JPH0876965A - Speech recognition system - Google Patents

Abstract

PURPOSE: To support a user by an operator by calling the operator when the user can not understand a using method for a system and hesitates to execute operation or the like. CONSTITUTION: When a call button 13b on an operation part 13 arranged on a user's terminal 10-1 is depressed by a user, a call request is transmitted from a control part 17 in the terminal 10-1 to a control part 25 in an operator's terminal 20 through a communication line 30-1 and the control part 25 calls an operator and switches a line so as to connect the line to the terminal 10-1. In said state, user's voice inputted from a voice input part 11 in the terminal 10-1 is sent to the terminal 20 and outputted from a voice output part 24 and an operator's voice inputted from a voice input part 21 in the terminal 20 is sent to the terminal 10-1 and outputted from a voice output part 16.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a voice recognition system suitable for use in an information guidance device, an automatic ticket issuing machine and the like.

[0002]

2. Description of the Related Art In recent years, various practical systems using voice recognition for dialog with a user are being considered. Among them, since the voice recognition system using the unspecified speaker voice recognition technology can be used by anyone without the user having to register the voice, information guidance devices, automatic ticketing machines, etc., used by an unspecified number of people Suitable for practical use in social systems.

However, when such a voice recognition system is introduced into the real world, the user is unfamiliar with the system and the degree of understanding of the voice recognition itself by the user is low, and a desired utterance can be obtained for the voice recognition device. It may not be, and the recognition rate will decrease,
In addition, the current voice recognition device has a drawback that the utility value of the voice recognition system is reduced due to the fact that it is difficult to always recognize the user's voice without error.

From the background described above, not only the voice input means accompanied by voice recognition is used as the input means from the user, but it can be used in combination with other input means (operation means) such as a touch panel. A system and a system which guides the usage method by a voice output means using a voice synthesis technique are considered.

As described above, in putting a voice recognition system into practical use in a social system, it is an important issue to realize the purpose so that even a person who uses the system for the first time can achieve the purpose. It wasn't always the case.

[0006]

As described above, the conventional voice recognition system has a problem that it is difficult to realize it so that even a person who uses the system for the first time can always achieve the purpose. .

The present invention has been made in consideration of the above circumstances, and an object thereof is to make a correct operation when a user cannot understand how to use a system (user terminal having a voice recognition function) or even if voice input is repeated many times. An object of the present invention is to provide a voice recognition system that can easily call an operator when not recognized and ask the operator to assist the operator.

Another object of the present invention is to provide a voice recognition system in which one operator can handle a plurality of user terminals. Still another object of the present invention is to provide
An object of the present invention is to provide a voice recognition system in which the operator can surely grasp the usage status of a user with respect to a user terminal.

Still another object of the present invention is to provide a voice recognition system in which the operator can operate the operation requested by the user instead. Still another object of the present invention is to provide
An object of the present invention is to provide a voice recognition system that enables a user to determine whether or not an operator can be called.

Still another object of the present invention is to provide a voice recognition system capable of automatically calling an operator by detecting a state in which the user cannot achieve the purpose of using the system and is in trouble. .

[0011]

A voice recognition system according to a first aspect of the present invention is a user terminal having a voice recognition function for carrying out a dialogue with a user, and a voice input operation by the user. User terminal side voice input means, voice recognition means for recognizing voice input from the user terminal side voice input means, user terminal side display means for displaying the recognition result and various information of the voice recognition means, User terminal side voice output means for notifying the person of various information by voice,
User terminal side operation means for input operation other than voice,
And a user terminal provided with a user terminal side control means for controlling each of the above means, and an operator terminal for a user and an operator of the user terminal to have a conversation, which are input by the user terminal side voice input means. Operator terminal side audio output means for listening to the user's audio input,
And an operator terminal equipped with an operator terminal side voice input means for notifying the user of the user terminal of the operator's voice.

[0012] A voice recognition system according to a second aspect of the present invention comprises a plurality of user terminals in the voice recognition system according to the first aspect, and the operator terminal has a plurality of user terminals. It is characterized by further comprising an operator terminal side control means for switching and connecting one to the operator terminal.

A voice recognition system according to a third aspect of the present invention is an operator terminal for monitoring the display content of a user terminal side display means on the operator terminal of the voice recognition system according to the first or second aspect. The present invention is characterized in that the side display means is further provided.

