JPS60246160A - Automatic response device of voice analysis and synthesis system - Google Patents

Abstract

PURPOSE:To store a lot of voice data expensively and to improve the utilizing efficiency of a memory section by storing a response message at a head of a tape in common use for response/business message recording to the memory section via a voice analysis section when a detecting means detects that the voice analysis data is destroyed or relinquished. CONSTITUTION:A response message and a business recording pattern are repeated as a response message recording part 12 and a business recording part on the tape in common use for response/business message recording. When a voice analysis data stored in the memory section 20 is destroyed or annihilated, the response message recording part 8 at the head of the tape in common use for response/business recording is reproduced and stored again in the memory section 20 via the voice analysis section 2. Thus, after the end of reception, that is, until the next reception after the end of the business message recording of a caller, it is possible to use freely the memory section 20 as an open state. Concretely, the data is transmitted/received between the memory section 20 and an external device through the intervention of a digital information input/output section 21.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Application) The present invention relates to an automatic response device using a voice analysis and synthesis method, and more particularly, to an automatic response device using a voice analysis and synthesis method. This invention relates to an automated response system.

(Background) FIG. 1 is a block diagram showing an example of a conventional voice analysis and synthesis type automatic response device.

In the figure, 1 is a voice input unit, 2 is a voice analysis unit, 3 is a memory unit for storing voice analysis data using CMO8RAM, 4 is a voice synthesis unit, 5 is a voice output unit, and 6 is for recording business matters. It's a tape.

As is known, the automatic response device of the voice analysis and synthesis method shown in figure W41 connected to a telephone line and a telephone first supplies a response message to the voice analysis section 2 through the voice input 81% as a preparatory step. The voice analysis data obtained as a result of the voice analysis is stored in the memory section 3. and,
Thereafter, when the other party makes a call to the apparatus, the voice analysis data is read out from the memory section 3, synthesized by the speech synthesis section 4, and sent to the other party via the voice output section 5.

After that, the other party (caller) can send the following information as necessary.

At this time, the business is sent via the telephone line.

The business is recorded on the business recording tape 6 via the voice input section 18.

By the way, in such a conventional automatic response device, an expensive CMO8RAM is used in the memory unit 3 for storing voice analysis data.
is used, and battery backup is performed to retain the data. Therefore, a large amount of speech analysis data (
If you try to store audio data, it will be very expensive.

Another disadvantage is that the voice analysis data is destroyed when the power is turned off, such as when replacing the battery. That is, when the variable power supply is turned off, the KAS user has to utter a new message, perform voice analysis, and store the voice analysis data again, which is inconvenient.

Furthermore, if the user wishes to use several types of response messages as appropriate, he or she must utter the message each time, perform voice analysis, and store it in the memory unit 3.
Furthermore, there is also the drawback that the memory section is always occupied for storing voice analysis data, resulting in very poor usage efficiency of the memory section.

On the other hand, unlike the conventional automatic answering device (answer recorder) using the voice analysis and synthesis method described above, this device records the response message and message on a single tape (a tape for both response and message recording) and plays it back. There is.

In other words, the response message is recorded at the beginning of a tape used for both response and message recording, the message is rewound to the beginning of the tape after recording, and when receiving, the response message is first sent out at the beginning of the tape, and then the response message is recorded as needed. When recording a message, the method is to fast-forward through the previous message recording section and then record the message.

However, with this method, the caller who made the call while the tape was being fast-forwarded was forced to remain on the line, and the message was recorded near the end of the tape that was used both for answering and recording the message. Those who do so will have to wait longer. For this reason, it had a fatal flaw as an absentee recorder.

