JP2015148648A - Dialogue system, speech controller, dialog unit, speech control method, control program of speech controller and control program of dialog unit - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent erroneous recognition or erroneous operation by detecting circumstances under which erroneous recognition of a voice or erroneous operation caused therefrom may occur.SOLUTION: A dialogue system 401 is a system for controlling a dialog unit which outputs a response voice responding to a voice output by a user. A speech control server 1 includes: a condition determination section 56 that, when a dialog unit 2 erroneously detects a voice other than a recognition object as the recognition object, determines a voice erroneous recognition condition is satisfied; and a speech control section 57 that, when the voice erroneous recognition condition is satisfied, controls not to output any voice. The determination whether or not the condition is satisfied is made based on operation information of a TV 3.

Description

  The present invention relates to a dialog system that recognizes voice, and more particularly to a dialog system that recognizes voice uttered by a user and outputs a response voice to the voice.

  2. Description of the Related Art Conventionally, there is an interactive device that performs processing corresponding to a voice recognition result for voice uttered by a user.

  For example, in Patent Document 1, switching between a normal conversation mode and a wireless communication device usage mode (a mode in which hands-free operation is performed by voice input) is performed by inputting voice (for example, a specific keyword “mode switching”). A portable wireless communication device is disclosed. This avoids the problem that the user does not intend to give an instruction to the portable wireless communication device but erroneously recognizes the voice of the normal conversation made by the user as the instruction.

  In addition, an information processing system for allowing a user to talk in consideration of the safety of the user during driving is disclosed. Specifically, in Patent Document 2, communication control for grasping a traveling situation based on vehicle sensor information and car navigation information and adding sound data describing the traveling situation to input voice data of a driver. An apparatus is disclosed. As a result, it is possible to inform the caller outside the vehicle who talks with the driver about the driving situation of the driver and to have a conversation in consideration of the driving situation.

  Furthermore, Patent Document 3 discloses a voice dialogue apparatus that can perform a smooth dialogue in accordance with the sensitivity of a speaker. Specifically, in a normal state, the voice interaction device outputs a response voice at a speed corresponding to the user's speaking speed, and when a predetermined event occurs while outputting the response voice, Increase or decrease the output speed. Furthermore, the voice interactive apparatus exhibits a standby state in which the provision of the response voice is delayed according to the traveling state of the vehicle.

JP 2002-33794 A (published January 31, 2002) JP 2006-352356 A (released on December 28, 2006) JP 2008-26463 A (published February 7, 2008)

  In the above-described prior art, depending on the environment in which the user or the interactive device is placed, the speech recognition function of the interactive device does not work accurately, and the input speech is misrecognized. (Such as outputting an incorrect answer voice). As an environment in which such erroneous recognition or malfunction may occur, for example, an environment in which voice is output from another voice output device and the voice is detected by the dialogue device, or the user does not respond to the dialogue device. There may be an environment where the user is speaking to another object (person or device).

However, in the above-described conventional technology, there is a problem that an appropriate process for avoiding erroneous recognition or malfunction cannot be performed because the above-described environment in which erroneous recognition or malfunction can occur cannot be detected. Specifically, in the technique of Patent Document 1, in order to switch the dialogue apparatus to the mode (not recognizing voice), an intentional operation (voice input) for instructing switching by the user is required. Furthermore, erroneous recognition cannot be avoided if the user's effort increases and the user fails to perform the above operation. In addition, the techniques of Patent Documents 2 and 3 detect the traveling state of the vehicle, and cannot detect an environment in which erroneous recognition or malfunction may occur.

  The present invention has been made in view of the above-mentioned problems, and its object is to detect that there is an environment in which erroneous recognition of speech or a malfunction associated therewith can occur and to avoid erroneous recognition or malfunction. It is to realize a dialogue system that can do this.

  In order to solve the above problem, a dialog system according to an aspect of the present invention is a dialog system that controls a dialog device that outputs a response voice in response to a voice uttered by a user. A non-recognition voice that is not a recognition target voice, and a condition determination unit that determines that the voice recognition condition of the dialog device is satisfied when the dialog device can be erroneously detected as a recognition target voice; Utterance control means for controlling so that the reply voice is not output by the interactive device when it is judged by the means that the voice error recognition condition is satisfied, and the condition judgment means directly outputs the non-recognition target voice. Or the success or failure of the said audio | voice recognition condition is judged based on the operation information which shows the operation condition of the non-target audio | voice generation source apparatus to generate | occur | produce indirectly.

  In order to solve the above problem, an utterance control device according to an aspect of the present invention is an utterance control device that controls an interactive device that outputs a response voice in response to a voice uttered by a user. A non-recognition voice that is not a recognition target voice that is generated with respect to the above-mentioned condition determination means for determining that the voice recognition condition of the dialog apparatus is satisfied when the dialog apparatus can be erroneously detected as a recognition target voice; Utterance control means for controlling so that the reply voice is not output by the dialogue device when the condition judgment means judges that the voice error recognition condition is satisfied, and the condition judgment means includes the non-recognition voice Is determined based on the operation information indicating the operation status of the out-of-target audio generating device that directly or indirectly generates the above.

  In order to solve the above-described problem, an interactive apparatus according to an aspect of the present invention is an interactive apparatus that outputs a response voice in response to a voice uttered by a user, and a recognition target that the user utters to the interactive apparatus. When a non-recognition voice that is not a voice can be erroneously detected as a recognition target voice by the dialogue apparatus, a condition judgment unit that judges that the voice recognition condition of the dialogue apparatus is satisfied, and the voice judgment error by the condition judgment unit A dialogue control means for suppressing the output of the reply voice when it is judged that a recognition condition is satisfied, and the condition judgment means generates the non-recognized voice directly or indirectly. The success or failure of the voice error recognition condition is determined based on the operation information indicating the operation status of the device.

  In order to solve the above-described problem, an utterance control method according to an aspect of the present invention is an utterance control method for controlling an interactive apparatus that outputs a response voice in response to a voice uttered by a user. A condition determination step for determining that a speech error recognition condition of the interactive device is satisfied when a non-recognition speech that is not a recognition target speech that is issued to the device may be erroneously detected as a recognition target speech by the interactive device; An utterance control step for controlling so that the reply voice is not output by the dialogue device when it is determined in the condition determination step that the voice error recognition condition is satisfied. Based on the operation information indicating the operation status of the non-target audio source device that generates sound directly or indirectly, the success or failure of the above-mentioned sound misrecognition condition is judged. .

  According to one aspect of the present invention, it is possible to appropriately suppress voice output in accordance with the environment in which the interactive device is placed, and as a result, it is possible to avoid voice misrecognition or associated malfunction. There is an effect that can be done.

It is a functional block which shows the principal part structure of each apparatus in the dialogue system which concerns on Embodiment 1 of this invention. It is a figure which shows the outline of the said dialogue system. (A) is a figure which shows the specific example of the information transmitted to information collection server from TV, (b) is a figure which shows the specific example of an apparatus state management table, (c) is an information collection server It is a figure which shows the specific example of the information transmitted to an utterance control server from FIG., (D) is a figure which shows the specific example of an apparatus apparatus corresponding table, (e) is a figure which shows the specific example of a rule table. is there. It is a flowchart which shows the flow of the operation information transmission process which TV performs in the said interactive system, and the flow of the operation information transmission process which an information collection server performs. It is a flowchart which shows the flow of the speech control process which an speech control server performs in the said dialog system, and the flow of the dialog mode control process which a dialog apparatus performs. It is a figure which shows the outline of the dialogue system which concerns on Embodiment 2 (or Embodiment 4) of this invention. It is a functional block which shows the principal part structure of each apparatus in the said interactive system. (A) is a figure which shows the specific example of apparatus ID stored in the memory | storage part of a dialog apparatus, (b) is a figure which shows the specific example of the operation information which a dialog apparatus produces | generates, (c) These are figures which show the specific example of an apparatus state management table, (d) is a figure which shows the specific example of the information transmitted to a speech control server from a dialogue apparatus. It is a flowchart which shows the flow of the operation information transmission process which a dialog apparatus performs in the said dialog system, and the flow of the utterance control process which an utterance control server performs. It is a figure which shows the outline of the dialogue system which concerns on Embodiment 3 of this invention. It is a functional block which shows the principal part structure of each apparatus in the said interactive system. (A) is a figure which shows the specific example of apparatus ID stored in the memory | storage part of a dialogue apparatus, (b) is a figure which shows the specific example of a ringing tone table, (c) is an apparatus state. It is a figure which shows the specific example of a management table, (d) is a figure which shows the specific example of the information transmitted to a speech control server from a dialogue apparatus, (e) is a figure which shows the specific example of a rule table. is there. It is a flowchart which shows the flow of the operation information transmission process which a dialog apparatus performs in the said dialog system, and the flow of the utterance control process which an utterance control server performs. (A) is a figure which shows the specific example of the operation information which an interactive apparatus produces | generates, (b) is a figure which shows the specific example of an apparatus state management table, (c) is an utterance control server from an interactive apparatus. FIG. 4D is a diagram illustrating a specific example of information transmitted to (1), and FIG. 4D is a diagram illustrating a specific example of a rule table. 10 is a flowchart showing a flow of processing of each device of the interactive system in Modification 2. 10 is a flowchart showing a flow of processing of each device of the interactive system in Modification 2. It is the block diagram which illustrated the composition of the computer which can be used as each device (apparatus, server) included in the dialog system of the present invention.

Embodiment 1
Embodiment 1 of the present invention will be described below with reference to FIGS. In the first embodiment, as an example, a case where the dialogue system of the present invention is applied to the dialogue system shown in FIG. 2 will be described. The dialogue system shown in FIG. 2 is merely an example, and does not limit the dialogue system of the present invention.

[Outline of Dialogue System]
FIG. 2 is a diagram showing an outline of the dialogue system according to Embodiment 1 of the present invention. A dialogue system 401 shown in FIG. 2 includes an utterance control server 1, a dialogue device 2, a digital television (hereinafter referred to as TV) 3, and an information collection server 4.

  The dialogue device 2 recognizes the user's utterance as a voice and interacts with the user by outputting a reply voice. The dialogue device 2 is, for example, a self-propelled cleaner, but any information processing device may be used as long as it has the above dialogue function and the function to communicate with the utterance control server 1. It may be a humanoid robot, a personal computer, a tablet terminal, a smartphone, or the like.

  The utterance control server (speech control device) 1 controls the voice output of the dialog device 2 by transmitting an instruction related to the utterance to the dialog device 2.

  For the instruction related to the utterance, for example, an “output suppression instruction” for instructing the dialogue apparatus 2 to switch to the dialogue mode that does not output the response voice, and the dialogue apparatus 2 to return to the dialogue mode that outputs the answer voice. "Suppression cancellation instruction" to be instructed.

  The utterance control server 1 determines the environment in which the interactive device 2 is placed based on the operation information of the TV 3 transmitted from the information collection server 4, and issues an output suppression instruction or suppression cancellation instruction according to the determined environment. It transmits to the dialogue apparatus 2.

  It should be noted that a publicly known function may be adopted as a function for determining what kind of response voice is output for what kind of user's utterance and determining the timing for outputting the answer voice. The function is provided in at least one of the utterance control server 1 and the dialogue apparatus 2.

  Further, in the first embodiment, the dialogue apparatus 2 may spontaneously output a sound in response to an instruction from the utterance control server 1 by using a predetermined event other than the user's utterance as a trigger. When the sound output in this way is distinguished from the response sound output in response to the user's utterance, it is referred to as a spontaneous sound. In other words, the utterance control server 1 may transmit a “spontaneous speech output instruction” that instructs the output of the spontaneous speech to the dialogue apparatus 2 as an instruction related to the utterance.

  For example, when the interactive device 2 is a self-propelled cleaner, the utterance control server 1 reports a spontaneous sound when reporting a predetermined event such as completion of cleaning or abnormality of the self-propelled cleaner to the user. Is sent to the dialogue apparatus 2 so as to output. Alternatively, the utterance control server 1 controls the interactive device 2 so as to report that when it detects an abnormality of the devices in cooperation with other devices in the building where the interactive device 2 is installed. To do. Alternatively, the utterance control server 1 controls the dialogue apparatus 2 so as to report that when a disaster alarm such as heavy rain / earthquake is received in cooperation with an external information providing server (not shown).

  The TV (non-target audio generating source device / audio output device) 3 is cited as an example of an audio output device that brings an environment in which misrecognition may occur to the dialogue apparatus 2. Specifically, the TV 3 outputs a voice (non-recognition target voice) that is not emitted for the purpose of dialogue with the dialogue apparatus 2.

  The information collection server (information collection device) 4 collects information on the TV 3, manages the operating status of the TV 3, and notifies the utterance control server 1.