A voice recognition system according to a fourth aspect of the present invention is a voice recognition system according to the first, second or third aspect, wherein an operator terminal of the voice recognition system is input by a user using a user terminal side operation means. It is characterized in that it further comprises an operator terminal side operation means for the operator to act on behalf of the operator.

A voice recognition system according to a fifth aspect of the present invention enables a user terminal of the voice recognition system according to the first, second, third or fourth aspect to have a conversation with an operator of an operator terminal. It is characterized by further comprising a state output means for externally outputting the state of whether or not.

A voice recognition system according to a sixth aspect of the present invention is a user terminal of the voice recognition system according to the first, second, third, fourth or fifth aspect, wherein the user terminal and the user. A state detecting means for detecting a state in which the dialogue with the operator does not end even if it exceeds a predetermined time, and an operator automatic calling means for making an operator call to the operator terminal according to the state detection of the state detecting means. It is characterized by having.

[0017]

In the voice recognition system according to the first aspect, the operator can hear the voice input of the user through the voice output means on the operator terminal side, and the voice input by the operator is input by the voice input means on the operator terminal side. The user can be notified through the terminal side audio output means. For this reason, when the user cannot understand how to use the system (user terminal), or when it is not correctly recognized even after repeatedly inputting voice, call the operator and ask if it does not know how to use it. When the voice is not recognized, the purpose of use can be achieved by instructing an input method other than the voice input.

Next, in the voice recognition system according to the second aspect, one of the plurality of user terminals can be selectively switched and connected to the operator terminal by the control means of the operator terminal. Therefore, one operator can handle a plurality of user terminals.

Next, in the voice recognition system according to the third aspect, as a result of the dialogue between the user and the user terminal,
The recognition result of the input voice displayed on the display unit on the user terminal side and various information are also displayed on the display unit on the operator terminal side. Therefore, the operator can grasp the state of the dialogue between the user and the user terminal in detail through the display means.

Next, in the voice recognition system according to the fourth aspect, the operator operates the operator terminal side operation means to obtain the same effect as that of the user operating the user terminal side operation means. . Therefore, the operator can perform the input operation requested by the user by using the operator terminal side operation means instead.

Next, in the voice recognition system according to the fifth aspect, the user can know whether or not the operator can be called by the contents of the status output from the user terminal.

Next, in the voice recognition system according to the sixth aspect, the predetermined time is set because the user is confused about the operation of the user terminal or is not correctly recognized even if the voice is input many times. If the purpose of use of the user terminal cannot be achieved even after the elapse of time, that is, if the dialogue between the user and the user terminal does not end within a predetermined time, the state is detected by the state detecting means. Has been
An operator call is automatically made to the operator terminal.

[0023]

DESCRIPTION OF THE PREFERRED EMBODIMENTS A voice recognition system according to an embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram of a voice recognition system showing the same embodiment.

The system of FIG. 1 interacts with a user to provide information requested by the user, for example, n user terminals 10-1 to 10-n as information guide devices, and the user terminals. One operator terminal 20 to which 10-1 to 10-n are connected, and a communication line 30-1 to be used for communication between these user terminals 10-1 to 10-n and the operator terminal 20.
It consists of 30-n.

The user terminal 10-1 includes a voice input unit 11, a voice recognition unit 12, an operation unit 13, a state detection unit 14, and a display unit 1.
5, a voice output unit 16 and a control unit 17. The voice input unit 11 is used by the user to input a voice and has a microphone 11a.

The voice recognition unit 12 is for recognizing the voice input by the voice input unit 11. Operation part 1
Reference numeral 3 is for the user to perform an input operation other than voice input. For example, the operation button 13a used for cursor up / down / left / right cursor operation, etc.
Has an operator call button 13b used as an operating means for calling the operator.

The state detecting unit 14 is for detecting the use state of the user with respect to the user terminal 10-1, and is a sensor for optically detecting that the user stands in front of the user terminal 10-1. 14a.

The display unit 15 is for displaying the recognition result of the input voice uttered by the user, the usage guidance for the user, various information requested by the user, and the like.
Displays such as displays and liquid crystal displays 15
a.

The voice output unit 16 is for outputting various information to the user by voice, for example, a speaker 16a.
have. The control unit 17 has a function of controlling the entire user terminal 10-1 to control the dialogue with the user, and also has a call button 1
A call request according to the operation of 3b, an input voice from the voice input unit 11, display contents of the display unit 15, and the like are displayed on the operator terminal
It has the function of transmitting to 0.