(Object of the Invention) The object of the present invention is to eliminate the drawbacks of the prior art described above, to be able to store a large amount of audio data at a low cost, and to easily provide audio data without the need for constant battery backup to retain audio data. In addition to being able to analyze and store the analyzed data, if you want to use several types of response messages as appropriate, you can utter them each time, analyze them, and then store them in the memory. An object of the present invention is to provide an automatic response device using a voice analysis and synthesis method that can store voice analysis data and also improve the usage efficiency of the memory section since it is not necessary to occupy the memory section with voice analysis data all the time.

(Summary of the Invention! The l!!jgL of the present invention includes a voice analysis unit that analyzes voice signals and creates voice analysis data, a memory unit that stores the voice analysis data, and a memory unit that is read from the memory unit. In an automatic response device using a voice analysis and synthesis method, which consists of a voice synthesis section that synthesizes voice analysis data and sends it out, a response message is recorded at the beginning of the response message, and is sent via a telephone line as necessary at the rear of the response message. a tape mechanism for driving and recording/playing a response/message recording tape that alternately and sequentially records messages from a caller and response messages read from the memory unit via the voice synthesis unit; and a detection means for detecting whether or not the voice analysis data in the memory section is destroyed or erased when reading from the memory section, and the voice analysis data is destroyed or erased by the detection means. When this is detected, the response message at the beginning of the response/message recording tape is stored in the memory section via the voice analysis section.

(Example) Hereinafter, the present invention will be explained using the drawings.

FIG. 2 is a block diagram showing one embodiment of the present invention.

In this figure, the same reference numerals as in FIG. 1 indicate the same or equivalent parts. 20 is a memory unit for storing voice analysis data using dynamic RAM (DRAM); 21 is a digital information input/output unit for transmitting and receiving data between the memory unit 20 and external equipment; and 22 is a response/message recording unit. A tape mechanism has a dual-purpose tape, drives the core tape, and performs recording/reproduction, and 31 and 32 are an audio amplification unit and a low-pass filter that constitute the audio input unit 1.

40 is a controller section, 41 is a recording button, 42 is an answering button, and 51 to 53 are an impedance conversion circuit, a low-pass filter, and an audio amplification section that constitute the audio output section 5.

FIG. 3 is a partial recording pattern diagram of the response/message recording tape, showing an example of the recording pattern of the response/message recording tape driven, recorded and reproduced by the tape mechanism 22 of FIG.

In the same figure, 8 is a response/message recording tape (hereinafter referred to as
(sometimes simply called tape) is the beginning of the recorded response message. 9 and 11 are the recording part of the business, 10.1
2 is a response message recording part.

1 and 2 in FIG. 4 refer to the controller section 4 in FIG.
2 is a flowchart for explaining an example of the operation of 0.

The operation of this embodiment will be explained below based on this flowchart.

First, a user connects the device of this embodiment to a telephone line and a telephone set, as is known.

In step S1, when the device is powered on,
In step 82, the speech analysis section 2 and the speech synthesis section 4
A reset signal is output to each. As a result, the voice analysis section 2 and the voice synthesis section 4 are reset. After that, the process advances to step S3.

In step S3, it is determined whether the recording button 41 has been pressed and the response message recording mode is set. Then, if the recording mode is not trapped, wait until the mode is set, and if it is determined that the recording mode is set, step 8
Proceed to step 4.

In step S4, a voice analysis start signal is output to the voice analysis section 2, and a recording signal is output to the tape mechanism 22.

After that, when a response message is issued, this audio signal is amplified by the audio amplification section 31 of the audio input section 1 and is transmitted to the tape machine #I.
22, and is further supplied to the one-pass filter 32.
The audio signal whose high frequency has been cut is supplied to the audio analysis section 2.

The audio signal supplied to the audio analysis section 2 is subjected to audio analysis here, and is sequentially output as audio analysis data.

This output audio analysis data is stored in the memory section 20 at 0 by the write address signal output in step 85.
Stored sequentially starting from address. - Then, the audio signal supplied to the tape mechanism 22 will be heard at the beginning of the tape used for both response and message recording.