  FIG. 2 shows only a pair of one TV 3 and one dialogue device 2, but the utterance control server 1 and the information collection server 4 use the service of this dialogue system 401. For each user, a plurality of pairs of TVs 3 (may be other audio output devices) and interactive devices 2 are managed. Therefore, the utterance control server 1 and the information collection server 4 can uniquely identify the TV 3 using the identification information (hereinafter referred to as device ID) of the TV 3, and the identification information (hereinafter referred to as device ID) of the interactive device 2. By using it, the interactive device 2 can be uniquely specified. A plurality of audio output devices (other TVs 3, recording / playback apparatuses, music playback apparatuses, etc.) may be managed in association with one dialog apparatus 2.

  In the first embodiment, when the user operates the TV 3 using the remote control, the TV 3 that has received the control signal of the remote control transmits operation information indicating the content of the control signal to the information collecting server 4 together with the device ID of the own device. The information collection server 4 determines the operating status of the TV 3 based on the received operation information, and transmits operating information indicating the operating status of the TV 3 together with the device ID of the TV 3.

  As described above, the utterance control server 1 determines the environment in which the dialogue device 2 paired with the TV 3 is placed based on the operation information, and transmits an instruction related to the utterance suitable for the environment to the dialogue device 2. To do. In this way, the dialogue apparatus 2 can appropriately suppress the voice output according to the environment by following the instruction of the utterance control server 1, and as a result, the voice recognition error or the accompanying malfunction is avoided. can do.

  The configuration of each device of the interactive system 401 for realizing erroneous recognition or avoiding malfunction will be described in detail below.

[Functional configuration of each device in the interactive system]
FIG. 1 is a functional block showing the main configuration of each device in the interactive system according to the first embodiment of the present invention.

(Main components of the utterance control server 1)
As shown in FIG. 1, the utterance control server 1 includes a control unit 10 that controls the utterance control server 1 and a storage unit 11 that stores various data used by the control unit 10. The utterance control server 1 includes blocks such as a communication unit for controlling the interactive device 2 and accessing information on the Internet, and an input unit for inputting data to the utterance control server 1. The illustration of this block is omitted.

  The control unit 10 includes an operation information receiving unit 54, a control target specifying unit 55, a condition determining unit 56, and an utterance control unit 57 as functional blocks.

  The storage unit 11 stores a device device correspondence table 71 and a rule table 74.

  The operation information receiving unit 54 receives operation information transmitted from the information collecting server 4 via the communication unit.

The operation information is information indicating the operation status of the audio output device (TV 3 in the present embodiment) managed by the information collection server 4. The operation information is transmitted from the information collection server 4 to the utterance control server 1 when the operation status of the TV 3 changes or periodically. In the dialogue system 401, a plurality of TVs 3 are managed. In this case, the operation information is transmitted to the operation information receiving unit 54 in association with a device ID for uniquely identifying the TV 3. The operation information receiving unit 54 supplies the received device ID to the control target specifying unit 55 and the received operation information to the condition determining unit 56. A specific example of the operation information will be described later in detail with reference to FIG.

  The control target specifying unit 55 specifies the interactive device 2 to be controlled based on the device ID supplied from the operation information receiving unit 54. Specifically, since the device ID specifies which TV 3 the supplied operation information belongs to, the control target specifying unit 55 selects the interactive device 2 associated with the TV 3 based on the device ID. It can be specified by referring to the device device correspondence table 71. The dialog device 2 specified in this way is recognized by the utterance control unit 57 as an object whose utterance should be controlled according to the operation information of the TV 3. A specific example of the device device correspondence table 71 will be described in detail later with reference to FIG.

  Based on the operation information of the TV 3 supplied from the operation information receiving unit 54, the condition determination unit 56 is a condition (hereinafter, referred to as “a voice recognition error”) in an environment where the interactive device 2 that is paired with the TV 3 is placed. It is determined whether or not (voice misrecognition condition) is satisfied. In the first embodiment, when the operation information indicates that the audio output device is in an operation state of outputting (or predicted to output) an audio that may be erroneously recognized as a user's utterance, The condition determining unit 56 determines that the voice recognition condition is satisfied. For example, you may refer to the rule table 74 with which the success or failure of voice recognition conditions is matched for every content which operation information shows. A specific example of the rule table 74 will be described in detail later with reference to FIG.

  The utterance control unit 57 transmits an instruction related to the utterance to the interactive device 2 specified by the control target specifying unit 55 according to the success or failure of the voice error recognition condition determined by the condition determination unit 56. . Specifically, the utterance control unit 57 transmits an output suppression instruction for controlling the dialogue apparatus 2 so as not to output a reply voice to the dialogue apparatus 2 when it is determined that the voice recognition condition is satisfied. On the other hand, when it is determined that the voice recognition condition is not satisfied, a suppression release instruction for controlling the dialog device 2 to output a response voice is transmitted to the dialog device 2.

  In addition, when operation information is periodically received, the success or failure of the speech error recognition condition determined based on the latest operation information is the success or failure of the speech error recognition condition determined based on the operation information received immediately before. It is assumed that there will be no change. In such a case, since the utterance control unit 57 continuously transmits the same instruction, the utterance control unit 57 may be configured to cancel the transmission.

  Further, the utterance control unit 57 controls each part in the own device (that is, without sending an instruction related to the utterance to the dialogue device 2), so that the answering sound is not finally output by the dialogue device 2. It may be configured to be controlled. The configuration of the utterance control unit 57 will be described in detail in Modification 3.

(Main part configuration of the dialogue device 2)
As shown in FIG. 1, the dialogue apparatus 2 includes a control unit 20 that controls the dialogue apparatus 2 in an integrated manner, a storage unit 21 that stores various data used by the control unit 20, a voice input unit 22 that receives voice input, and An audio output unit 23 that outputs audio is provided. The dialog device 2 has functions as a communication unit for receiving an instruction related to an utterance from the utterance control server 1, a processing unit for processing a speech recognition result to realize a dialog, and a self-propelled cleaner. Although each part for implement | achieving is provided, illustration is abbreviate | omitted about these blocks.

  The control unit 20 includes a dialogue mode control unit (conversation control unit) 58, a voice recognition unit (voice recognition unit) 59, and a voice output control unit (voice output control unit) 60 as functional blocks.

The voice recognition unit 59 analyzes the digital signal of the voice input via the voice input unit 22 and converts words included in the voice into a text format. The converted text data is handled by each downstream processing unit (not shown) for generating a suitable response voice.
As the voice recognition unit 59, a known voice recognition technique may be adopted as appropriate.

  The voice output control unit 60 supplies the voice data (response voice or spontaneous voice) generated in the utterance control server 1 or the dialogue apparatus 2 to the voice output unit 23 and outputs the voice data as voice that can be heard by the user. The audio output unit 23 is controlled.

  The dialogue mode control unit 58 controls the operations of the voice recognition unit 59 and the voice output control unit 60 in accordance with an instruction related to the utterance of the utterance control server 1. Specifically, when an output suppression instruction is received from the utterance control server 1, the dialogue mode control unit 58 causes the voice recognition unit 59 and the voice output so that the dialogue device 2 enters a dialogue mode in which no response voice is output. At least one operation of the control unit 60 is suppressed. On the other hand, when the suppression release instruction is received from the utterance control server 1, the dialogue mode control unit 58 returns the voice recognition unit 59 and the voice output control unit so that the dialogue apparatus 2 returns to the dialogue mode in which a response voice is output. The suppression of the operation for 60 is released. It should be noted that the dialog mode control unit 58 sets the voice output control unit 60 to output the spontaneous voice when receiving the spontaneous voice output instruction from the utterance control server 1 even in the interactive mode in which no reply voice is output. You may control.

  In the following, a configuration in which the dialogue mode control unit 58 disables the voice recognition function of the voice recognition unit 59 is adopted as a configuration for switching to the dialogue mode in which no reply voice is output. “Disabling the voice recognition function” means that even if the voice input unit 22 detects some voice and inputs the voice data to the control unit 20, the voice data is processed to generate text data. This means that the voice recognition unit 59 is not executed. As a result, even if some kind of sound is generated around the interactive device 2, the interactive device 2 can be prevented from outputting a response sound. However, the configuration of the interactive mode control unit 58 is not limited to the above. For example, the dialogue mode control unit 58 may suppress the operation of the voice output control unit 60 so that the reply voice is not output to the voice output unit 23. As a result, even if some kind of voice is input, voice is recognized by the voice recognition unit 59, and a reply voice is generated based on the recognition result, the dialogue apparatus 2 can be prevented from outputting the reply voice. Alternatively, the dialogue mode control unit 58 may suppress the operation of the processing unit so that the processing unit (not shown) does not process the text data of the speech recognition result. As a result, even if some voice is input, voice recognition is performed by the voice recognition unit 59, and text data is generated, the corresponding response voice is not generated. Therefore, the dialogue apparatus 2 outputs a response voice to the voice. You can avoid it.

(Main components of audio output equipment)
As shown in FIG. 1, a TV (audio output device) 3 includes a control unit 30 that controls the TV 3 in an integrated manner, a storage unit 31 that stores various data used by the control unit 30, and a remote control for a user to operate the TV 3. The operation unit 32 that functions as: The TV 3 includes a communication unit for transmitting various types of information to the information collection server 4 and each unit for realizing a function as a digital television. However, these blocks are not illustrated. .

  The control unit 30 includes an operation information transmission unit (operation information transmission unit) 50 as a functional block.

The operation information transmission unit 50 receives a control signal sent from the operation unit 32 and transmits operation information indicating the content of the control signal to the information collection server 4. Specifically, the operation unit 32 as a remote controller includes a power on / off button, a number button, a cross (up, down, left, right) button, a determination button, a return button, a data broadcast display button, and four colors (blue, red). , Green, yellow) buttons etc. are provided. The operation information transmission unit 50 transmits information on the button pressed by the user to the information collection server 4 as operation information. At this time, the operation information transmission unit 50 associates the device ID of the own device stored in the storage unit 31 with the operation information, and the information collection server 4
Send to. A specific example of the operation information will be described in detail later with reference to FIG.

  Alternatively, the operation information transmitting unit 50 may be configured to transmit the operation information every time a state change of the TV 3 occurs, instead of transmitting the operation information every time a button operation occurs. More specifically, the operation information transmitting unit 50 determines that the pressed button is only in the case where the state of the TV 3 changes as the TV 3 performs some operation according to the button of the operation unit 32 being pressed. Information is transmitted to the information collection server 4 as operation information. For example, the operation information transmission unit 50 displays the operation information of the button when the power of the TV 3 is switched on / off by pressing the button, when the viewing channel is switched, or when the input is switched to the external input. It transmits to the information collection server 4. Thereby, transmission of the operation information can be omitted when the state of the TV 3 does not change even when the button is pressed.

(Main components of the information collection server 4)
As shown in FIG. 1, the information collection server 4 includes a control unit 40 that controls the information collection server 4 in an integrated manner, and a storage unit 41 that stores various data used by the control unit 40. The information collection server 4 receives information transmitted from the TV 3 or transmits information to the utterance control server 1 and an input unit for inputting data to the information collection server 4. Although these blocks are provided, illustration of these blocks is omitted.

  The control unit 40 includes an operation information reception unit (operation information reception unit) 51, an operation information generation unit (operation information setting unit) 52, and an operation information transmission unit (operation information transmission unit) 53 as functional blocks.

  The storage unit 41 stores a device state management table 70.

  The operation information receiving unit 51 receives operation information from the TV 3 together with the device ID of the TV 3. Then, the device state stored in association with the device ID received in the device state management table 70 is updated based on the received operation information. For example, the operation information receiving unit 51 refers to state transition information (not shown) stored in the storage unit 41. The state transition information is information in which the state immediately before the TV 3, the event (button press), and the state after the transition are associated with each other. Thereby, the operation information receiving part 51 can determine which state is changed when which button is pressed when the TV 3 is in which state based on the received operation information.

  The operation information receiving unit 51 may be configured to receive the latest device state from the TV 3 instead of the operation information. In this case, the above-described function of determining the latest device state based on the operation information in the operation information receiving unit 51 is provided in the TV 3. Then, the operation information receiving unit 51 only needs to update the device state stored in the device state management table 70 to the latest received device state, and the configuration of the operation information receiving unit 51 can be simplified.

  The operation information generation unit 52 generates or updates operation information based on the state of the TV 3 managed for each TV 3 in the device state management table 70. When the operation information reception unit 51 updates the state of the TV 3 stored in the device state management table 70, the operation information generation unit 52 updates the operation information of the TV 3 based on the latest state. To do. The operation information is information indicating whether the TV 3 is operating and what operation is being performed when the TV 3 is operating. The operation information is used by the utterance control server 1 to determine whether or not an audio output device such as the TV 3 provides an environment that satisfies the audio error recognition condition. A specific example of the device state management table 70 will be described later in detail with reference to FIG.

  The operation information transmission unit 53 transmits the operation information generated by the operation information generation unit 52. Specifically, the operation information transmission unit 53 transmits the operation information to the utterance control server 1 when the operation information is updated in the device state management table 70 or periodically. At this time, it is preferable to transmit the device ID in association with the operation information so that the utterance control server 1 can identify the operation information of which TV 3.

