JPH0641195Y2 - Electronic device equipped with voice recognition means - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention relates to an electronic device including a voice recognition means for recognizing input voice data and performing a predetermined operation.

[Prior Art] In a conventional voice recognition device incorporated in an electronic device or the like, when a voice is input, the voice data is recognized for a predetermined time, and after the recognition is completed, a predetermined operation is started according to the recognition result. It is like this.

[Problems of conventional technology]

When the above-mentioned conventional voice recognition device is incorporated in, for example, a stopwatch device and an attempt is made to start (end) measurement by voice input of "start"("stop"), the following problems occur. That is, when there is a voice input “start” (“stop”), the operation is not started at the same time as the input, but the operation is started after the recognition time as described above has passed. There is a gap between the time and the operation start time, and thus an accurate measurement time cannot be obtained. Therefore, a stopwatch device by voice input has not been realized.

It should be noted that this kind of problem is not limited to the stopwatch device, and occurs in various devices in which data controlled according to voice input fluctuates with time.

[Purpose of device]

The present invention has been made in view of the above conventional problems, and an object of the present invention is to provide an electronic device equipped with a voice recognition means capable of performing an accurate operation by eliminating a time lag associated with voice recognition.

[Key points of device]

In order to achieve the above object, the present invention counts the time from voice input to the end of voice recognition, and uses the count value to correct the time gap between the voice input and the end of voice recognition. The point is to do so.

〔Example〕

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing an internal circuit of a stopwatch device to which an embodiment of the present invention is applied.

In the figure, ROM1 is a fixed memory containing a microprogram and data for controlling the entire system. The RAM 2 is a memory that can read and write various data, and is composed of various storage areas as shown in FIG. Here, the display register X, the stopwatch register Y, and the timekeeping register Z respectively store a register for storing data to be displayed on the display unit 11 described later, a register for storing measurement data during stopwatch operation, and a current time. A register that stores the time data that is represented, and also a plurality of lap storage memories
M ₀ , M ₁ , ... Are memories that sequentially store lap data during stopwatch operation. Furthermore, register C is
As will be described in detail later, this is a register that starts counting when voice is input and stops counting when voice recognition is completed.

Further, in FIG. 1, the voice recognition unit 3 is a circuit for recognizing voice data input from the microphone 4, and the voice detection unit 5 is a circuit for detecting that a voice input from the microphone 4 is above a certain volume. is there. Further, the oscillator 6 is a circuit that outputs a clock signal of a constant cycle, and the frequency dividing circuit 7 is a circuit that divides the clock signal and outputs a clock signal of a constant cycle (for example, 1/32 seconds).

Based on the program in the ROM 1, the control unit 8 performs voice processing according to the outputs of the voice recognition unit 3 and the voice detection unit 5, key input processing according to the key input of the key input unit 9, and the frequency dividing circuit 7. It is a circuit that performs a timing process or the like according to the timing signal of. The time data, stopwatch data, and the like obtained by these processes are displayed on the display unit 11 by the decoder driver 10.

FIG. 3 is a general flow showing the overall processing by the control unit 8.

In the figure, in step a ₁ ~a _3, respectively shown in FIG. 1, clocking the carry signal from the frequency divider circuit 7, a key input unit 9
It is determined whether or not there is a key input from, or a voice input from the microphone 4 (voice detection by the voice detection unit 5). When it is determined in step a ₁ that there is a time-carry signal, time-measurement processing is performed in step a ₄ . This time counting process is a process of counting the current time data, such as hours, minutes, seconds, and stopwatch data, based on the above-mentioned timekeeping carry signal. Also, if it is determined that there is a key input in step a _2, performs the key input processing in step a _5. This key input process is a process corresponding to various keys operated in the key input unit 9 (for example, a mode switching key and a time adjustment key). Also, if it is determined that there is the voice input in step a _3, performs audio processing in Step a _6. This voice processing is specifically shown in FIG.

In FIG. 4, first, in step b ₁ , it is judged whether the data is voice data, and if it is voice data, the process proceeds to step b ₂ .
The voice recognition by the voice recognition unit 3 shown in FIG. 1 is started. At the same time, in step b _3, the counter C shown in FIG. 2
Start counting by. This counting is continued until it is determined in step b ₄ that the voice recognition by the voice recognition unit 3 is completed. That is, the counter C counts the time from the voice input to the completion of the voice recognition.