Then, in step $6, it is determined whether the response message recording start edge and the scheduled recording time, for example, 30 seconds, have elapsed. As a result, if the elapsed time has not elapsed, the audio recording state on the memory unit 20 and the tape described above is continued, and when the elapsed time has elapsed, the process advances to step S7. (At this time, the response message recording portion 8 at the beginning of the tape shown in FIG. 3 has been formed.

In step S7, a voice analysis cancellation signal is output to the voice analysis section 2, and a recording stop signal is output to the tape machine 1122. From this K, the voice recording state is canceled and the process advances to step S8.

In step S8, a cue signal is output to the tape mechanism 22. As a result, the response message recording portion 80 at the beginning of the tape is located.

From this point K, the preparation of the automatic response device of this embodiment is completed.

In step S9, it is determined whether +1, the answering button 42 has been pressed, and the apparatus has been set to answering mode.

If the answering mode is not selected, the process waits until it is selected, and when the answering mode is selected, the process proceeds to step 5IOK.

In step 810, the device receives a call and determines whether the receiving bell has rang three times.

That is, it is determined whether or not the bell sound (bell signal) has been counted a preset number of three times. Note that the bell signal is input from the terminal 60 in this embodiment.

If the ringing stops after less than three times, in this embodiment, it is determined that there is no reception and the standby state continues in the answering mode. On the other hand, if it is determined that the bell has rang three times, the process advances to step 811.

In step 811, a response message reproduction signal is output to the tape mechanism 22. As a result, the response message recording portion 8 at the beginning of the tape is played back, and the closing message is sent to the other party (caller) from the audio output s5.

In step S12, after reproducing the entire response message recording portion 8 at the beginning of the tape, a beep sound is emitted to the caller through the audio output unit 5. In other words, the timing to start recording the business is notified.

In step 813, a message recording signal is output to the tape mechanism 22. From this trap, the tape that can be used to record responses and business matters is
Driving begins in recording mode.

In step S14, it is determined whether the first caller has hung up the telephone.
That is, it is determined via the voice amplifying section 31 of the voice input section 1 whether or not the caller's handset is turned on.

If the handset is on, step 3
Proceed to step S17, and if it is not in the on state, proceed to step S1
Proceed to step 5.

In this step S15, after outputting the beep sound,
It is determined whether or not a message 1g has been received from the caller via the voice amplifying section 31 within the scheduled time. As a result, when it is determined that no business has been received within the scheduled time, the process proceeds to step 817K, and on the other hand, when it is determined that business is rarely received, the process proceeds to step S16.

In step 816, it is determined whether or not the predetermined recording time of the business, for example, 3 minutes, has elapsed in this embodiment. If the elapsed time has not elapsed, the process returns to step S14, but
When it is determined that the three minutes have elapsed, step 3 is performed.
Proceed to step 17.

In step 817, a message recording stop signal is output to the tape taIN22. As a result, the recording mode of the tape mechanism 22 is canceled, and the tape for both response and message recording stops running. Furthermore, if the other party has issued a message by this time, a message recording section 9 is formed on the tape following the response message recording section 8, as shown in FIG.

In step 318, as described above, it is determined whether the voice analysis data of the response message stored in the memory unit 20 in step 35 is stored without being destroyed or erased. As a result, if the voice analysis data is stored without disappearing, the process proceeds to step 823; if for some reason it disappears and is not stored,
Proceed to step S19.

In step 819, a tape cue signal is output to the tape mechanism 22. As a result, the response/message recording tape is rewound to the beginning and then stopped.

In step 320, a voice analysis start signal is output to the voice analysis section 2. Thereafter, the process proceeds to step 821.

At step 821, a response message reproduction signal is output to the tape mechanism 22, and a write address signal is output to the memory section 2. As a result, the response message recording portion 8 at the beginning of the chive is played back, the played audio signal is analyzed by the audio analysis section 2, and the resulting audio analysis data is stored again in the memory by the address signal. The data are sequentially stored in the section 2o starting from address 0. Thereafter, the process advances to step 522.