  When the operation information is periodically received, it is assumed that the operation information updated based on the latest operation information does not change from the operation information updated based on the immediately previous operation information. In such a case, the operation information transmission unit 53 continuously transmits the same operation information, and thus may be configured to cancel the transmission.

[About information and tables]
In the interactive system 401 according to the first embodiment, information and tables handled by each device are shown in (a) to (e) of FIG. FIG. 3A is a diagram illustrating a specific example of information transmitted from the TV 3 to the information collection server 4. FIG. 3B is a diagram illustrating a specific example of the device state management table 70. FIG. 3C is a diagram illustrating a specific example of information transmitted from the information collection server 4 to the utterance control server 1. FIG. 3D is a diagram illustrating a specific example of the device device correspondence table 71. FIG. 3E is a diagram showing a specific example of the rule table 74.

  FIG. 3 shows a specific example of various information for the purpose of facilitating understanding, and does not limit the configuration of each device. Also, in FIG. 3, the various information data structures shown in the table format are merely examples, and the data structures are not intended to be limited to the table format. Hereinafter, the same applies to other figures for explaining the data structure.

  A specific example will be described. First, in the interactive system 401, when the user is viewing two channels on the TV 3, in order to switch the channel to the previous channel, the upper button of the operation unit 32 (remote control) is pressed. Suppose that it is pressed. The TV 3 selects one channel according to this operation. At this time, the operation information transmitting unit 50 of the TV 3 transmits the operation information indicating the content of the input control signal “up button pressed” and the device ID of the own device to the information collecting server 4 in association with each other. A specific example of the device ID and operation information transmitted at this time is shown in FIG.

When the operation information receiving unit 51 of the information collection server 4 receives the device ID and the operation information shown in FIG. 3A, the information stored in the device state management table 70 using the device ID and the operation information. Update. Specifically, in the device state management table 70 shown in FIG. 3B, the state “power on-2CH selected” associated with the device ID “TV0001” is set to “power on-1CH selected”. Update to "Busy". As described above, the operation information receiving unit 51 changes the state of the TV 3 to “power-on-1CH channel selection” when the “up button pressed” event occurs in the state “power-on-2CH channel selection”. It can be determined based on the state transition information (not shown). Alternatively, when the numeric button is pressed when an arbitrary channel is selected on the TV 3, the operation information receiving unit 51 can determine that the numeric channel is being selected. Alternatively, when the OSD (On-Screen Display) image such as EPG (Electronic Program Guide) is displayed and the cross button and the determination button are pressed, the operation information receiving unit 51 selects the TV 3 in the above state. It is possible to determine that the selected channel is being selected.

When the device state is updated, the operation information generation unit 52 of the information collection server 4 updates the operation information based on the updated device state. In the first embodiment, when the state of the TV 3 indicates “power off”, the operation information generation unit 52 sets (generates or updates) the associated operation information to also indicate “power off”. On the other hand, when the state of the TV 3 indicates “power on” and “selecting (selecting any channel)”, the operation information generating unit 52 indicates “operation (viewing any channel) by the user”. Generate or update information. In addition, the operation information generation unit 52 generates or updates operation information indicating “non-viewing in use” when the state of the TV 3 indicates “power on” and “(any operation other than channel selection)”. To do.

  In the above-described specific example, the state of the device ID “TV0001” “power-on-2CH being selected” is updated to “power-on-1CH being selected”. Therefore, the operation information has been updated, but as a result remains “viewing”. Note that the operation information generation unit 52 may store the date and time when the operation information is generated or updated as shown in FIG. 3B in the last update date and time column.

  In another specific example, when the operation information receiving unit 51 updates the state “power off” of the device ID “TV0002” to “powering on—one channel is being selected”, the operation information generating unit 52 The operation information of the device ID “TV0002” is updated from “power off” to “viewing”.

  When the operation information is generated or updated by the operation information generation unit 52, the operation information transmission unit 53 transmits the latest operation information together with the associated device ID to the utterance control server 1. For example, FIG. 3C illustrates a specific example of the device ID and the operation information transmitted by the operation information transmission unit 53 when the operation information “viewing” of the device ID “TV0001” is updated.

  When the operation information receiving unit (operation information receiving unit) 54 of the utterance control server 1 receives the device ID and the operation information shown in (c) of FIG. 3, the device ID is transmitted to the control target specifying unit 55. It supplies to the condition judgment part 56, respectively.

  First, the control target specifying unit (control target specifying means) 55 is a dialog device 2 that is a pair of the TV 3 indicated by the device ID, using the supplied device ID, and is a dialog for which utterance control is performed. The device 2 is specified. According to the device device correspondence table 71 shown in FIG. 3D, the device ID of the interactive device 2 to be controlled is stored in association with the device ID. Therefore, the control target specifying unit 55 can specify the interactive device 2 to be controlled by referring to the device device correspondence table 71. In the above specific example, since the received device ID indicates “TV0001”, the control target specifying unit 55 specifies that the interactive device 2 to be controlled is the interactive device 2 having the device ID “DE0001”.

  Next, the condition determination unit (condition determination means) 56 uses the supplied operation information to determine whether or not a voice error recognition condition is satisfied in an environment where the interactive device 2 that is paired with the TV 3 is placed. Judging. According to the rule table 74 shown in (e) of FIG. 3, success / failure of the voice error recognition information is associated with each content of the operation information. Therefore, the condition determination unit 56 can determine whether or not the voice error recognition condition is successful based on the operation information by referring to the rule table 74. In the above specific example, since the received operation information indicates “viewing”, the condition determination unit 56 determines that the voice recognition error condition is satisfied in the environment where the interactive apparatus 2 is currently placed. In this way, since the TV 3 of “TV0001” is viewed by the user, the dialog system 401 grasps the situation that the dialog device 2 of “DE0001” is highly likely to cause erroneous recognition of voice.

Therefore, in this specific example, the utterance control unit (speech control means) 57 transmits an output suppression instruction as an instruction related to the utterance to the identified dialog device 2 of “DE0001”. The dialogue mode control unit 58 of the dialogue apparatus 2 that has received the output suppression instruction controls the voice recognition unit 59 or the voice output control unit 60 as described above in accordance with this instruction. As a result, even if some kind of sound is generated around the interactive device 2, the response voice of the sound is not output from the sound output unit 23 of the interactive device 2. In this way, it is possible to prevent the interactive device 2 from causing erroneous recognition of voice or malfunction under a situation where the interactive device 2 is likely to cause erroneous recognition of speech. In particular, in the above-described specific example, it is possible to avoid that the dialogue apparatus 2 outputs an erroneous response voice in response to the voice output from the TV 3 while the user is watching the TV 3. As a result, the user's viewing of the TV 3 can be prevented from being disturbed by a malfunction.

  The device device correspondence table 71 may be stored in the storage unit 41 of the information collection server 4. In this case, the control target specifying unit 55 is provided in the control unit 40, and the operation information transmitting unit 53 associates the device ID specified by the control target specifying unit 55 with the operation information and supplies it to the utterance control server 1. That's fine.

[Process flow]
FIG. 4 is a flowchart showing the flow of operation information transmission processing executed by the TV 3 and the flow of operation information transmission processing executed by the information collection server 4 in the interactive system 401. FIG. 5 is a flowchart showing the flow of the utterance control process executed by the utterance control server 1 and the flow of the dialogue mode control process executed by the dialog device 2 in the dialog system 401.

(Operation information transmission flow)
Referring to FIG. 4, when TV 3 receives a control signal sent from operation unit 32 by a user's remote control operation (YES in S101), operation information transmission unit 50 receives the device ID of the own device and the control signal. Operation information indicating the contents is transmitted to the information collection server 4 (S102) (for example, (a) in FIG. 3). Alternatively, a case where a certain time or more has passed without receiving the control signal since the TV 3 received the previous control signal (YES in S101) can be considered. In this case, instead of indicating the pressed button, the operation information transmitting unit 50 transmits operation information indicating that there is no pressed button to the information collecting server 4 (S102).

(Operation information transmission flow)
When the operation information receiving unit 51 of the information collection server 4 receives the device ID and the operation information (YES in S103), the operation information receiving unit 51 reads the device ID in the device state management table 70 ((b) of FIG. 3). The state of the TV 3 specified by is updated based on the operation information (S104).

  When the state of the TV 3 indicates that the power is on and the channel is being selected (YES in S105 and YES in S106), the operation information generation unit 52 generates the operation information “viewing” of the TV 3 Alternatively, the operation information is updated to “viewing” (S107). Then, the operation information transmission unit 53 transmits the device ID of the TV 3 and the operation information “viewing” to the utterance control server 1 (S108). On the other hand, when the status of TV3 indicates power-off or power-on other than during channel selection (NO in S105 or NO in S106), operation information generating unit 52 displays the operation information of TV3 as “power off”. "Or" Non-viewing in use "is generated or updated (S109). Then, the operation information transmission unit 53 transmits the device ID of the TV 3 and the operation information “power off” or “not in use” to the utterance control server 1 (S110).

  If the operation information receiving unit 51 has not received operation information for a certain time after receiving the previous operation information for one TV 3 (NO in S103, YES in S111), the state of the TV 3 May be updated to “power off”. In this case, S109 and S110 are executed.

Alternatively, the operation information receiving unit 51 may periodically check the last update date and time of the device state management table 70 shown in FIG. Then, the operation information receiving unit 51 extracts all TVs 3 that have passed a predetermined time or more from the last update date and time (or since the previous reception of the operation information), and sets the state of these TVs 3 to “power off”. Update. In this case, the operation information is updated to “power off” by the operation information generation unit 52 together with all the extracted TVs 3 and transmitted to the utterance control server 1 by the operation information transmission unit 53.

  Note that when the information collection server 4 holds the device device correspondence table 71, the control target specifying unit 55 provided in the information collection server 4 performs the device in any step after S103 and before S108 or S110. A step of specifying the device ID based on the ID is executed. In this case, in S108 or S110, the device ID is transmitted together with the operation information instead of the device ID.

(Speech control flow)
Referring to FIG. 5, when the operation information receiving unit 54 of the utterance control server 1 receives the device ID and the operation information (for example, (c) of FIG. 3) (YES in S112), the control target specifying unit 55 With reference to the device device correspondence table 71 ((d) of FIG. 3), the dialogue device 2 that is paired with the TV 3 and is an utterance control target is specified (S113). Specifically, the device ID corresponding to the received device ID is specified. If the device ID is received together with the operation information instead of the device ID, S113 can be omitted.

  On the other hand, the condition determination unit 56 refers to the rule table 74 ((e) of FIG. 3), and based on the received operation information, whether or not a voice error recognition condition is satisfied in the environment of the interactive device 2 described above. Is determined (S114). In the example shown in (e) of FIG. 3, the condition determination unit 56 determines that the voice recognition condition is satisfied when the operation information indicates “viewing”, and the operation information indicates “power off” or “non-viewing”. When “in use” is indicated, it is determined that the voice recognition condition is not satisfied.

  When it is determined that the erroneous voice recognition condition is satisfied (YES in S114), the utterance control unit 57 instructs the interactive device 2 specified in S113 to suppress the output of the reply voice, that is, the output suppression instruction. Is transmitted (S115). On the other hand, when it is determined that the voice recognition condition is not satisfied (NO in S114), the utterance control unit 57 instructs the interactive device 2 specified in S113 to cancel the output suppression of the response voice, that is, Then, a suppression release instruction is transmitted (S116).

(Interactive mode control flow)
When interactive mode control unit 58 of interactive apparatus 2 receives an instruction related to speech (YES in S117), interactive mode control unit 58 analyzes the content of the instruction. When the received instruction is an output suppression instruction (1 in S118), the dialogue mode control unit 58 switches the dialogue mode of the dialogue apparatus 2 to a dialogue mode that does not output a reply voice (S119). Specifically, the output of the response voice is suppressed by invalidating the voice recognition function of the voice recognition unit 59. On the other hand, when the received instruction is a suppression release instruction (2 in S118), the dialogue mode control unit 58 switches the dialogue mode to a dialogue mode that outputs a response voice (S120). Specifically, by suppressing the voice recognition function of the voice recognition unit 59, the output suppression of the reply voice is released.

  When the conversation mode control unit 58 analyzes in S118 that the instruction related to the utterance is a spontaneous voice output instruction, the voice output control is performed even if the current mode is the conversation mode in which no response voice is output. The unit 60 may be instructed to output a spontaneous sound.

<< Embodiment 2 >>
The second embodiment of the present invention will be described below with reference to FIGS. For convenience of explanation, members having the same functions as those described in the previous embodiment are denoted by the same reference numerals and description thereof is omitted. The same applies to the following embodiments.

  In the second embodiment, the audio output device (TV 3) is not connected to an external network such as the Internet and does not have a function of communicating with an external device. That is, the audio output device cannot transmit its own state to the outside. As described above, even when the voice output device is not connected to the network, the speech control server 1 can grasp the status of the voice output device and the environment of the interactive device 2 and appropriately avoid the erroneous recognition of the voice. Is possible.