When the voice recognition is completed, whether the recognition result is "start" that instructs the start of the stopwatch operation or "lap" that instructs the writing of the lap time that is the intermediate elapsed time, or the result of the stopwatch operation Select whether to use "stop" to stop or "clear" to clear measurement data.
b _5, it is determined in _{b 8, b 10, b 13} . If it is determined that the "start" at step b ₅ proceeds to Step b _6, after the addition of the contents of the counter C stopwatch register Y shown in FIG. 2, the start process in step b ₇ (i.e. register Y The process of measuring the time after the start is started by sequentially adding 1 to 1. In this case, the time of actual start to the start time of the start process is delayed by the time required for speech recognition than (i.e. when the sound is input as "start") occurs, the time delay is above Step b ₆ The content of the stopwatch register Y becomes an accurate measurement time from the actual start because it is corrected by the process of. For example, even if it takes 2 seconds for voice recognition, the delay of 2 seconds is added to the stopwatch register Y, so that the content shows an accurate measurement time.

On the other hand, the process proceeds to step b ₉ if the determination is "wrapped" in Step b _8, wrap the value obtained by subtracting the contents of the counter C from the register for stopwatch Y (i.e. the time measured by the time it is speech input) storage memory M _k (k = 0, 1, 2
...) Also in this case, if the contents of the stopwatch register Y at the time of voice recognition completion are used as the wrap time as they are, this will include an extra time corresponding to the time required for voice recognition. to subtract a such time in the process at step b _9, accurate lap time is obtained.

Further, the process proceeds to step b ₁₁ If it is determined as "stop" in step b _10, after subtraction of the contents of the counter C from the register for stopwatch Y, stop the process at step b ₁₂ (counting by register for stopwatch Y Process to stop). Also in this case, if the content of the stopwatch register Y at the time of completion of voice recognition is used as the final elapsed time as it is, this includes an extra time required for voice recognition as in the case of the above lap. but will, to subtract the extra time in the process at step b _11, the exact elapsed time can be obtained.

Also, if it is determined to be "clear" in step b ₁₃ is
Performing clearing process for clearing the contents of the stopwatch register Y in step b _14. In this case, since the time lag does not cause any particular problem, the correction by the counter C is not necessary.

According to this embodiment, the time from the voice input to the end of recognition is counted by the counter C as described above (step
b ₃ ), using this count value, the time lag between voice input and voice recognition end is corrected (steps b ₆ , b ₉ , b ₁₁ ).
As a result, a very accurate measurement time can be obtained. Therefore, a stopwatch device by voice input is realized.

Next, a data bank device to which another embodiment of the present invention is applied will be described. The overall structure of the internal circuit is the same as that of FIG. 1, and the microprogram of ROM1 and RAM2
Only the content of is different.

The contents of RAM 2 according to this embodiment are shown in FIG. In the figure, a plurality of data memories A ₀ , A ₁ , ...
It is a memory that stores various data such as telephone numbers and schedules. The page pointer P is a pointer for designating any address (page) of the data memory,
The in-search flag F is a flag in which "1" is written during data search and "0" is written when the search is completed. Further, the counter C is a counter that counts a 32 Hz signal and a signal of {circle around (1)} that is deviated from this by a half cycle, and the counter N is a counter that counts a signal every 0.5 seconds.

Next, the voice processing at the time of data search according to this embodiment will be described with reference to FIG.

First, did the 32Hz signal come in at step c ₁ ?
▼ Determine if a signal comes in, and if it is a 32 Hz signal, move to step c ₂ . In this step c ₂ , it is judged whether there is voice input, and if there is voice input, step
At c ₃ , the counter C is incremented by 1 (1 corresponds to 1/32 second). Then, in step c ₄ , the content of the counter C is 0.
Determine if it is 5 seconds. Here, for 0.5 seconds, each data in the data memory A ₀ , A ₁ , ... Is sequentially switched and displayed as shown in FIG. 7 (step c described later).
_{10 to} c ₁₃ )). If the content of the counter C is 0.5 seconds, the process proceeds to step c ₅ , where 1 is added to the other counter N and the content of the counter C is set to zero. The counting by the counter N is continued until it is judged in step c ₆ that the voice recognition is completed. That is, the counter N counts the number of display switchings from the time of voice input to the completion of voice recognition.