The configuration of the other user terminals 10-2 to 10-n is the same as that of the user terminal 10-1 described above. The operator terminal 20 includes a voice input unit 21 and an operation unit 2
2, a display unit 23, a voice output unit 24, and a control unit 25.

The voice input unit 21 is used by the operator to input voice, and is for notifying the user of the user terminal 10-i (i = 1 to n) of the voice of the operator.

The operation section 22 is used for operator operation, and is for the operator to perform input operation by the user of the user terminal 10-i. The display unit 23 has a display (not shown) used for display for the operator, and the display content on the user terminal 10-i (in the example of the user terminal 10-1, the display 15a of the display unit 15). The contents of the screen displayed in (1) are monitored by the operator.

The voice output unit 24 is used for voice output to the operator, and is for the operator to listen to the voice input of the user of the user terminal 10-i.
The control unit 25 controls the entire operator terminal 20,
Control the communication line 30-i.

The configurations of the other user terminals 10-2 to 10-n are also as follows.
The configuration is the same as that of the user terminal 10-1 shown in FIG. FIG. 2 shows the appearance of the user terminal 10-1 in FIG.

As shown in FIG. 2, a microphone 11a and an operation button 1 are provided on the front surface of the housing of the user terminal 10-1.
3a, a call button 13b, a sensor 14a, and a display 15a are provided. Also, the display 15a
An operator standby display area 15b for displaying information on whether or not a conversation with the operator of the operator terminal 20 is possible, that is, whether or not the operator is on standby, is secured in a predetermined area of the display screen of There is. The external appearances of the casings of the other user terminals 10-2 to 10-n are the same as the external appearances of the casing of the user terminal 10-1 shown in FIG.

Next, the operation of the configuration shown in FIGS. 1 and 2 will be described. First, when the user approaches the front of the user terminal 10-1, for example, the fact is detected by the sensor 14a in the state detection unit 14. Then, the state detector 14 to the controller 17
In response, a state detection notification indicating that the user has stood in front of the user terminal 10-1 is sent. Upon receiving the state detection notification, the control unit 17 sends the user guidance (operation guidance) to the display unit 15 and the voice output unit 16, and displays the user guidance on the display screen of the display 15a and the speaker 16a. The user guidance voice output is performed.

The user has a display unit 15 and a voice output unit 16.
While listening to and listening to the usage guidance using the display 15a and the speaker 16a, the voice is input by speaking into the microphone 11a of the voice input unit 11.

The voice input from the microphone 11a is passed to the voice recognition unit 12 by the voice input unit 11. The voice recognition unit 12 performs the recognition process of this input voice and passes the recognition result to the control unit 17.

The control unit 17 outputs the recognition result of the voice recognition unit 12 and various information according to (the request indicated by) the recognition result from the display 15a and the speaker 16a. In this way, the user inputs the voice by inputting the user terminal 10-1
Achieve your goals by interacting with. In addition to this, the user performs an auxiliary operation by using the operation button 13a, or performs an input operation by using an input means (not shown) such as a touch panel as an alternative to the voice input, whereby the user terminal 10-1. You can also interact with and achieve your goals.

Here, when the user cannot understand how to use the user terminal 10-1 (in the voice recognition system), or even if the voice is input many times, the user terminal 10-1 (the voice recognition unit 12 of the user terminal 10) correctly recognizes it. Suppose that some cases have not been recognized.

In such a case, the user presses the call button 13b of the operation unit 13 to call the operator of the operator terminal 20. The pressing of the call button 13b is detected by the control unit 17 of the user terminal 10-1. When the control unit 17 detects that the call button 13b has been pressed, it transmits a call request to the control unit 25 of the operator terminal 20 via the communication line 30-1.

Upon receiving the call request from the control unit 17 of the user terminal 10-1, the control unit 25 of the operator terminal 20 displays a predetermined message for calling the operator on the display unit 23. Alternatively, the message may be output by voice from the voice output unit 24. Further, the control unit 25 switches the line (communication line) so that it can be connected to the user terminal 10-1 of the call request source among the n user terminals 10-1 to 10-n connected to the operator terminal 20. . Thereby, the operator terminal 20
Of the user terminal 10-1 can respond to the call from the user terminal 10-1, and a voice conversation (dialogue) is performed between the user of the user terminal 10-1 and the operator of the operator terminal 20 as described below. You will be told.