[Outline of Dialogue System]
FIG. 6 is a diagram showing an outline of a dialogue system according to Embodiment 2 of the present invention. The dialogue system 402 shown in FIG. 6 includes the utterance control server 1, the dialogue device 2, and an audio output device (for example, TV 3).

  The operation information is configured to be generated by the information collection server 4 and supplied to the utterance control server 1 in the interactive system 401 of the first embodiment, but the information collection server 4 is not provided in the interactive system 402 of the second embodiment. Instead of the information collection server 4, the dialogue apparatus 2 supplies the utterance control server 1.

  In the second embodiment, the dialogue apparatus 2 has a function of communicating with the TV 3 via infrared rays, and similarly to the operation unit 32, sends a control signal to operate the TV 3 remotely and grasps the state of the TV 3. Is possible. Thereby, since the dialogue apparatus 2 grasps the operation information received by the TV 3 and the latest state of the TV 3 by itself, it can generate the operation information based on these information. That is, the dialogue apparatus 2 can transmit the apparatus ID of the own apparatus and the operation information generated by the own apparatus to the utterance control server 1. The utterance control server 1 can return an instruction related to the utterance to the dialogue apparatus 2 in accordance with the received operation information.

  As a result, while the TV 3 is being viewed by the user, it is possible to obtain the same effect as in the first embodiment, in which it is possible to avoid erroneous voice recognition or malfunction associated therewith.

[Functional configuration of each device in the interactive system]
FIG. 7 is a functional block showing a main configuration of each device in the dialogue system according to Embodiment 2 of the present invention. 8A to 8D are diagrams illustrating specific examples of information and tables handled by each device in the interactive system 402 according to the second embodiment.

(Main components of the utterance control server 1)
The utterance control server 1 shown in FIG. 7 is different from the utterance control server 1 of the first embodiment shown in FIG. 1 in the following points.

  In the second embodiment, the operation information receiving unit 54 receives the device ID together with the operation information instead of the device ID. Therefore, the utterance control server 1 can directly grasp the interactive device 2 to be controlled from the device ID. Therefore, the control unit 10 may not include the control target specifying unit 55. The storage unit 11 may not store the device device correspondence table 71. The operation information reception unit 54 supplies the received device ID to the utterance control unit 57 and transmits the received operation information to the condition determination unit 56.

(Main part configuration of the dialogue device 2)
The interactive device 2 shown in FIG. 7 is different from the interactive device 2 of the first embodiment shown in FIG. 1 in the following points.

The dialogue apparatus 2 includes an infrared transmission unit 24 that transmits a control signal for operating the TV 3. The control unit 20 further includes a device operation unit (device operation means) 61 as a functional block,
An operation information generation unit 52 and an operation information transmission unit 53 are included.

  The storage unit 21 stores a device state management table 70 and a device ID 72. The device ID 72 is identification information individually assigned to the interactive device 2 as shown in FIG. The device ID 72 is used when the operation information transmission unit 53 transmits operation information.

  The device operation unit 61 is for remotely operating the TV 3 in response to the occurrence of a predetermined event (a predetermined time is reached or the user instructs it). The device operation unit 61 is a method similar to that of the operation information receiving unit 51 of the first embodiment, and the state of the TV 3 stored in the device state management table 70 according to the operation information indicating the content of the remote operation described above. Update.

  For example, it is assumed that the user is instructed to switch the channel of the TV 3 from channel 2 to channel 1. In this case, the device operation unit 61 controls the infrared transmission unit 24 with a control signal corresponding to, for example, pressing the upper button of the remote control so that the state of the TV 3 transitions from channel selection to channel selection. Then, it is sent to the TV 3. The device operation unit 61 generates operation information indicating that the control signal (pressing the upper button) is sent to the TV 3, and based on this, the state of the TV 3 stored in the device state management table 70 is displayed. Update. Specifically, the device operation unit 61 first generates operation information shown in FIG. Then, the record of the operation target TV 3 (device ID “TV0001”) is read out from the device state management table 70 shown in FIG. And based on the said operation information, the state of TV3 is updated by the method similar to the operation information receiving part 51 of Embodiment 1. FIG.

  The operation information generation unit 52 generates or updates the operation information, similarly to the operation information generation unit 52 of the first embodiment. Although not shown, the date and time when the operation information is generated or updated as in the first embodiment may be stored in the device status management table 70 as the last update date and time. Note that the device state management table 70 shown in FIG. 8C is a table for managing the state of one or more audio output devices that can be remotely operated by the interactive apparatus 2. Therefore, in addition to the TV 3, records such as a recording / playback device and a music playback device that can be remotely operated by the interactive device 2 may be managed in the device state management table 70.

  The operation information transmission unit 53 transmits the operation information generated or updated by the operation information generation unit 52 to the utterance control server 1, similarly to the operation information transmission unit 53 of the first embodiment. However, in the second embodiment, not the device ID but the device ID 72 that is identification information of the own device is transmitted in association with the operation information. FIG. 8D shows a specific example of the device ID and operation information transmitted by the operation information transmission unit 53.

(Main components of audio output equipment)
In the second embodiment, the TV 3 may not include the operation information transmission unit 50 in the communication unit and the control unit 30. The TV 3 includes at least an infrared receiving unit 33 for receiving a control signal transmitted from the interactive apparatus 2. The TV 3 includes various units for executing functions as a digital television in accordance with the control signal received by the infrared receiver 33, but the illustration of these blocks is omitted.

  According to the above configuration, when the operating status of the audio output device (TV 3 or the like) changes due to remote operation of the dialog device 2, the dialog device 2 transmits the change information to the utterance control server 1 To report to. As in the first embodiment, the utterance control server 1 determines the success or failure of the speech misrecognition condition based on the operation information, and outputs an output suppression instruction or a suppression release instruction to the report source dialog apparatus 2 according to the determination result. Reply to.

  As described above, even when the voice output device is not connected to the network, the speech control server 1 can grasp the status of the voice output device and the environment of the interactive device 2 and appropriately avoid the erroneous recognition of the voice. Is possible.

[Process flow]
FIG. 9 is a flowchart showing the flow of the operation information transmission process executed by the dialogue apparatus 2 and the flow of the utterance control process executed by the utterance control server 1 in the dialogue system 402.

(Operation information transmission flow)
When the device operation unit 61 of the interactive apparatus 2 detects the occurrence of a predetermined event for remotely operating the audio output device (herein referred to as TV 3) (YES in S201), the device operation unit 61 transmits the infrared transmission unit 24. To control the TV 3 remotely. And based on the operation information which shows the operation content, the state of TV3 is updated in the apparatus state management table 70 (S202).

  When the state of the TV 3 indicates that the power is on and the channel is being selected (YES in S203 and YES in S204), the operation information generation unit 52 generates the operation information “viewing” of the TV 3 Alternatively, the operation information is updated to “viewing” (S205). Then, the operation information transmission unit 53 transmits the device ID 72 and the operation information “viewing” to the utterance control server 1 (S206). On the other hand, when the status of TV3 indicates power-off or power-on other than during channel selection (NO in S203 or NO in S204), operation information generating unit 52 displays the operation information of TV3 as “power off”. "Or" Non-viewing in use "is generated or updated (S207). Then, the operation information transmission unit 53 transmits the device ID 72 and the operation information “power off” or “not in use” to the utterance control server 1 (S208).

(Speech control flow)
When the operation information receiving unit 54 of the utterance control server 1 receives the device ID 72 and the operation information (YES in S209), the operation information receiving unit 54 supplies the device ID 72 to the utterance control unit 57 and uses the operation information as a condition determination unit. 56. The subsequent processing is the same as S114 to S116 shown in FIG.

(Interactive mode control flow)
This is the same as S117 to S120 shown in FIG.

<< Embodiment 3 >>
The third embodiment of the present invention will be described below with reference to FIGS.

  The cause of causing an environment in which erroneous recognition may occur to the dialogue apparatus 2 is not limited to the audio output device such as the TV 3. Even in an environment where the user is speaking to another object, not to the interactive device, the interactive device 2 may cause erroneous recognition and erroneous operation. For example, it is conceivable that the user speaks to a remote call partner using a call device such as a telephone, a mobile phone, a smartphone, or an interphone. This utterance is obviously not directed to the dialogue apparatus 2, but the dialogue apparatus 2 near the user may misrecognize the voice of the utterance. In the third embodiment, the interactive system of the present invention can grasp an environment in which a user speaks by using a telephone device, and can appropriately avoid voice misrecognition.

[Outline of Dialogue System]
FIG. 10 is a diagram showing an outline of a dialogue system according to Embodiment 3 of the present invention. The dialogue system 403 shown in FIG. 10 includes the utterance control server 1, the dialogue device 2, and one or more telephone devices (for example, the telephone 3a).

  In the third embodiment, the call device has a function of outputting a predetermined ringing tone and notifying the user of an incoming call when an incoming call is received from the device of the other party. Thereafter, the user who notices the incoming call is expected to speak to the other party using the telephone device.

  The dialog device 2 detects a specific predetermined ringing tone for each telephone device (telephone 3a), thereby identifying the telephone (non-target audio source device / calling device) 3a, and the telephone 3a is busy Can be grasped. Based on the fact that the telephone 3a is in a call state, the operation information of the telephone 3a can be generated. Thereafter, as in the second embodiment, the dialogue apparatus 2 may transmit the apparatus ID of the own apparatus and the operation information of the telephone 3 a to the utterance control server 1.

  As a result, while the telephone 3a is in a call state, it is possible to obtain an effect that it is possible to avoid erroneous voice recognition or a malfunction associated therewith.

[Functional configuration of each device in the interactive system]
FIG. 11 is a functional block showing a main configuration of each device in the dialogue system according to Embodiment 3 of the present invention. (A)-(d) of FIG. 12 is a figure which shows the specific example of the information and table which each apparatus handles in the interactive system 403 of Embodiment 3. FIG.

(Main components of the utterance control server 1)
The utterance control server 1 shown in FIG. 11 is different from the utterance control server 1 of the second embodiment shown in FIG. 7 in the following points.

  In the third embodiment, the operation information receiving unit 54 receives the operation information (for example, (d) of FIG. 12) of the telephone device (telephone 3a) together with the device ID 72. In the third embodiment, the operation information indicates “busy” (a state in which the user uses the telephone 3a and the user is talking to the other party) or “standby” (a state in which an incoming call is awaited).

  The condition determining unit 56 refers to the rule table 74 and determines whether or not a voice error recognition condition is satisfied in the environment where the interactive device 2 is placed, based on the received operation information. For example, according to the rule table 74 shown in FIG. 12E, the condition determination unit 56 determines that the voice recognition error condition is satisfied when the operation information of the telephone 3a indicates “busy”, and the operation is performed. When the information indicates “standby”, it is determined that the voice recognition condition is not satisfied.

(Main part configuration of the dialogue device 2)
The dialogue apparatus 2 shown in FIG. 11 is different from the dialogue apparatus 2 according to the second embodiment shown in FIG. 2 in the following points.

  The dialogue apparatus 2 may not include the infrared transmission unit 24.

  Moreover, the control part 20 does not need to include the apparatus operation part 61, and is provided with the sound determination part (sound determination means) 62 as a functional block instead.

  The storage unit 21 further stores a ringing tone table 73.

The sound determination unit 62 analyzes the sound data input to the control unit 20 via the sound input unit 22 and determines what sound it is. Further, the sound determination unit 62 updates the state of each telephone device managed in the device state management table 70 (for example, (c) of FIG. 12) according to the determination result.

  FIG. 12B is a diagram showing a specific example of the ringing tone table 73. In the third embodiment, the sound determination unit 62 compares the ringing sound sample for each telephone device registered in the ringing sound table 73 in advance with the input voice data, and the sound of the voice data is registered. If it is the ringing sound of the telephone device that is being used, the ringing sound is identified from the samples in the ringing sound table 73. When there is only one telephone device 3a, the sound determination unit 62 only needs to determine whether the sound is a ringing sound of the telephone 3a. When the ring determination sound is specified, the sound determination unit 62 can specify a call device that has received an incoming call based on the device ID associated with the ringing sound. Then, the sound determination unit 62 updates the state of the call device associated with the device ID stored in the device state management table 70. Specifically, when the sound determination unit 62 determines that the ringing tone of the telephone 3a is sounded at 17:30, the state of the device ID “PH1001” in the device state management table 70 of FIG. Update to “17:30 Incoming Call”.

  Furthermore, the sound determination unit 62 compares the sound data of the input voice data by comparing a sample of human voice sound (not shown) stored in the storage unit 21 with the input sound data. It is preferable to determine whether or not is a human voice. The sound determination unit 62 only needs to determine whether or not it is a human voice. Compared with the voice recognition unit 59 that converts the utterance contents into text, the configuration can be simplified and the processing load is also reduced. The The human voice sample may be prepared in advance, or may be a registered voice of the user of the dialogue apparatus 2.