In step 822, a fast-forward signal to the recording end is output to the tape mechanism 22. As a result, the tape used for both response and message recording is as follows, up to the end where the message is recorded.
The program is fast-forwarded to the end of the business recording section 9, and the traveling is stopped there.

Next, the process proceeds to step 323.

In step 823, in response to the process in step S22 or the determination in step 818 that the voice analysis data is stored in the memory unit 20 without disappearing,
A speech synthesis start signal is output to the speech synthesis section 4. after that,
The process advances to step S24.

In step S24, a response message recording signal is output to the tape mechanism 22, and a read address signal is output to the memory section 20. As a result, the voice analysis data stored in the memory section 20 is sequentially read out in accordance with the read address signal and supplied to the voice synthesis section 4, where it is synthesized into a response message, and a response/message recording tape is used. It will be recorded in the part following the business recording part 9. As a result, a response message recording section 10 is formed. After that, the process advances to step S26.

In step S26, a response message cue signal is output to the tape mechanism 22. As a result, here, the response message recording part 1017) is output (f(), and then it enters a standby state in the answering mode and prepares for the next reception.

When there is a next reception, at this time, the response message bell tone part 10 is played back in the same way as the first reception described above.
A cough reproduction response message is sent to the other party via the audio output unit 5.

Thereafter, if necessary, the message is recorded following the response message recording section 10 as described above. That is,
It becomes a trap to form the business recording part 11.

Thereafter, as described above, the recording pattern of the response message and the message is repeated on the response/message recording tape, such as the response message recording portion 12 and the message recording portion.

As is clear from the above explanation, in this embodiment, when the voice analysis data stored in the memory unit 20 is destroyed or disappears, the response message recording portion 8 at the beginning of the response/message recording tape is played back and stored in the memory unit 20 again through the voice analysis unit 2.

For this reason, in this embodiment, it is possible to freely use the memory unit 20 as an open warrant after the reception is completed, that is, after the recording of the caller's message is completed until the next reception. be.

Specifically, the digital information person output unit 21fr: As a mediator, the digital information person output unit 21fr is busy in being able to send and receive data between the memory unit 20 and external devices.

Next, in this example, due to reasons such as a power outage,
A case in which the power is once turned off and the voice analysis data in the memory unit 20 is destroyed or deleted will be explained.

In this case, in the present embodiment, when the power outage or the like is canceled, the device remains in a standby state in the answering mode. That is, in FIG. 4, in step SIO, the judgments in steps 89 and 810 are repeated until the ringing is counted three times and it is judged that there is an incoming telephone call.

Then, in step 810, when it is determined that there is a telephone call, 1'! , processing and judgment will be performed in the same flow as described above. However, in this case,
Since the voice analysis data (response message) in the memory unit 20 has been destroyed, etc., it is always determined in step 818 that the voice analysis data does not exist. As a result, after the judgment,
The process proceeds to step S23 through the processes of steps 819 to S22.

That is, in this case, by playing back the response message recording portion 88 at the beginning of the tape, the voice analysis data of the response message is stored in the memory section 2o again, and then the response message recording section 88 is played back, and then the voice analysis data of the response message is stored in the memory section 2o again, and then the response message recording part 88 is played back, and then the voice analysis data of the response message is stored in the memory section 2o again. Fast forward to now.

In this way, in this embodiment, even if the voice analysis data in the memory unit 20 is destroyed, there is no need to re-utter and store the response message again, and it is possible to back it up using a tape that can be used for both response and message recording. I can do it.

Furthermore, as can be easily inferred from the above explanation, according to the embodiment, when a different response message is stored in the memory unit 20 as needed, the streetcar utters the response message and stores it in the memory unit. There is no need to store it in 20.