First, the line is switched as described above in response to the operator call request from the user terminal 10-1, and the user terminal 10-1 is operated by the user while the operator terminal 20 is connected to the user terminal 10-1. When a voice is input from (the microphone 11a of) the voice input unit 11 of -1, the input voice (voice signal) is controlled by the control unit 17 of the user terminal 10-1 via the communication line 30-1 by the control unit of the operator terminal 20. 25 is output to the voice output unit 24 of the operator terminal 20 by the control unit 25. As a result, the operator can hear the user's voice through the voice output unit 24.

Similarly, the operator operates the operator terminal 20.
When a voice is input from the voice input unit 21 of the above, the voice (voice signal) is sent by the control unit 25 to the control unit 17 of the user terminal 10-1 via the communication line 30-1, and the control unit 17 concerned.
Is output from (the speaker 16a of) the audio output unit 16 of the user terminal 10-1. As a result, the user can hear the operator's voice through the voice output unit 16.

As described above, in the present embodiment, when the user of the user terminal 10-1 presses the call button 13b to call the operator of the operator terminal 20, the operator uses the voice output unit 24. It is possible to listen to the voice of the person, convey the voice to the user by the voice input unit 21, and have a conversation with the user. on the other hand,
The user of the user terminal 10-1 talks to the operator using the microphone 11a of the voice input unit 11 and the speaker 16a of the voice output unit 16, and when he / she does not know how to use the user terminal 10-1, asks for it. Further, when the voice is not correctly recognized, the object can be achieved by instructing an input method other than the voice input.

In addition to this, in the present embodiment, the user of the user terminal 10-1 and the operator of the operator terminal 20 have the control unit 17 and the operator terminal 2 in the user terminal 10-1.
When a conversation is performed under the control of the control unit 25 within 0, the screen content (display screen information) displayed on (the display 15a of) the display unit 15 of the user terminal 10-1 is changed to the control unit 17 Operator terminal 20 via communication line 30-1
Transferred to the side. And the transferred screen information is
It is displayed on the display unit 23 by the control unit 25 of the operator terminal 20.

Therefore, the operator is the user terminal 10
-1 indicates the content displayed on the display unit 15 (that is, the state of the dialogue between the operator and the user terminal 10-1).
It is possible to give an instruction to the user while monitoring the display unit 23.

Further, in the present embodiment, the operating section 22 provided on the operator terminal 20 side can be used as follows. First, as described above, the user terminal 10-1
When the operator operates the operation unit 22 when an operator is called from the user terminal 1, the operation information is controlled by the control unit 25 through the communication line 30-1 to the user terminal 1.
Sent to 0-1. The control unit 17 of the user terminal 10-1 receives this operation information as if it was input information on the user terminal 10-1 by a user operation, and performs processing according to the information.

As described above, since the operator can substitute the input operation of the user by using the operation unit 22 of the operator terminal 20, it is possible to perform the auxiliary operation of the user or achieve the purpose of the user. It will be possible.

The above is the same when the operator call is made to the operator terminal 20 from the other user terminals 10-2 to 10-n. As described above, in the present embodiment, the voice recognition system is configured by connecting the n user terminals 10-1 to 10-n to the one operator terminal 20.
By switching the line so as to connect to the user terminal 10-i that has issued the operator call request, the operator can respond to the call from the user of the user terminal 10-i. Normally, it is unlikely that a plurality of user terminals will simultaneously make an operator call request. Therefore, with such a system configuration, the operator 1
The number of operators can be reduced by coping with the users of n user terminals 10-1 to 10-n.

In the present embodiment, as will be described below, the user of the user terminal 10-i can identify whether or not a conversation with the operator is possible by the user terminal 10-i. ing.

First, the operator of the operator terminal 20 inputs the status information indicating whether or not it is possible to answer the call from the user terminals 10-1 to 10-n by operating the operation unit 22, for example. This status information is received by the control unit 25. The control unit 25 sends the information to the communication lines 30-1 to 30-n.
To the user terminals 10-1 to 10-n. Then, for example, in the user terminal 10-1, the control unit 17 receives the status information from (the control unit 25 of) the operator terminal 20, and the information, that is, the operator is displayed in the operator standby display area 15b of the display 15a of the display unit 15. Information indicating whether or not the call can be accepted (whether or not the operator is waiting) is displayed. The user of the user terminal 10-1 can visually recognize whether or not a conversation with the operator is possible when he / she wants to call the operator by the display contents of the operator standby display area 15b. This is useful when the operator may be absent. The above operator standby display is similarly performed on user terminals other than the user terminal 10-1.