  Specifically, the sound determination unit 62 monitors the input of a human voice after determining that the ringing sound of the registered telephone device has sounded. For example, after determining that the ringing tone of the interphone (not shown) having the device ID “PH1002” is sounded at 15:40, the input of a human voice is monitored. Thereafter, when it is confirmed that the input of the person's voice has been interrupted for a predetermined time or more, it is determined that the call started with the ringing tone as a trigger is terminated (for example, terminated at 15:42). At this time, as shown in (c) of FIG. 12, the sound determination unit 62 updates the interphone state of the device ID “PH1002” from “15:40 received” to “15:42 call ended”. With the above configuration, the interactive device 2 can grasp the period during which the telephone device is in a telephone conversation.

  The operation information generation unit 52 generates or updates the operation information of the calling device. Although not shown, the date and time when the operation information is generated or updated as in the first embodiment may be stored in the device status management table 70 as the last update date and time. The operation information generation unit 52 updates the operation information from “standby” to “busy” when the state of the call device is updated from “call end” to “call received”. In addition, when the state of the calling device is updated from “present incoming call” to “end call”, the operation information is updated from “busy” to “standby”.

  Similarly to the operation information transmission unit 53 of the second embodiment, the operation information transmission unit 53 generates the operation information generated or updated by the operation information generation unit 52 together with the device ID 72 (for example, (a) in FIG. 12) and the utterance control server. 1 to send. FIG. 12D shows a specific example of the device ID and operation information transmitted by the operation information transmission unit 53.

(Main components of telephone equipment)
The telephone 3a as a telephone device only needs to include each unit (at least a voice output unit for notifying a user of an incoming call) for executing a general function as a telephone, and is not illustrated.

When a call device (such as the telephone 3a) around the interactive device 2 receives an incoming call and outputs a ringing tone, the user then speaks to the other party, thereby providing an environment in which voice misrecognition is likely to occur. It is. According to the above configuration, when the call device responds to an incoming call with a ringing tone as a trigger, and the operating status of the call device changes, the interactive device 2 detects the change by collecting the ringing sound. Then, the changed operation information is generated and transmitted to the utterance control server 1, and the change is reported to the utterance control server 1. Furthermore, the dialogue apparatus 2 monitors the input of the voice of the person (user) collected after the ringing tone is heard, and when the input of this voice is interrupted for a certain time or longer, the dialogue apparatus 2 is started with the ringing tone as a trigger. It can be determined that the call has ended. Then, operation information indicating the end of the call is generated, transmitted to the utterance control server 1, and reported to the utterance control server 1. As in the second embodiment, the utterance control server 1 determines the success or failure of the voice recognition condition based on the operation information, and outputs an output suppression instruction or a suppression cancellation instruction to the reporting source interactive apparatus 2 according to the determination result. Reply to.

  As described above, the utterance control server 1 can grasp the environment in which the user utters a target other than the interactive device 2 based on the situation of the telephone device. That is, it is possible to appropriately avoid misrecognition of voice during a period in which the call device is in a call state.

[Process flow]
FIG. 13 is a flowchart showing the flow of the operation information transmission process executed by the dialogue apparatus 2 and the flow of the utterance control process executed by the utterance control server 1 in the dialogue system 403.

(Operation information transmission flow)
When the sound determination unit 62 of the interactive device 2 detects that the calling device has made a ringing sound (YES in S301), the sound determination unit 62 continues in the ringing sound table 73 ((b) of FIG. 12). Based on the device ID associated with the detected ringing tone, it is specified which calling device has received the incoming call (S302). Here, it is assumed that the telephone 3a is specified. And the sound determination part 62 updates the state of the telephone 3a stored in the apparatus state management table 70 ((c) of FIG. 12) (S303).

  When the state of the telephone 3a indicates that there is an incoming call (YES in S304), the operation information generating unit 52 generates the operation information “busy” of the telephone 3a or updates the operation information to “busy”. (S305). Then, the operation information transmission unit 53 transmits the device ID 72 ((a) in FIG. 12) and the operation information “busy” ((d) in FIG. 12) to the utterance control server 1 (S306). On the other hand, when the state of the telephone 3a indicates the end of the call (NO in S304), the operation information generating unit 52 generates or updates the operation information of the telephone 3a so as to indicate "standby" (S307). . Then, the operation information transmission unit 53 transmits the device ID 72 and the operation information “standby” to the utterance control server 1 (S308).

  Although not shown, the sound determination unit 62 monitors the input of a human voice after the step of S306 and before the step of S309. When sound determination unit 62 detects that the input of human voice has been interrupted for a certain period of time or longer (NO in S301, YES in S309), device status management table for telephone 3a that has been updated to a state with an incoming call immediately before. The state of 70 is updated to “end of call” (S303). Then, the subsequent processes (S304 to 308) are repeated.

(Speech control flow)
When the operation information receiving unit 54 of the utterance control server 1 receives the device ID 72 and the operation information (YES in S309), the operation information receiving unit 54 supplies the device ID 72 to the utterance control unit 57 and uses the operation information as a condition determination unit. 56. The subsequent processing is the same as S114 to S116 shown in FIG. However, the condition determination unit 56 is different from the second embodiment in that it refers to the rule table 74 shown in FIG.

(Interactive mode control flow)
This is the same as S117 to S120 shown in FIG.

<< Embodiment 4 >>
The following description will discuss Embodiment 4 of the present invention with reference to FIGS.

  The cause of causing an environment in which erroneous recognition may occur to the dialogue apparatus 2 is not limited to the audio output device such as the TV 3. In recent years, household electrical appliances (hereinafter referred to as home appliances) having a function of outputting voice guidance have been widely used. For example, home appliances such as washing machines, rice cookers, microwave ovens, and water heaters output voice guidance to users in various processes during operation (for example, the washing machine "starts washing" “Rinse completed”, “Dehydrating, ca n’t open the lid”, etc.) Even in an environment in which such voice guidance is output from the home appliance, the interactive device 2 may cause erroneous recognition and erroneous operation. In Embodiment 4, the dialogue system of the present invention can grasp the environment in which voice guidance is output while the home appliance is in operation, and can appropriately avoid voice misrecognition.

[Outline of Dialogue System]
As an example of the dialogue system 404 according to the fourth embodiment of the present invention, a configuration substantially the same as that of the dialogue system 402 according to the second embodiment shown in FIG. 6 is employed. Alternatively, the interactive system 401 according to the first embodiment illustrated in FIG. 2 may be employed.

  The dialog system 404 of the fourth embodiment is different from the dialog system 402 of the second embodiment shown in FIG. 6 in the following points.

  In the fourth embodiment, the dialogue system 404 includes one or more home appliances (for example, a washing machine, a rice cooker, a microwave oven, etc.) instead of the audio output device (TV3) as the non-target audio generation source device.

  In the fourth embodiment, the dialogue apparatus 2 is configured to operate these home appliances remotely and supply operation information of each home appliance to the utterance control server 1.

  As a result, it is possible to obtain an effect that it is possible to avoid voice misrecognition or accompanying malfunction while the home appliance is in operation.

[Functional configuration of each device in the interactive system]
(A)-(d) of FIG. 14 is a figure which shows the specific example of the information and table which each apparatus handles in the interactive system 404 of Embodiment 4. FIG.

(Main components of the utterance control server 1)
The utterance control server 1 of the fourth embodiment is different from the utterance control server 1 shown in FIG. 7 in the following points.

  In the fourth embodiment, the operation information receiving unit 54 receives operation information (for example, (c) of FIG. 14) of home appliances (for example, a washing machine). In the fourth embodiment, the operation information indicates “in operation”, “standby”, or “power off”, as will be described in detail later.

The storage unit 11 stores a rule table 74 shown in FIG. The condition determining unit 56 refers to the rule table 74 and determines whether or not a voice error recognition condition is satisfied in the environment where the interactive apparatus 2 is placed, based on the received operation information. For example, according to the rule table 74 shown in FIG. 14D, the condition determination unit 56 determines that the voice recognition error condition is satisfied when the operation information of the washing machine indicates “in operation”, and the operation is performed. When the information indicates “standby” or “power off”, it is determined that the voice recognition condition is not satisfied.

(Main part configuration of the dialogue device 2)
The interaction device 2 of the fourth embodiment is different from the interaction device 2 of the second embodiment shown in FIG. 7 in the following points.

  The device operation unit 61 is for remotely operating home appliances. Specifically, the device operation unit 61 controls the power on / off of the home appliance to be operated, or instructs the home appliance to execute a predetermined function when the home appliance is powered on. To do. In the fourth embodiment, when operating the home appliance, the device operation unit 61 supplies the operation information generating unit 52 with the device ID and operation information of the operation target home appliance. In the fourth embodiment, the operation information indicates “power on” when the device operation unit 61 switches the power of the home appliance from off to on, and indicates “power off” when the power is switched from on to off. When the home appliance is instructed to execute a predetermined function, “execution instruction” is indicated. For example, when the device operation unit 61 instructs the washing machine to start washing, a specific example of the device ID and operation information supplied from the device operation unit 61 to the operation information generation unit 52 is shown in FIG. a).

  Based on the device ID and operation information supplied from the device operation unit 61, the operation information generation unit 52 generates or updates the operation information of the home appliance in the device state management table 70 (for example, FIG. 14B). To do. In the above example, the operation information of the washing machine with the device ID “WA0001” indicates “execution instruction”. That is, since the washing machine has started washing in accordance with the instruction from the interactive device 2, the operation information generation unit 52 updates the operation information of the washing machine to “in operation”. When the operation information indicates “power on”, the washing machine is not yet in operation, but transitions to a state in which washing can be started at any time when an instruction from the user or the interactive device 2 is received. Therefore, the operation information generation unit 52 updates the operation information to “standby”. When the operation information indicates “power off”, since the washing machine is not in an operable state, the operation information generation unit 52 updates the operation information to “power off”. Although not shown, the date and time when the operation information is generated or updated as in the first embodiment may be stored in the device status management table 70 as the last update date and time.

  The operation information transmission unit 53 transmits the operation information generated or updated by the operation information generation unit 52 to the utterance control server 1. FIG. 14C illustrates a specific example of the device ID and the operation information transmitted by the operation information transmission unit 53.

(Main components of home appliances)
7, in addition to the blocks shown in FIG. 7, each home appliance (such as a washing machine) replacing the TV 3 performs a general function as each home appliance (at least in a predetermined process, voice). A voice output unit for outputting guidance) is provided, but this block is not shown.

According to the above configuration, when the operation status of the household electrical appliance (such as a washing machine) is changed by remote operation of the dialog device 2, the dialog device 2 transmits the change to the utterance control server 1. To report to. The utterance control server 1 determines the success / failure of the voice misrecognition condition based on the operation information, and returns an output suppression instruction or a suppression release instruction to the reporter interactive apparatus 2 according to the determination result. Specifically, the utterance control server 1 does not output a response voice to the dialogue device 2 when the operation information indicates “in operation” and the home appliance is in a state of outputting some voice guidance. To instruct. On the other hand, when the operation information indicates “standby” or “power off” and the home appliance is not in a state of outputting voice guidance, the interactive device 2 is instructed to output a response voice.

  As described above, even when the home appliance that outputs the voice guidance is in the vicinity of the dialogue device 2, the utterance control server 1 grasps the situation of the home appliance and the environment of the dialogue device 2, and recognizes erroneous voice recognition. It is possible to avoid it appropriately.

  In the above-described example, the dialog system 404 employs a configuration in which the dialog device 2 remotely operates the home appliance, but is not limited thereto. In the fourth embodiment, the dialog device 2 and the home appliance are connected to each other via a local network such as a LAN (Local Area Network) so that the home appliance can turn on the power and start executing the function. 2 can also be adopted as the interactive system 404 of the fourth embodiment. In this case, power-off is determined by the dialog device 2 not receiving the report for a certain period of time.

[Process flow]
The processing flow of each device in the interactive system 404 of the fourth embodiment will be described below with reference to FIGS. 9 and 5.

(Operation information transmission flow)
As illustrated in FIG. 9, when the device operation unit 61 of the interactive apparatus 2 detects the occurrence of a predetermined event for remotely operating the home appliance (here, a washing machine) (YES in S201), the device operation unit 61 Controls the infrared transmitter 24 to remotely operate the washing machine. Then, the operation information generation unit 52 is supplied with the operation target device ID and operation information indicating the operation content (S202). In the fourth embodiment, since the state of the home appliance is not stored in the device state management table 70, the device state management table 70 is not updated here.

  In the fourth embodiment, step S203 is omitted. Then, instead of the step of S204, a step of determining whether or not the operation information is an “execution instruction” (not shown; S204 ′) is executed. When the operation information of the washing machine indicates an execution instruction (YES in S204 ′), the operation information generation unit 52 generates the operation information “in operation” of the washing machine or sets the operation information to “in operation”. (S205). Then, the operation information transmission unit 53 transmits the device ID 72 and the operation information (“in operation” in the fourth embodiment) to the utterance control server 1 (S206). On the other hand, when the operation information of the washing machine indicates power on or power off (NO in S204 ′), the operation information generation unit 52 sets the operation information of the washing machine as “standby” or “power off”. Generate or update as shown (S207). Then, the operation information transmission unit 53 transmits the device ID 72 and the operation information (in the fourth embodiment, “standby” or “power off”) to the utterance control server 1 (S208).