The reason for this is that in this case, if you prepare several types of response/message recording tapes in advance with different response messages recorded at the beginning, you can exchange those tapes as needed. This is because the response message can be easily stored in the memory section 20 via the voice analysis section 2.

Specifically, in this embodiment, for example, when a new response/message recording tape is loaded, a loading detection signal is supplied from the tape mechanism 22 to the controller unit 40. Upon receiving this, the controller unit 40 sends a signal to the tape mechanism 22, rewinds the response/message@sound tape, and sends a reset signal to the voice analysis unit 2 and voice synthesis unit 4, and sends a reset signal to the memory unit 20.
Outputs a clear signal to. As a result, the memory section 20
Preliminary pressure Even if there is voice analysis data up to that point, the data is erased. .

Thereafter, when the answering button 42 is pressed, the machine enters the waiting state in the above-mentioned answering mode.

At this time, if the other party calls,
In this case, the response message at the beginning of the tape is stored in the memory section 20 in the same manner as when the voice analysis data is destroyed due to the power outage, and then the response message at the beginning of the tape is stored in the memory section 20, and then the response message is stored at the end of the recording of the tape. The response message is recorded. Further thereafter, the response message is cued. In other words, it prepares for the next reception.

(Effects of the Invention) As is clear from the above description, the present invention has the following effects.

(1) Easily analyze the voice of the response message at the beginning of the response/message recording tape and store the analyzed data in the memory unit, without having to use battery backup to retain the voice data. I can do it.

(2) If the user wants to change the response message depending on the purpose of use, there is no need to say it each time and store it in the memory, but the response message can be recorded at the beginning of the message and can be used for multiple types of response/message recording. If a tape is prepared, voice analysis data corresponding to any response message can be easily stored in the memory section simply by replacing the tape.

(3) If you prepare a tape for both response and message recording with a response message recorded at the beginning, you can freely use the memory section for other purposes after reception ends until the next reception. This makes it possible to improve the usage efficiency of the memory section.

(4) Compared to conventional automatic response devices that use tapes that do not use voice analysis and synthesis to record responses and messages, it is possible to start recording messages immediately after sending a response message, which significantly reduces the waiting time for the other party. can be shortened to

(5) Since a low-cost RAM4 is used instead of an expensive CMO3RAM in the memory section, it is possible to store a large amount of audio data in the memory section at low cost.

[Brief explanation of the drawing]

Fig. 1 is a block diagram showing an example of an automatic response device using a conventional voice analysis and synthesis method, Fig. 2 is a block diagram showing an embodiment of the present invention, and Fig. 3 is a recording pattern of a tape used for both response and message recording. 1 and 2 of FIG. 4 are flowcharts for explaining an example of the operation of the controller section of FIG. 2. FIG. 2...Speech analysis section, 4...Speech synthesis section, 20.
...Memory section, 22...Tape mechanism, 40--Controller section Michihito Hiraki, Patent Attorney Figure 1

Claims (2)

[Claims] (1) a voice analysis unit that analyzes voice signals and creates voice analysis data; a memory unit that stores the voice analysis data;
In an automatic response device using a speech analysis and synthesis method, which includes a speech synthesis section that synthesizes and sends out speech analysis data read from the memory section, a response message is recorded at the beginning of the response message, and a response message is recorded at the rear of the response message as necessary. Driving a response/message recording tape that alternately and sequentially records messages sent from a caller via a telephone line and response messages read from the memory section via the voice synthesis section; comprising a tape mechanism for recording and reproducing, and a detection means for detecting whether or not the voice analysis data in the memory section is destroyed or erased when reading from the memory section,
When the detection means detects that the voice analysis data is destroyed or deleted, the response message at the beginning of the response/message recording tape is stored in the memory section via the voice analysis section. An automatic response device using a voice analysis and synthesis method, which is characterized by: (2) The automatic response device using a speech analysis and synthesis method as set forth in item 1 of the patent, wherein the memory section is configured using a dynamic RAM.