Further, (the control unit 2 of the operator terminal 20
In 5), when the operator call request from the user terminal 10-i is received, the control unit 25 controls
An operator standby display indicating that the operator cannot answer the call is displayed on each user terminal other than the user terminal 10-i. This is useful when the operator is being called by a user of another user terminal.

Further, in the present embodiment, the operator can be automatically called based on the user terminal usage state of the user. Regarding this, the user terminal 10-1
The automatic call from the operator will be described as an example.

The control unit 17 of the user terminal 10-1 uses the sensor 1 provided in the state detection unit 14 of the user terminal 10-1.
The usage time of the user is measured using 4a, and the usage status of the user terminal of the user is detected from this usage time and the internal status of the user terminal 10-1. Then, the control unit 17 automatically calls the operator based on the detected user terminal usage state. Hereinafter, the flow of this series of processes will be described with reference to the flowchart of FIG.

First, the control unit 17 of the user terminal 10-1
The time measurement value t is cleared (step S1). Next, the control unit 17 checks whether or not the output of the sensor 14a is turned on (ON) (step S2), and if it is not turned on, the process returns to step S1.

On the other hand, if the user stands in front of the user terminal 10-1 and the output of the sensor 14a is turned on,
The control unit 17 increments the time measurement value t by 1 (step S3). Next, the control unit 17 checks whether or not the output of the sensor 14a remains on (step S4),
If the user terminal 10 remains on (that is, the user
If the time measurement value t does not exceed the predetermined time (threshold value) T, it is checked (step S5). A value obtained by adding a certain margin to the user's normal system usage time (standard system usage time) is used as the threshold value T. Also, if the output of the sensor 14a is turned off (that is, if the user is the user terminal 10-1).
(If left), return to step S1.

If it is determined in step S5 that the time measurement value t does not exceed the threshold value T, step S3
Then, the time measurement value t is incremented by 1. On the other hand, if it is determined in step S5 that the time measurement value t exceeds the threshold value T, it is considered that the user's normal system usage time has been exceeded.
Checks whether the purpose of the user has been achieved, for example, based on the internal state of the control unit 17 itself (step S
6). Here, whether or not the purpose of the user has been achieved can be determined by whether or not a ticket has been purchased, for example, if the present system (user terminal 10-1) is an automatic ticket vending machine.

Now, the inspection in step S6 is performed in 1
This is because it is considered that one user continuously uses the system multiple times. In other words, factors that exceed the normal system usage time include the case where the user has some trouble, such as the user getting lost in operation or not being recognized correctly even if voice is input many times, This is because it is considered that some users may continuously use this system multiple times. If the user's purpose is not achieved even if the normal system usage time is exceeded (even if the time measurement value t exceeds the threshold value T), it is determined that the user has some problem. On the contrary, if the user's purpose is achieved, it can be judged that one user is using this system a plurality of times in succession.

When it is determined in step S6 that the user's purpose has been achieved, the control section 17 returns to step S1. On the other hand, when it is determined in step S6 that the purpose of the user has not been achieved, the control unit 17 determines that the user has some problem and the communication line 3
The operator call request is notified to the control unit 25 of the operator terminal 20 via 0-1, and the operator is automatically called (step S7).

As described above, in the voice recognition system of the present embodiment, when the user cannot understand how to use the system (user terminal), or when the user does not correctly recognize the voice input many times, the operator manually operates. Or call it automatically, let the operator understand the situation in detail, ask the operator if you do not know how to use it, and ask the operator to enter an input method other than voice input if voice is not recognized, or substitute the operation. The purpose can be achieved by receiving.

The present invention is not limited to the above embodiment. For example, in the voice recognition system of the embodiment, the system configuration is shown assuming an information guidance device.
Even if an automatic ticket issuing machine, a vending machine, or the like equipped with a money inserting means, a ticket extracting means, etc. is configured, the present invention can be variously modified and implemented without departing from the scope of the invention.

In the configuration of the user terminal 10-1 shown in FIG. 2, the microphone 11a and the speaker 16a are integrated into a structure like a telephone receiver, or a touch panel is used instead of the input means realized by the operation button 13a. It is also possible to adopt the configuration used. Further, the number of sensors forming the state detecting unit 14 is not limited to one, and it is possible to obtain a more detailed user state by using a plurality of sensors. Furthermore, it is possible to use public telephone lines as the communication lines 30-1 to 30-n and to operate from a remote place.