(Speech control flow)
When the operation information receiving unit 54 of the utterance control server 1 receives the device ID 72 and the operation information (YES in S209), the operation information receiving unit 54 supplies the device ID 72 to the utterance control unit 57 and uses the operation information as a condition determination unit. 56. The subsequent processing is the same as S114 to S116 shown in FIG. However, the condition determination unit 56 is different from the second embodiment in that it refers to the rule table 74 shown in FIG.

(Interactive mode control flow)
This is the same as S117 to S120 shown in FIG.

<< Embodiment 5 >>
[Example of software implementation]
Control blocks of the utterance control server 1, the interactive device 2, the TV 3, and the information collection server 4 (
In particular, the control units 10, 20, 30 and 40) may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or by software using a CPU (Central Processing Unit). It may be realized.

  In the latter case, the utterance control server 1, the interactive device 2, the TV 3, and the information collection server 4 can be configured using a computer (electronic computer) as shown in FIG. FIG. 17 is a block diagram illustrating the configuration of a computer 100 that can be used as the utterance control server 1, the interactive device 2, the TV 3, or the information collection server 4.

  As shown in FIG. 17, the computer 100 includes an arithmetic device 120, a main storage device 130, an auxiliary storage device 140, and an input / output interface 150 connected to each other via a bus 110. The arithmetic device 120, the main storage device 130, and the auxiliary storage device 140 may be, for example, a CPU, a random access memory (RAM), and a hard disk drive, respectively. The main storage device 130 may be a computer-readable “non-temporary tangible medium”. For example, a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like can be used.

  The input device 200 and the output device 300 are connected to the input / output interface 150. The input device 200 and the output device 300 of each server receive data transmitted from another server, the TV 3 or the interactive device 2, and transmit data to the other server, the TV 3 or the interactive device 2. The input device 200 and the output device 300 of the dialogue apparatus 2 receive data from the utterance control server 1 and acquire a user's voice, transmit data to the utterance control server 1, and utter the user. The input device 200 and the output device 300 of the TV 3 acquire an operation instruction from the user, receive data from each server or interactive device 2, and transmit data from each server or interactive device 2.

  The auxiliary storage device 140 stores programs for operating the computer 100 as the utterance control server 1, the interactive device 2, the TV 3, or the information collection server 4. Then, the arithmetic device 120 expands the respective programs stored in the auxiliary storage device 140 on the main storage device 130, and executes instructions included in the respective programs expanded on the main storage device 130. The computer 100 is caused to function as each unit included in the utterance control server 1, the interactive device 2, the TV 3, or the information collection server 4. Further, the auxiliary storage device 140 may be a ROM (Read Only Memory) in which the above programs and various data are recorded so as to be readable by the arithmetic device 120 (or CPU).

  Here, the configuration for causing the computer 100 to function using each of the programs recorded in the auxiliary storage device 140, which is an internal recording medium, has been described, but a program recorded on an external recording medium may be used. The program may be supplied to the computer via an arbitrary transmission medium (such as a communication network or a broadcast wave) that can transmit the program. The present invention can also be realized in the form of a data signal embedded in a carrier wave in which the program is embodied by electronic transmission.

<< Modification 1 >>
The dialogue mode control unit 58 of the dialogue device 2 of each embodiment outputs a spontaneous voice from the utterance control server 1 during the dialogue mode in which no reply voice is output, that is, while the voice recognition function is disabled. When the output instruction is received, the voice output control unit 60 may be controlled to output the designated spontaneous voice. In this case, it is preferable that the dialogue mode control unit 58 cancels the output suppression of the response voice only for a certain period after the output of the spontaneous voice, even in a period during which the output of the response voice is suppressed.

  When the dialogue apparatus 2 speaks spontaneously, there is a possibility that the user may return some response to the dialogue apparatus 2 in response to the said speech even while watching the TV 3. It is unnatural as a dialogue that the dialogue device 2 does not respond to the response. According to the above configuration, since the output suppression of the response voice is canceled for a certain period after the spontaneous voice is output, the dialogue apparatus 2 can further output the response voice in response to the response. As a result, it is possible to eliminate the unnaturalness described above.

  As a use case in which a spontaneous voice output instruction is transmitted from the utterance control server 1, for example, the following is assumed. The user can perform timer setting on the speech control server 1 in advance so that the dialog device 2 speaks predetermined content at a predetermined time. Specifically, it is assumed that the user has set in advance in the utterance control server 1 so that the dialogue apparatus 2 speaks about the weather of the next day when a predetermined time (for example, 20:00) comes. In this case, the utterance control server 1 obtains the weather information of the next day from the external information providing server according to the above settings, generates a spontaneous speech “It is sunny tomorrow”, and outputs it at 20:00. To the interactive device 2.

  The voice output control unit 60 of the dialogue apparatus 2 outputs a spontaneous voice “It is sunny tomorrow”. Here, the dialogue mode control unit 58 releases the output suppression of the response voice for a certain period (for example, 5 seconds). In response to the spontaneous voice, the user speaks, for example, “I understand, thank you”. Since the output suppression is released, the voice recognition unit 59 can recognize the user's speech as a voice. For example, the voice output control unit 60 can output “you are welcome” as a response voice to the user's utterance. The dialog mode control unit 58 returns the dialog mode of the dialog apparatus 2 to the dialog mode in which no response voice is output again after 5 seconds have elapsed since the spontaneous voice was output. That is, the voice recognition function of the voice recognition unit 59 is disabled.

<< Modification 2 >>
In each embodiment, the determination as to whether or not the voice recognition condition is successful is performed by the condition determination unit 56 provided in the utterance control server 1. However, the interactive system of the present invention is not limited to the above configuration. The dialog system of the present invention can employ a configuration in which the dialog device 2 determines whether or not the voice recognition error condition is successful by providing the operation information reception unit 54 and the condition determination unit 56 in the dialog device 2. In this case, device operation information necessary for success / failure determination is supplied from the information collection server 4 to the dialogue device 2 via the utterance control server 1 in the second modification of the first embodiment. Alternatively, the dialogue apparatus 2 is acquired by making a direct request to the information collection server 4. Further, in the second modification of the second and fourth embodiments, the interactive device 2 communicates with each device and directly obtains operation information from each device, or grasps the state of each device through the communication and is based on it. The dialogue device 2 itself may generate the operation information. Further, in the second modification of the third embodiment, the interactive device 2 may monitor the state of each device with various sensors, grasp the state of each device, and generate operation information based thereon. Then, in response to the determination of whether or not the erroneous speech recognition condition is satisfied by the condition determination unit 56, the dialogue mode control unit 58 controls at least one of the speech recognition unit 59 and the voice output control unit 60 to switch the dialogue mode. Do.

  Furthermore, in the modified example 2, it is preferable that the dialogue apparatus 2 periodically obtains the operation information at regular time intervals, and determines whether or not the voice recognition condition is successful each time. According to the above configuration, the operation information supplied from the device on the operation information supply side (the speech control server 1, the information collection server 4, or each device such as a voice output device, a telephone device, and a home appliance) is a communication error. Even if the communication device 2 is stuck, the interactive device 2 makes a voluntary request to always accurately determine the environment in which the interactive device 2 is placed from the latest operating information of each device, Output can be controlled.

  FIG. 15 and FIG. 16 are flowcharts showing the processing flow of each device of the interactive system in the second modification.

  Referring to FIG. 15, when the operation information receiving unit 54 of the dialog device 2 determines that a certain time has elapsed from the time when the previous operation information is acquired for the device paired with the dialog device 2 (YES in S401). The operation information receiving unit 54 requests the utterance control server 1 for operation information (S402). Specifically, the operation information receiving unit 54 generates an operation information request including the device ID of the interactive device 2 and transmits it to the utterance control server 1.

  In the second modification, the utterance control server 1 executes an operation information proxy acquisition process for acquiring operation information from the information collection server 4 as a proxy. When the utterance control server 1 receives the operation information request (YES in S403), the control target specifying unit 55 refers to the device / device correspondence table 71 and specifies the TV 3 corresponding to the received device ID (S404). . Then, the operation information receiving unit 54 of the utterance control server 1 generates an operation information request including the identified device ID of the TV 3 and transmits it to the information collection server 4 (S405).

  When the information collection server 4 receives the operation information request (YES in S406), the operation information generation unit 52 specifies which device operation information should be output based on the received device ID (S407). For example, the operation information generation unit 52 determines that operation information of the TV 3 is requested. The operation information generation unit 52 acquires the operation information of the TV 3 from the device state management table 70 ((b) of FIG. 3) (S408). Here, it is assumed that the operation information generation unit 52 determines that the last update date and time associated with the operation information to be acquired is information that is older than a predetermined time compared to the current time. In this case, the operation information generation unit 52 may acquire an error message indicating that the operation information is old and invalid, instead of the operation information. Alternatively, when the operation information is old, the operation information receiving unit 51 requests the latest operation information from the TV 3, updates the state of the TV 3 to the latest state based on the operation information, and the operation information generation unit 52 The operating state may be updated based on the latest state.

  The operation information transmission unit 53 sends an operation information response including the operation information (or error message) acquired in S408 and the device ID received in S406 to the utterance control server 1 as a response to the operation information request. Return (S409).

  The operation information receiving unit 54 of the utterance control server 1 receives the operation information response (S410). The control target specifying unit 55 acquires the device ID corresponding to the received device ID from the device device correspondence table 71, and specifies the interactive device 2 to which the operation information response is to be returned (S411). The operation information transmission unit of the utterance control server 1 (not shown) returns an operation information response including operation information or an error message to the specified dialog device 2 as a response to the operation information request received in S403 (S412). ).

  Referring to FIG. 16, the operation information receiving unit 54 of the dialogue apparatus 2 receives the operation information response (S413).

If the operation information response includes operation information indicating “viewing” (1 in S414), the condition determination unit 56 of the interactive device 2 uses the voice for the environment in which the device is placed. It is determined that the misrecognition condition is satisfied. Then, the dialogue mode control unit 58 switches the dialogue mode to a dialogue mode that does not output a reply voice (S415). If the operation information response includes operation information indicating “power off or not in use” (2 in S414), the voice recognition condition is satisfied for the environment in which the device is located. Judge that it is not. Then, the dialogue mode control unit 58 switches to the dialogue mode that outputs a response voice (S416). Alternatively, when the operation information response includes the error message (3 in S414), the dialogue mode control unit 58 switches to the default dialogue mode (S417). The default interactive mode may be an interactive mode that outputs a response voice.

<< Modification 3 >>
In each of the above-described embodiments, the dialogue apparatus 2 includes the voice recognition unit 59, recognizes the voice input to the own apparatus, and generates the reply voice generated by the own apparatus or the reply supplied from the utterance control server 1. It was the structure which outputs an audio | voice. However, the configuration of the dialog device 2 in the dialog system of the present invention is not limited to the above configuration.

  For example, the dialogue apparatus 2 that does not have the voice recognition unit 59 can be employed in the dialogue system of the present invention. That is, the client side (dialogue device 2) does not have a voice recognition function, the dialogue device 2 transmits the voice picked up by the voice input unit 22 (microphone) to the utterance control server 1, and the utterance control server 1 recognizes the voice. To do. Such an interactive system also falls within the scope of the present invention. More specifically, the voice recognition unit 59 is provided in the control unit 10 of the utterance control server 1. The control unit 20 of the interactive apparatus 2 includes a voice data transmission unit (voice data transmission unit) (not shown) instead of the voice recognition unit 59. The voice data transmission unit transmits the voice data input via the voice input unit 22 to the utterance control server 1. The voice recognition unit 59 of the utterance control server 1 recognizes the voice data received from the voice data transmission unit. The utterance control server 1 generates a response voice in each downstream processing unit (not shown) for generating a response voice (not shown). Then, the utterance control unit 57 of the utterance control server 1 returns the generated response voice to the dialogue apparatus 2. In the dialogue apparatus 2, the dialogue mode control unit 58 instructs the voice output control unit 60 to output the response voice supplied from the utterance control server 1, and the voice output control unit 60 sends the response voice via the voice output unit 23. Is output. Thus, even when the dialogue apparatus 2 does not have the voice recognition unit 59, the dialogue apparatus 2 can interact with the user.

  In the above-described configuration, it is possible to further suppress output of response voice as follows.

  First, the utterance control unit 57 of the utterance control server 1 instructs the dialogue apparatus 2 not to transmit voice data when the condition determination unit 56 determines that the voice recognition condition is satisfied. In accordance with this instruction, the dialogue mode control unit 58 of the dialogue apparatus 2 prohibits the voice data transmission unit from transmitting voice data. According to the above configuration, since the voice data is not processed by the utterance control server 1, a reply voice is not generated and cannot be supplied to the dialogue apparatus 2.

