invention has the advantage that it is possible to easily check the synthesizer, which can be said to be the heart of a speech synthesis type voice response device, while on-line.

[Brief explanation of the drawing]

The attached figure is a block diagram showing an embodiment of the abnormality detection method for a voice response device according to the present invention.

Claims (1)

[Claims] 1. Set frequently used specific information in a buffer in advance, and compare the preset specific information and synthesis information when the specific information is used to perform voice synthesis work online; 1. A method for detecting an abnormality while a voice response device is online, characterized in that it is configured to check a synthesizer.