JPH0731670B2 - Information processing method - Google Patents

Description

TECHNICAL FIELD The present invention relates to an information processing method for outputting a voice corresponding to an input character string.

[Prior art]

With the recent advances in speech synthesis technology, it has been attempted to add a speech synthesizer to a Japanese word processor and convert the input character information into speech when outputting characters to the Japanese word processor. ing.

One of these forms outputs a voice corresponding to an input character each time one character is input, and the other form outputs a voice to an input character string to be converted at every Kana-Kanji conversion. However, in both cases, the voice generation speed is constant, so when characters are temporarily input at high speed or when it takes time to convert Kana-Kanji characters, the result is a character buffer. It was not possible to utter a large amount of accumulated input characters in a short time, and the voice was uttered much later than the input, which significantly impaired the effect of input confirmation by voice.

〔Purpose〕

The object of the present invention is made in view of the above point, and by changing the output speed of the voice to be output according to the voice output instruction, there is no delay in voice output, and a confirmation operation of the input character string by voice is performed. The present invention realizes an information processing method capable of easily and surely.

〔Example〕

 An embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of a Japanese word processor according to the present invention. Reference numeral 1 is a key input device, 2 is a central processing unit, and 2A is an operation explanation described later, for example, FIG.
This is a memory in which a control program for sequentially executing the procedure shown in FIG. 10 is stored. Reference numeral 3 is a first storage device for storing the input kana character string, 4 is a second storage device for storing the kana-kanji converted sentence information, 5 is a third storage device for storing dictionary data, and 6 is a central processing unit. 2 is an auxiliary storage device, 7 is a display device, 8 is a voice rule synthesizing device, 9 is a fourth storage device used by the voice rule synthesizing device 6, 10 is a fifth storage device for storing a monosyllabic parameter file, 11 Is a speaker for audio output.

The voice rule synthesizer 8 stores the monosyllabic parameter corresponding to the Kana character string transferred from the central processing unit 2 in the fifth storage device.
The data is sequentially taken out from 10, and combined using a rule using the fourth storage device 9, converted into a sound signal having naturalness, and outputted from the speaker 11 as sound. At this time, if the speed information of the voice output is transferred from the central processing unit 2, the speed of the voice output is reset and thereafter the voice output is performed at that speed.

FIG. 2 shows a part of the keyboard of the key input device 1. In addition to various keys for a normal Japanese word processor, a "voice confirmation" key and a "variable speed" key for input confirmation function. Input confirmation by voice is performed only when the "Voice confirmation" key is set, and when the "Variable speed" key is set, the voice output speed is adaptively changed. Let

FIG. 3 shows another part of the key input device. In the figure, "conversion", punctuation ",", and punctuation "." When any of the keys is pressed, Kana-Kanji conversion is performed.

The operation of the embodiment configured as described above will be described with reference to FIG.

The key information input from the key input device 1 is converted into kana characters by the central processing unit 2 according to the control procedure and sequentially stored in the first storage device 3. The input is the above-mentioned "conversion", ",", Reading point ". , Any of the keys, the central processing unit 2 retrieves a kana character string from the first storage device 3 and performs kana-kanji conversion processing while referring to the dictionary data stored in the third storage device 5, The processing result is stored in the second storage device 4. At this time, the central processing unit 2 checks in step S1 whether or not the "voice confirmation" key on the key input device 1 is set.
The kana character string fetched from the storage device is transferred to the voice rule synthesizing device 8 and converted into voice before the kana-kanji conversion. Next, if the "variable speed" key is set in step S2, the central processing unit 2 is pressed in step S3.
Calculates the length of the kana character string stored in the first storage device 3, converts it into voice output speed information, and then, in step S4
Then, the kana character string is transferred to the voice rule synthesizing device 8 before being transferred, the voice output speed is changed in step S5, and then the kana character string is transferred and converted into voice. The voice rule synthesizer 8 sequentially extracts the monosyllabic parameters corresponding to the transferred Kana character string from the fifth storage device 10, combines them using the fourth storage device 9 according to the rule, and outputs the voice of the set speed. The signal is converted into a signal and the sound is output from the speaker 11.