  Secondly, the speech control unit 57 of the speech control server 1 receives the voice data transmitted from the voice data transmission unit of the interactive apparatus 2 when the condition determination unit 56 determines that the voice recognition condition is satisfied. To refuse. According to the above configuration, since voice data is not received, a reply voice is not generated in the utterance control server 1 and a reply voice cannot be supplied to the dialogue apparatus 2.

  Third, the utterance control unit 57 of the utterance control server 1 generates a response voice so as not to process the received voice data when the condition judgment unit 56 determines that the voice error recognition condition is satisfied. Control each processing unit. According to the above configuration, the response voice is not generated in the utterance control server 1 and the response voice cannot be supplied to the dialogue apparatus 2.

  Fourth, the utterance control unit 57 of the utterance control server 1 does not transmit the response voice generated by the own apparatus to the dialogue apparatus 2 when the condition determination unit 56 determines that the voice recognition condition is satisfied. . According to the above configuration, the response voice is not supplied to the dialogue apparatus 2.

  According to each of the above configurations, as a result, even if a voice that is not a recognition target is generated when the voice recognition condition is satisfied, the output of the reply voice is suppressed, and erroneous recognition and malfunction of the interactive device 2 can be avoided. it can.

<< Modification 4 >>
In the first embodiment, whether or not the erroneous voice recognition condition is satisfied may be determined by the information collection server 4. In this case, in the dialogue system of the present invention, the information collection server 4 is provided with a condition determination unit 56, and instead of the operation information transmission unit 53, a notification unit (not shown) for notifying the utterance control server 1 of the determination result of success / failure. Not shown).

<< Modification 5 >>
In the first embodiment, the information collection server 4 and the utterance control server 1 may be configured by a single computer. In this case, the operation information transmission unit 53 provided in the information collection server 4 and the operation information reception unit 54 provided in the speech control server 1 can be omitted.

<< Modification 6 >>
In the second to fourth embodiments, the information collection server 4 for managing the operation status of each device is omitted. However, when there is a purpose to centrally manage the operation status of each device for each user, each interactive system 402 to 404 may be constructed including the information collection server 4. Specifically, the dialog device 2 is configured to transmit the device ID and the operation information transmitted to the utterance control server 1 to the information collection server 4 in the dialog systems 402 to 404. Alternatively, the dialog device 2 may be configured to transmit the device ID and the operation information only to the information collection server 4, and the information collection server 4 transfers the device ID and the operation information to the utterance control server 1. As a result, requests from the interactive device 2 are collected in the information collection server 4, so that the operation status of each device can be managed in the information collection server 4 and the processing load of the utterance control server 1 can be controlled. Can be dispersed.

[Summary]
A dialogue system (401 to 404) according to an aspect 1 of the present invention is a dialogue system that controls a dialogue device (2) that recognizes a voice uttered by a user and outputs a response voice to the voice. A condition determination that determines that a speech misrecognition condition of the interactive device is satisfied when a non-recognition speech that is not a recognition target speech uttered to the interactive device can be erroneously detected as a recognition target speech by the interactive device Utterance control means (speech control section 57) for controlling the answering voice not to be output by the dialogue device when it is determined by the means (condition judgment section 56) and the condition judgment means that the voice recognition error condition is satisfied. ), And the condition determination means includes a non-recognized sound source device (for example, a sound output device, a call device) that directly or indirectly generates the non-recognized sound. Such as home appliances, and more specifically, TV 3, based on the telephone 3a, operation information indicating the operation status such as washing machines), which determine the success of the speech misrecognition conditions.

  According to said structure, based on the operation information which shows the operation condition of a non-target audio | voice generation | occurrence | production source apparatus, a condition determination means determines the success or failure of the audio | voice misrecognition conditions of a dialogue apparatus. For example, the condition determination means may determine success or failure every time the operation information is set (every time it is newly generated or updated). If it is determined by the condition determining means that the voice recognition condition is satisfied, the utterance control means controls the dialog device so that the response voice is not output.

Specifically, when the non-target audio source device around the dialog device is in an operating state in which non-recognition target sound is generated, the dialog device is directly or indirectly connected by the non-target sound source device. It is considered that the non-recognition voice generated in the above can be erroneously detected as the recognition target voice. Therefore, the condition determination means determines that the voice recognition condition is not correct for the interactive device when the non-target sound source device is in an operating state in which non-recognition target sound is generated based on the operation information of the non-target sound source device. Judgment is made. In this case, the output of response voice is suppressed by the speech control means. For example, the utterance control means may suppress the output by instructing the dialogue apparatus not to output a reply voice. Alternatively, when the dialogue apparatus does not have a function of generating a response voice by itself, the utterance control unit may suppress the output by not supplying the answer voice to the dialogue apparatus.

  Note that “indirect generation of non-recognized audio” means that the non-target audio source device is “not indirectly recognized” when the non-target audio source device has executed a function of its own device. This means that an entity different from the sound source device (for example, a user or a device different from the non-target sound source device) generates non-recognition sound.

  Based on the above, in a situation where misrecognition is likely to occur depending on the operating status of the non-target speech source device, that is, in a situation where unrecognized voice can occur around the dialog device, the dialog device Even if external audio is generated, no response audio is output. As a result, even if a speech that is not a recognition target is generated, there is an effect that it is possible to avoid erroneous speech recognition or malfunction associated therewith.

  In the dialog system according to aspect 2 of the present invention, in the above aspect 1, the non-target sound generation source device is a sound output device having at least a sound output function for outputting non-recognition target sound. While the sound output device is executing the sound output function, operation information setting means (operation information generating unit 52) sets operation information of the sound output device to indicate that the sound output function is being executed. ) And the condition determination means indicates that the operation information of the voice output device set by the operation information setting means indicates that the voice output function is being executed. It may be determined that the recognition condition is satisfied.

  According to the above configuration, when the untargeted sound source device is a sound output device having at least a sound output function for outputting unrecognized sound, the sound output device directly outputs unrecognized sound. While the data is being output, it is determined that the voice recognition condition for the interactive device is satisfied.

  From the above, in a situation where the voice output device directly outputs non-recognition target sound around the dialog device, the dialog device responds to the occurrence of the non-recognition sound. Will not be output. As a result, there is an effect that it is possible to avoid erroneous recognition of voice or the accompanying malfunction. Note that the situation in which the audio output device directly outputs unrecognized audio is not limited to this, for example, but the TV selects an arbitrary channel and outputs the audio together with the video. A situation, a situation in which video and audio of a program recorded by the recording / playback apparatus are output, or a situation in which the music playback apparatus is playing back music are assumed. When the user is viewing non-recognized audio output using these audio output devices, it is particularly advantageous because the viewing can be prevented from being disturbed by erroneous recognition and malfunction of the interactive device. Is big.

  In the dialogue system according to aspect 3 of the present invention, in the above aspect 1, the non-target audio source device is a call device for a user to make a call with a remote call partner, and the conversation system includes the call device. Operation information setting means (operation information) for setting operation information of the calling device so as to indicate that the call is in progress until a voice uttered by the user is interrupted for a predetermined time or more after outputting a ringing tone for notifying an incoming call. The condition determining means includes a generating unit 52), and the condition determination means determines that the voice recognition condition of the interactive device is in a state where the operation information of the calling device set by the operation information setting means indicates that a call is in progress. It may be determined that it is established.

  According to the above configuration, when the non-target audio generation source device is a call device, and the call device triggers the start of a call by outputting a ringing tone, the dialogue apparatus during the call Is determined to be satisfied.

  From the above, in a situation where the calling device indirectly generates unrecognized sound by outputting a ringing tone, that is, in a situation where the user is talking to the other party after receiving a call, the interactive device is not recognized. Even if voice (the voice of the user who is talking to the other party) is generated, no response voice is output. As a result, there is an effect that it is possible to avoid erroneous recognition of voice or the accompanying malfunction. When the user is making a call using a call device, the merit is particularly great because the call can be prevented from being disturbed by erroneous recognition and malfunction of the interactive device. In addition, although it is not limited to this as a telephone call apparatus, a fixed telephone, a mobile phone, a smart phone, an intercom etc. are assumed.

  In the dialog system according to aspect 4 of the present invention, in the above aspect 1, the non-target audio generation source device has at least a sound output function for outputting non-recognition target sound at a predetermined timing while the device is operating. An operation information setting unit (operation information generation unit 52) that sets operation information of the home appliance to indicate that it is in operation while the home appliance is in operation. And the condition determination means determines that the voice recognition condition for the interactive device is satisfied when the operation information of the home appliance set by the operation information setting means indicates that the operation is in progress. May be.

  According to the above configuration, when the non-target sound source device is a home device that has at least a sound output function of outputting non-recognition target sound at a predetermined timing while the device is operating, While operating, it is determined that the voice recognition condition of the interactive device is satisfied.

  From the above, in a situation where home appliances are operating around the interactive device and it is not always possible to output unrecognized speech, the interactive device On the other hand, no response voice is output. As a result, there is an effect that it is possible to avoid erroneous recognition of voice or the accompanying malfunction. The situation in which the home appliance directly outputs the non-recognized voice is not limited to this, for example, but the situation in which voice guidance for notifying the user of the operation status of the own device is output, A situation in which a voice guidance for notifying the user of the occurrence of an error and outputting a voice guidance for prompting the user to perform an operation on the device is assumed. While the user is using these home appliances, the interactive device does not speak due to erroneous recognition and malfunction. Therefore, since the user can use the home appliance without annoying thoughts, the merit is particularly great.

  In the dialog system according to aspect 5 of the present invention, in the above aspects 1 to 4, when the dialog apparatus acquires an instruction to suppress the output of the reply voice from the utterance control unit, the voice input to the own apparatus Dialogue control means (dialogue mode control) for disabling at least one of the functions of voice recognition means (voice recognition unit 59) for recognizing voice and voice output control means (voice output control unit 60) for executing output of the reply voice Part 58).

  According to said structure, the said dialogue apparatus acquires the instruction | indication which suppresses the output of the said response voice from the said speech control means in the condition which can detect unrecognized audio | voice as recognition object audio | voice accidentally. In this case, the dialog control means of the dialog device disables the function of at least one of the voice recognition means and the voice output control means.

As a result, if the function of the voice recognition unit is disabled, the voice recognition unit does not subject the voice recognition process to the voice recognition process even if a voice that is not a recognition target is generated. Therefore, it is possible to suppress a response voice from being erroneously output with respect to the non-recognition target voice. Further, if the function of the voice output control means is disabled, even if an unrecognized voice is generated and erroneously subjected to voice recognition processing as a recognition target voice, a reply voice for responding thereto is not output. As a result, even if a speech that is not a recognition target is generated, there is an effect that it is possible to avoid erroneous speech recognition or malfunction associated therewith.

  In the dialog system according to Aspect 6 of the present invention, in the Aspect 5, the speech control means further includes a speech corresponding to the event in response to the occurrence of the predetermined event even if the user does not input the voice. The dialogue control means instructs the dialogue apparatus to voluntarily output a voice having the content as a spontaneous voice, and the dialogue control means outputs the spontaneous voice according to an instruction of the utterance control means to output the spontaneous voice. After outputting, it is preferable to release the suppression of the response voice output for a certain period.

  When the dialog device speaks spontaneously, the user is likely to receive some response and return some response to the dialog device. However, it is unnatural as a dialogue that the dialogue device does not respond to the user's response while the output of the reply voice is suppressed.

  However, according to the above configuration, the output suppression of the response voice is released for a certain period after the dialogue apparatus outputs the spontaneous voice. During this period, the dialogue apparatus can temporarily output a response voice in response to the user's utterance. As a result, it is possible to eliminate the unnaturalness described above.

  An utterance control device (speech control server 1) according to an aspect 7 of the present invention is an utterance control device that recognizes a voice uttered by a user and controls an interactive device that outputs a response voice to the voice. A condition determination that determines that a speech misrecognition condition of the interactive device is satisfied when a non-recognition speech that is not a recognition target speech uttered to the interactive device can be erroneously detected as a recognition target speech by the interactive device Utterance control means (speech control section 57) for controlling the answering voice not to be output by the dialogue device when it is determined by the means (condition judgment section 56) and the condition judgment means that the voice recognition error condition is satisfied. ), And the condition determination means is based on operation information indicating an operation status of the non-recognized sound generation source device that directly or indirectly generates the non-recognized sound, Serial to determine the success or failure of speech recognition error conditions.

  According to said structure, a condition judgment part judges the success or failure of the audio | voice misrecognition conditions of a dialogue apparatus based on the operation information which shows the operation condition of a non-target audio | voice generation source apparatus. If it is determined by the condition determining means that the voice recognition condition is satisfied, the utterance control means controls the dialog device so that the response voice is not output.

  As a result, in situations where misrecognition is likely to occur depending on the operating status of the non-target audio source device, that is, in situations where non-recognition target audio may occur around the dialog device, the dialog device No response voice will be output. As a result, even if a speech that is not a recognition target is generated, there is an effect that it is possible to avoid erroneous speech recognition or malfunction associated therewith.