[0064]

As described above in detail, according to the present invention, when the user cannot understand how to use the voice recognition system, or when the user does not recognize correctly even if the voice is input many times,
Since the operator can be called manually or automatically and the operator can give instructions and operation assistance, it is possible to provide a user-friendly system in which even a person who uses the system for the first time can always achieve the purpose. In addition, the user can learn how to use the system by talking with the operator, and can gradually become accustomed to the system. Further, the system provider side can know, through the operator, where the system is difficult to use and where the problem is, so that useful information for improving the system can be obtained and a better system can be obtained. Can be provided.

[Brief description of drawings]

FIG. 1 is a block configuration diagram of a voice recognition system according to an embodiment of the present invention.

FIG. 2 is an external view of a user terminal 10-1 in the embodiment.

FIG. 3 is a flowchart showing a flow of a series of processes in the user terminal 10-1 for automatically calling an operator in the embodiment.

[Explanation of symbols]

10-1 to 10-n ... User terminal, 11 ... Voice input unit (user terminal side voice input means), 11a ... Microphone, 1
2 ... Voice recognition section, 13 ... Operation section (user terminal side operation means), 13a ... Operation button, 13b ... Call button, 14
... state detection section, 14a ... sensor, 15 ... display section (user terminal side display means), 15a ... display, 15b ... operator standby display area (state output means), 16 ... voice output section (user terminal side voice output means) , 16a ... Speaker, 17 ... Control unit (user terminal side control means, state detection means, operator automatic calling means), 20 ... Operator terminal, 21 ... Voice input section (operator terminal side voice input means), 22 ... Operation section ( Operator terminal side operation means), 2
3 ... Display unit (operator terminal side display unit), 24 ... Voice output unit (operator terminal side voice output unit), 25 ... Control unit (operator terminal side control unit), 30-1 to 30-n ... Communication line.

Continuation of front page (51) Int.Cl. ⁶ Identification number Office reference number FI technical display location G10L 3/00 561 C

Claims (6)

[Claims] 1. A user terminal having a voice recognition function for interacting with a user, comprising a user terminal side voice input means for voice input operation by the user, and input from this user terminal side voice input means. Voice recognition means for recognizing the generated voice, user terminal side display means for displaying the recognition result of this voice recognition means and various information, user terminal side voice output means for notifying the user of various information by voice, and Is a user terminal provided with a user terminal side operating means for another input operation, and a user terminal side controlling means for controlling each of the means, and an operator for conversation between a user of the user terminal and an operator. A terminal, an operator terminal side voice output means for listening to the user's voice input inputted by the user terminal side voice input means, and an operator's voice Speech recognition system, characterized by comprising an operator terminal having an operator terminal side audio input means for notifying the user of the user terminal. 2. A plurality of user terminals having a voice recognition function for dialogue with a user, the user terminal side voice input means for voice input operation by the user, and the user terminal side voice input means. Voice recognition means for recognizing the voice input from the user, user terminal side display means for displaying the recognition result of this voice recognition means and various information, user terminal side voice output means for notifying the user of various information by voice, User terminal side operation means for input operation different from voice, and a plurality of user terminals each having a user terminal side control means for controlling the respective means, and a user and an operator of the user terminal have a conversation And an operator terminal for performing one of the plurality of user terminals.
To the operator terminal side control means for switching and connecting one to the operator terminal, and listening to the user's voice input inputted by the user terminal side voice input means of the user terminal switched and connected by the operator terminal side control means. And an operator terminal equipped with an operator terminal side voice input means for notifying the user of the user terminal of the switch-connected voice of the operator. Speech recognition system. 3. The voice recognition system according to claim 1, wherein the operator terminal further comprises an operator terminal side display means for monitoring the display content of the user terminal side display means. . 4. The operator terminal further comprises operator terminal side operation means for allowing an operator to act on behalf of an input operation by a user using the user terminal side operation means. Item 5. The voice recognition system according to any one of Items 3. 5. The user terminal further comprises status output means for externally outputting a status indicating whether or not a conversation with the operator of the operator terminal is possible. Item 5. The voice recognition system according to any one of items 4. 6. The state detection means for detecting, in the user terminal, a state in which the dialogue between the user terminal and the user does not end even if a predetermined time elapses, and the state detection means detects the state. The voice recognition system according to any one of claims 1 to 5, further comprising an automatic operator calling means for calling an operator to the operator terminal.