Many speech rule synthesizers prepare unit speech used in rule synthesis in the form of acoustic parameters, but a set of acoustic parameters is used to synthesize a fixed number of speech samples, resulting in a constant rate of speech. A synthetic voice is output. Therefore, in order to change the speed of voice output, a method of increasing or decreasing the number of voice samples to be synthesized with a set of acoustic parameters is often used. That is, if the number of voice samples is increased, the duration of the voice expressed by a set of acoustic parameters becomes longer, so that the voice output becomes slower, and if the number of voice samples is decreased, the voice output becomes faster. The voice rule synthesizing device 8 constituting the present embodiment has a function of changing the voice output speed by the above method, and accepts voice output speed information as input information in addition to the kana character string. As shown in FIG. 4, the voice output speed information is 2 or 3 indicating the number of voice samples.
Expressed by adding an escape code before and after each number. FIG. 4 shows that there are 75 voice samples per set of acoustic parameters.

The procedure for determining the voice output speed will be further described.
First, the central processing unit 2 determines the length of the kana character string stored in the first storage device 3. The first storage device 3 is a buffer on a ring of size N as shown in FIG.
The next kana character is stored in the address pointed to by, and the kana character in the buffer is extracted from the address pointed to by pointer P2. Therefore, the number of characters X in the buffer is calculated as follows. However, 0 ≦ X ≦ N−1.

X = P1−P2 (P1 ≧ P2) X = P1−P2 + N (P1 <P2) On the other hand, the voice rule synthesizing device 7 has the upper limit S1 and the lower limit S2 of the number of voice samples. When the number of characters X takes the maximum value N-1, it means the maximum speed.
When S2 is transferred and the minimum value is 1, S1 which means low speed again may be transferred. Therefore, the central processing unit 2 determines the number of voice samples y by the following equation and transfers it to the voice rule synthesizer 8 as speed information.

Next, the central processing unit 2 transfers the Kana character string which is the target of Kana-Kanji conversion to the voice rule synthesizing unit 8 to convert it into voice, and finally, in Step S6, performs the Kana-Kanji conversion on the Kana character string. A series of processing ends.

FIG. 7A shows how variables are arranged in the auxiliary storage device 6 used by the central processing unit 2 during the Kana-Kanji conversion process.

Shown in Fig. 8 (b) to create the text shown in Fig. 8 (a), "Tonight in the Tokyo region, the wind in the west will be cloudy and sometimes cloudy, and in the mountains there will be some sunsets." Taking the case of input as an example, the contents of the first storage device 3, the number of characters x, the number of voice samples y, the transfer data to the voice rule synthesizing device 8 and the standard speed in the above-mentioned kana-kanji conversion processing are 10
The relative speed of the output sound when 0 is set and the output sound from the speaker 11 are as shown in FIG. 9, and the output speed of the sound is adaptively controlled by the number of characters in the first storage device 3. Is shown. However, N = 16, S1 = 100, and S2 = 30, and the speed of the voice output with the number of voice samples 85 is the standard speed. The symbol / indicates a "conversion" key.

According to another embodiment of the present invention, a storage area for storing the number of characters in the input buffer is provided on the auxiliary storage device, and the number of characters in the buffer x and the number of previous characters z stored in the storage area are set in the Kana-Kanji conversion process. The voice sample speed y is calculated only when the difference exceeds a certain threshold value, and is transferred to the voice rule synthesizer as voice output speed information, so that the voice output speed when the number of characters in the buffer changes abruptly. Is a change.

The configuration of this embodiment is shown in FIG.
As shown in FIG. 7 (b), a storage area is newly prepared on the auxiliary storage device 6 used by the computer 1 for the number of preceding characters z in the buffer and the threshold value D.

An embodiment of the above configuration will be described in detail below with reference to FIG.

The key information input from the key input device 1 is converted into kana characters by the central processing unit 2 according to the control procedure and sequentially stored in the first storage device 3. The input is the above-mentioned "conversion", ",", Reading point ". , Any of the keys, the central processing unit 2 retrieves a kana character string from the first storage device 3 and performs kana-kanji conversion processing while referring to the dictionary data stored in the third storage device 5, The processing result is stored in the second storage device 4.