If the speech recognition has completed, the recognition results or an instruction to start data retrieval "start" of the, or to instruct the stopping of the data retrieval or "stop" of the respective determined in step c _7, c ₁₄ To do. If it is determined that the "start" in step c ₇ proceeds to step c _8, counter N
After the contents of C and C are set to zero, 1 is written in the searching flag F in step c ₉ . Start like this (start search)
When voice input is performed, the time gap between the voice input and the voice recognition completion does not cause any particular problem, and therefore the process for correcting the time gap is not necessary. While searching for data, until next voice input, ie step
While it is determined that F is 1 in c ₁₀ , the signal is counted by the counter C in step c ₁₁ .
When the content of the counter C in step c ₁₂ is determined to be 0.5 seconds, the process of displaying changed in step c _13. In this process, 1 is added to the page pointer P and the content of the counter C is set to zero. Therefore, during the search, the content of the page pointer P is incremented by 1 every time the counter C counts 0.5 seconds, and the contents of the data memories A ₀ , A ₁ , ... Are sequentially added as shown in FIG. It is switched and displayed.

On the other hand, if it is determined that the "stop" in step c ₁₄ if there is speech input, the flow advances to step c _15, see if the search flag F is 1 (i.e. whether Searching for data). F = If 1 proceeds to step c _16, wherein after subtracting the contents of the counter N from page pointer P, the contents of the counter N and C to zero, followed by step c ₁₇
Enter zero in the searching flag F. In this case, if the contents of the page designated by the page pointer P at the time of completion of the voice recognition are displayed as they are, the previous data for the time required for the voice recognition will be displayed here. Then, since the display data is corrected by subtracting the content of the counter C from the page pointer P, the page at the time when the "stop" voice is input is finally displayed. For example, as shown in FIG. 7, if "stop" is issued when the contents of the data memory A ₂ are displayed, and if it takes 1 second for the voice recognition, the data memory A ₄ which is two pages ahead will be displayed. After the contents are switched and displayed, the display is finally returned to the position two pages before that. Therefore, data that is not time-shifted with respect to the voice input is displayed.

According to the present embodiment, as described above, the counter N counts the number of display switchings from the time of voice input to the end of recognition (step c ₅ ), and this count value is used to determine the time of voice input. correcting the time lag of display data at the time of speech recognition terminal since the (step c ₁₆₎ above, it is possible to search at an accurate timing that matches the time of voice input. Therefore, a data bank device capable of searching by voice input is realized.

The present invention can be applied not only to the stopwatch device and the data bank device described above, but also to various devices in which data controlled according to voice input varies temporally.

[Effect of device]

As described above, according to the present invention, it is possible to accurately perform a predetermined operation such as a stopwatch operation and a data search operation by voice input without eliminating the time lag caused by voice recognition.

[Brief description of drawings]

FIG. 1 is a block diagram showing the internal circuit of a stopwatch device to which an embodiment of the present invention is applied, FIG. 2 is a diagram showing the contents of RAM 2, and FIG. 3 is a flow chart showing the overall processing of the above embodiment. 4, FIG. 4 is a flow chart specifically showing the voice processing in FIG. 3, FIG. 5 is a view showing the contents of RAM in a data bank device to which another embodiment of the present invention is applied, and FIG. 6 is the same. FIG. 7 is a flow chart specifically showing the voice processing of the embodiment, and FIG. 7 is a diagram showing switching of display accompanying voice input at the time of data search of the embodiment. 1 ... ROM, 2 ... RAM, 3 ... voice recognition section, 4 ... microphone, 5 ... voice detection section, C, N ... counter.

Claims (1)

[Scope of utility model registration request] 1. A voice recognition means for recognizing inputted voice data, a control means for starting a predetermined operation based on a recognition result of the voice recognition means, and a time from the voice input to the end of the recognition. An electronic device comprising a voice recognition means, comprising: a counting means and a correcting means for correcting data obtained by the predetermined operation based on a count value of the counting means.