  The speech control apparatus according to aspect 8 of the present invention provides the speech control apparatus according to aspect 7 described above, by communicating with the non-target sound generation source device via a communication network and collecting information on the non-target sound generation source device. Operation information receiving means (operation information receiving unit 54) that receives the operation information from an information collection device (information collection server 4) that generates operation information of the external sound source device may be provided.

According to the above configuration, when the non-target sound source device has a communication function, the information collection device grasps the operating status of each non-target sound source device, and each non-target sound source device Operation information can be generated. The utterance control device only has to determine whether or not the voice recognition condition is satisfied using the operation information supplied from the information collecting device, and the configuration of the utterance control device can be simplified.

  The speech control apparatus according to aspect 9 of the present invention provides the operation information of the non-target sound generation source device in the aspect 7 from the interactive device that controls the operation of the non-target sound generation source device by short-range wireless communication. Operation information receiving means (operation information receiving unit 54) for receiving may be provided.

  According to the above configuration, when the interactive device has a control function for remotely controlling the non-target audio source device, the interactive device grasps the operating status of each non-target audio source device, It is possible to generate operation information of a non-target audio source device. Even when the non-target sound source device does not have a communication function, the object of erroneous recognition of sound or the accompanying malfunction can be achieved.

  The dialogue apparatus (2) according to the tenth aspect of the present invention is a dialogue apparatus that recognizes a voice uttered by a user and outputs a response voice in response to the voice. Condition determination means (condition determination unit 56) for determining that a voice error recognition condition of the dialog device is satisfied when a non-recognition voice that is not a target voice can be erroneously detected as a recognition target voice by the dialog device; A dialogue control means (interaction mode control unit 58) that suppresses output of the reply voice when the condition judgment means judges that the voice error recognition condition is satisfied, and the condition judgment means excludes the recognition target The success or failure of the speech error recognition condition is determined based on operation information indicating the operation status of the non-target sound generation source device that directly or indirectly generates sound.

  According to said structure, a condition judgment part judges the success or failure of the audio | voice misrecognition conditions of a dialogue apparatus based on the operation information which shows the operation condition of a non-target audio | voice generation source apparatus. If it is determined by the condition determining means that the voice recognition condition is satisfied, the dialogue control means suppresses the output of the reply voice of the own device.

  As a result, in situations where misrecognition is likely to occur depending on the operating status of the non-target audio source device, that is, in situations where non-recognition target audio may occur around the dialog device, the dialog device No response voice will be output. As a result, even if a speech that is not a recognition target is generated, there is an effect that it is possible to avoid erroneous speech recognition or malfunction associated therewith.

  In the dialog device according to aspect 11 of the present invention, in the above aspect 10, the dialog control means executes speech recognition means (speech recognition unit 59) for recognizing the sound input to the own device and output of the reply voice. The output of the response voice may be suppressed by disabling at least one of the functions of the voice output control means (voice output control unit 60).

  According to the above configuration, the dialog control unit disables at least one of the functions of the voice recognition unit and the voice output control unit in a situation where the non-recognition target voice can be erroneously detected as the recognition target voice.

  As a result, if the function of the voice recognition unit is disabled, the voice recognition unit does not subject the voice recognition process to the voice recognition process even if a voice that is not a recognition target is generated. Therefore, it is possible to suppress a response voice from being erroneously output with respect to the non-recognition target voice. Further, if the function of the voice output control means is disabled, even if an unrecognized voice is generated and erroneously subjected to voice recognition processing as a recognition target voice, a reply voice for responding thereto is not output. As a result, even if a speech that is not a recognition target is generated, there is an effect that it is possible to avoid erroneous speech recognition or malfunction associated therewith.

  In the dialog device according to aspect 12 of the present invention, in the above aspect 10 or 11, the dialog control means further responds to the event according to the occurrence of a predetermined event without the input of the voice uttered by the user. It is preferable that the voice having the content of the voice to be output is spontaneously output as the spontaneous voice, and the suppression is canceled for a certain period after the spontaneous voice is output while the output of the reply voice is being suppressed. .

  When the dialog device speaks spontaneously, the user is likely to receive some response and return some response to the dialog device. However, it is unnatural as a dialogue that the dialogue device does not respond to the user's response while the output of the reply voice is suppressed.

  However, according to the above configuration, the output suppression of the response voice is released for a certain period after the dialogue apparatus outputs the spontaneous voice. During this period, the dialogue apparatus can temporarily output a response voice in response to the user's utterance. As a result, it is possible to eliminate the unnaturalness described above.

  An utterance control method according to an aspect 13 of the present invention is an utterance control method for recognizing a voice uttered by a user and controlling a dialog device that outputs a response voice to the voice. A condition determination step (S114) for determining that the voice recognition condition of the interactive device is satisfied when the non-recognition speech that is not the recognition target speech generated in the above can be erroneously detected as the recognition target speech by the interactive device; An utterance control step (S115) for controlling the interactive device not to output the response voice when it is determined in the condition determining step that the voice error recognition condition is satisfied, and in the condition determining step, Based on the operation information indicating the operation status of the non-target audio generating device that directly or indirectly generates the non-recognized audio, the speech misrecognition condition To determine the success or failure of.

  According to the above method, in the condition determination step, whether or not the voice error recognition condition of the interactive device is satisfied is determined based on the operation information indicating the operation status of the non-target audio generation source device. If it is determined in the condition determination step that the erroneous voice recognition condition is satisfied, then in the utterance control step, control is performed so that the response voice is not output by the dialogue apparatus.

  As a result, in situations where misrecognition is likely to occur depending on the operating status of the non-target audio source device, that is, in situations where non-recognition target audio may occur around the dialog device, the dialog device No response voice will be output. As a result, there is an effect that it is possible to avoid erroneous recognition of voice or the accompanying malfunction.

  Each device included in the speech control device, the interactive device, and the interactive system according to each aspect of the present invention may be realized by a computer. In this case, a control program for the utterance control apparatus that causes the utterance control apparatus (or the interactive apparatus) to be realized by the computer by operating the computer as each unit included in the utterance control apparatus (or the interactive apparatus), and A computer-readable recording medium on which it is recorded also falls within the scope of the present invention.

  The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims, and embodiments obtained by appropriately combining technical means disclosed in different embodiments. Is also included in the technical scope of the present invention. Furthermore, a new technical feature can be formed by combining the technical means disclosed in each embodiment.

INDUSTRIAL APPLICABILITY The present invention can be used for an interactive apparatus that recognizes a voice uttered by a user and outputs a response voice in response to the voice and controls the voice.

1 Speech control server (speech control device)
2 Dialogue device 3 TV (excluded audio source device / audio output device)
3a Telephone (non-target audio source / calling equipment)
4 Information collection server (information collection device)
22 voice input unit 23 voice output unit 24 infrared transmission unit 32 operation unit 33 infrared reception unit 50 operation information transmission unit (operation information transmission means)
51 Operation information receiving unit (operation information receiving means)
52 Operation information generation unit (operation information setting means)
53 Operation information transmission unit (operation information transmission means)
54 Operation information receiving unit (operation information receiving means)
55 Control object specifying part (control object specifying means)
56 Condition Judgment Unit (Condition Judgment Unit)
57 Speech control unit (speech control means)
58 Dialog mode control unit (dialog control means)
59 Voice recognition unit (voice recognition means)
60 Audio output control unit (audio output control means)
61 Device operation unit (device operation means)
62 Sound determination unit (sound determination means)
401, 402, 403, 404 Dialogue system

Claims (15)

A dialogue system for controlling a dialogue device that outputs a response voice in response to a voice uttered by a user,
A condition for determining that a speech misrecognition condition of the interactive device is satisfied when a non-recognition speech that is not a recognition target speech uttered by the user to the interactive device can be erroneously detected as a recognition target speech by the interactive device Judgment means,
Utterance control means for controlling so that the reply voice is not output by the dialogue device when the condition judgment means judges that the voice error recognition condition is satisfied,
The condition judging means judges success or failure of the speech misrecognition condition based on operating information indicating an operating status of a non-target speech generating source device that directly or indirectly generates the non-recognized speech. Interactive system. The non-target sound generation source device is a sound output device having at least a sound output function for outputting non-recognition target sound,
The dialogue system sets operation information of the audio output device so as to indicate that the audio output function is being executed while the audio output device is executing the audio output function. Including
The condition judging means satisfies the voice recognition condition for the interactive device when the voice output device operating information set by the voice information setting means indicates that the voice output function is being executed. The dialogue system according to claim 1, wherein the dialogue system is determined to be. The non-target audio source device is a call device for a user to call a remote call partner,
The interactive system displays the operation information of the call device to indicate that the call device is busy during a period from when the call device outputs a ringing tone notifying an incoming call until the voice uttered by the user is interrupted for a predetermined time or more. Including operation information setting means to set,
The condition determining means determines that the voice recognition condition of the interactive device is satisfied when the operating information of the calling device set by the operating information setting means indicates that a call is in progress. The interactive system according to claim 1. The non-target audio generation source device is a household electrical appliance having at least a sound output function for outputting non-recognition target sound at a predetermined timing while the device is operating,
The interactive system includes operation information setting means for setting operation information of the home appliance to indicate that the home appliance is in operation while the home appliance is operating,
The condition determining means determines that a voice error recognition condition of the interactive device is satisfied when the operation information of the household electrical appliance set by the operation information setting means indicates that it is operating. The interactive system according to claim 1.   When the dialogue apparatus obtains an instruction to suppress the output of the reply voice from the utterance control means, the voice recognition means for recognizing the voice input to the own apparatus and the voice output control for executing the output of the reply voice The dialogue system according to any one of claims 1 to 4, further comprising dialogue control means for disabling at least one of the functions of the means. Further, the speech control means may spontaneously output a speech having a speech corresponding to the event as a spontaneous speech in response to the occurrence of a predetermined event, even if the speech uttered by the user is not input. Instructing the dialogue device,
6. The dialog control unit according to claim 5, wherein after the speech control unit outputs the spontaneous speech in accordance with an instruction to output the spontaneous speech, the suppression of the output of the response speech is canceled for a certain period. Interactive system. An utterance control device that controls an interactive device that outputs a response voice to a voice uttered by a user,
A condition for determining that a speech misrecognition condition of the interactive device is satisfied when a non-recognition speech that is not a recognition target speech uttered by the user to the interactive device can be erroneously detected as a recognition target speech by the interactive device Judgment means,
Utterance control means for controlling the interactive voice so that the reply voice is not output by the dialogue device when the voice judgment condition is judged to be satisfied by the condition judgment means,
The condition judging means judges success or failure of the speech misrecognition condition based on operating information indicating an operating status of a non-target speech generating source device that directly or indirectly generates the non-recognized speech. An utterance control device.   From the information collection device that generates the operation information of the non-target sound source device by collecting information on the non-target sound source device by communicating with the non-target sound source device through a communication network, The speech control apparatus according to claim 7, further comprising an operation information receiving unit that receives the operation information.   The operation information receiving means for receiving operation information of the non-target audio source device from the interactive device that controls the operation of the non-target audio source device by short-range wireless communication. 8. The utterance control device according to 7. An interactive device that outputs a response voice in response to a voice uttered by a user,
A condition for determining that a speech misrecognition condition of the interactive device is satisfied when a non-recognition speech that is not a recognition target speech uttered by the user to the interactive device can be erroneously detected as a recognition target speech by the interactive device Judgment means,
Dialogue control means for suppressing the output of the answer voice when the condition judgment means judges that the voice error recognition condition is satisfied,
The condition judging means judges success or failure of the speech misrecognition condition based on operating information indicating an operating status of a non-target speech generating source device that directly or indirectly generates the non-recognized speech. Interactive device.   The dialogue control means disables the function of at least one of the voice recognition means for recognizing the voice input to the device and the voice output control means for executing the output of the reply voice, thereby The interactive apparatus according to claim 10, wherein output is suppressed. Further, the dialogue control means spontaneously outputs a voice having a speech corresponding to the event as a spontaneous voice in response to the occurrence of a predetermined event without the input of the voice uttered by the user. Yes,
The interactive apparatus according to claim 10 or 11, wherein after the spontaneous voice is output while the output of the reply voice is suppressed, the suppression is released for a certain period. An utterance control method for controlling an interactive device that outputs a response voice in response to a voice uttered by a user,
A condition for determining that a speech misrecognition condition of the interactive device is satisfied when a non-recognition speech that is not a recognition target speech uttered by the user to the interactive device can be erroneously detected as a recognition target speech by the interactive device A decision step;
An utterance control step for controlling so that the answering voice is not output by the dialogue device when it is determined in the condition determining step that the voice error recognition condition is satisfied,
In the condition determining step, the success or failure of the speech misrecognition condition is determined based on operation information indicating an operation status of an untargeted sound generation source device that directly or indirectly generates the unrecognized sound. Utterance control method.   A control program for causing a computer to function as the speech control apparatus according to any one of claims 7 to 9, wherein the control program causes the computer to function as each of the means.   A control program for causing a computer to function as the interactive apparatus according to any one of claims 10 to 12, wherein the control program causes the computer to function as each of the above means.

Images