At this time, the central processing unit 2 checks whether or not the "voice confirmation" key on the key input device 1 is set as shown in the first step of FIG. 10, and if it is set, it is taken out from the first storage device. Prior to the Kana-Kanji conversion, the Kana character string is transferred to the voice rule synthesizing device 8 to check whether or not it is converted into voice. Next, if the "variable speed" key is set in step S11, the central processing unit 2 first determines the number x of kana character strings stored in the first storage device 3 before transferring the kana character string. Determined in step S12.

X = P1−P2 (P1 ≧ P2) X = P1−P2 + N (P1 <P2) However, 0 ≦ x ≦ N−1.

Next, in step S13, the central processing unit 2 obtains the absolute value of the difference between the previous character number z stored in the auxiliary storage device 6 and the above character number x, compares this with a threshold value D, and the absolute value of the difference is the threshold value D. At the above time, x is stored in z in step S15, the number y of voice samples is calculated in step S16, and transferred to the voice rule synthesizer 8 as speed information to change the speed of voice output. The number of voice samples y is calculated by the following equation.

If the absolute value of the difference is less than the threshold value D, the central processing unit 2 does not transfer the speed information and the speed of voice output does not change.

After that, the central processing unit 2 transfers the kana character string to be converted into kana-kanji character to the voice rule synthesizing device 8 in step S17 and converts it into voice. Next, the kana character string is transferred and converted into voice. The voice rule synthesizer 8 sequentially extracts the monosyllabic parameters corresponding to the transferred Kana character string from the second storage device 10, combines them using the rule using the fourth storage device 9, and outputs the voice of the set speed. The signal is converted into a signal and the sound is output from the speaker 11. Next, Kana-Kanji conversion is performed.

Shown in Fig. 8 (b) to create the text shown in Fig. 8 (a), "Tonight in the Tokyo region, the wind in the west will be cloudy and sometimes cloudy, and in the mountains there will be some sunsets." Taking the case of input as an example, the first in each Kana-Kanji conversion process
Contents of storage device 3, number of previous characters z, number of characters x, number of voice samples y, transfer data to voice rule synthesizer 8, relative speed of output voice when standard speed is 100, and speaker 1
The output sound from 1 is as shown in FIG. 11, and the first storage device 3
It is shown that the output speed of the voice is adaptively controlled only when the number of characters in fluctuates rapidly. However, N = 1
6, S1 = 100, S2 = 30, D = 2, and the speed of the voice output with the number of voice samples 85 is the standard speed. The symbol / indicates a "conversion" key.

[Effects of the Invention] As described above, according to the present invention, by changing the output speed of the voice to be output according to the voice output instruction, there is no delay in voice output, and a confirmation operation of the input character string by voice is performed. Can be done easily and reliably.

[Brief description of drawings]

FIG. 1 is a block diagram showing the configuration of the present invention, FIG. 2 is a partial conceptual diagram of a keyboard showing an example of arrangement of a “voice confirmation” key and a “variable speed” key, and FIG. 3 is a key related to kana-kanji conversion. FIG. 4 is a partial conceptual diagram of a keyboard, FIG. 4 is a diagram showing formation of output speed information of rule-synthesized speech, FIG. 5 is a conceptual diagram of input buffer, and FIG. 6 is a flow chart of Kana-Kanji conversion according to the present invention. FIG. 7 is a block diagram showing the inside of the auxiliary storage device, FIG. 8 is a diagram showing an example sentence, FIG. 9 is a diagram showing a processing result when an example sentence is input, and FIG. 10 is another example. FIG. 11 shows a flow chart, and FIG. 11 is a diagram showing a processing result of another embodiment when an example sentence is input. 1 ... Key input device 2 ... Central processing unit 3 ... First storage device 4 ... Second storage device 5 ... Third storage device 6 ... Auxiliary storage device 7 ... Display device 8 ... Voice rule synthesis Device 9 ... Fourth storage device 10 ... Fifth storage device 11 ... Speaker

Claims (1)

[Claims] 1. An information processing method for storing an input character string and outputting a voice corresponding to the stored character string according to a voice output instruction of the stored character string, wherein the stored character An information processing method, comprising: determining a length of a string, and changing an output speed of a voice corresponding to the character string according to the determined length.