Voice interaction device and control method for voice interaction device

Abstract

To provide a voice interaction device capable of enhancing a sense of unity between a driver and the voice interaction device when the driver of a vehicle interacts with the voice interaction device.SOLUTION: A voice interaction device (1) includes: a vehicle control state acquisition unit (111) configured to acquire information on a control state of a vehicle; an inference unit (113) configured to infer a mood of a driver of the vehicle, based on the acquired information on the control state of the vehicle; and a dialogue reply method setting unit (114) that sets a manner of replying to the driver according to the inferred mood of the driver.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a voice dialogue device and a control method for a voice dialogue device. [Background technology]

[0002] Patent document 1 discloses an in-vehicle route guidance device that controls at least one of the timing, number of times, or content of output of guidance information providing guidance to a guide point based on the driver's current load state and future load state. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2017 / 085833 Summary of the Invention [Problem to be solved by the invention]

[0004] According to the invention described in Patent Document 1, it is possible to output guidance information that is more appropriate for the driver in terms of the driver's load while driving. However, the invention described in Patent Document 1 only changes the timing, number of times, or content of output of guidance information based on the driver's load state, and is unable to enhance the sense of unity of the dialogue with the driver when the in-vehicle route guidance device dialogues with the driver.

[0005] The present invention has been made in consideration of the above, and its object is to provide a voice dialogue device and a control method for a voice dialogue device that can enhance a sense of unity between the driver and the voice dialogue device when the driver of a vehicle interacts with the voice dialogue device. [Means for solving the problem]

[0006] A voice dialogue device and a control method for a voice dialogue device according to one embodiment of the present invention acquire information regarding the control state of a vehicle, estimate the mood of the vehicle driver based on the acquired information regarding the control state of the vehicle, and set the manner of responding to the driver depending on the estimated driver's mood. Effect of the Invention

[0007] According to the present invention, when a driver of a vehicle interacts with a voice dialogue device, a sense of unity between the driver and the voice dialogue device can be enhanced. [Brief description of the drawings]

[0008] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a voice dialogue device according to an embodiment. [Diagram 2] FIG. 2 is a table showing a specific example in which the driver's mood estimating unit estimates the driver's mood based on information related to the control state of the vehicle. [Diagram 3] FIG. 3 is a table showing a specific example of how the dialogue response method setting section 114 sets a response method based on the driver's mood. [Figure 4] FIG. 4 is a flowchart illustrating an example of a process executed by the voice dialogue device according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] [Configuration of the spoken dialogue device] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. Fig. 1 shows an example of the configuration of a voice dialogue device 1 according to an embodiment. The voice dialogue device 1 according to this embodiment is installed in a vehicle. The voice dialogue device 1 includes a dialogue generation device 10, a vehicle control state input unit 20, a voice input unit 30, and a voice output unit 40.

[0010] The vehicle control state input unit 20 is a variety of sensors, switches or devices mounted on the vehicle, and inputs to the dialogue generation device 10 information relating to the vehicle control state that can be set by the driver's operation.

[0011] The vehicle control state input unit 20 inputs, for example, information indicating the current driving mode of the vehicle to the dialogue generation device 10. The driving mode of the vehicle is a driving specification that varies the responsiveness of the drive actuator to the required driving force by the accelerator operation, and includes, for example, a "sport mode" in which the responsiveness of the drive actuator to the required driving force is high, and an "eco mode" in which the responsiveness of the drive actuator to the required driving force is low to improve fuel efficiency. The driving mode is selected by the driver.

[0012] In addition, when the vehicle is under automatic driving control, the vehicle control state input unit 20 inputs automatic driving control information indicating at least one of the following during automatic driving control: the distance to the preceding vehicle, the acceleration, speed, frequency of lane changes, or the driving mode of the vehicle. The driver selects one of multiple settings for the distance between the vehicles, such as "short distance," "medium distance," and "long distance." The acceleration and speed are autonomously controlled so that the vehicle reaches the set speed set by the driver. The start of lane changes is approved by the driver, and the number of times lane changes are approved within a predetermined time is calculated as the frequency of lane changes. In addition, the driving mode during autonomous driving control is selected by the driver.

[0013] The vehicle control state input unit 20 also inputs information about the driver's intervention (override) in the automatic driving control of the vehicle to the dialogue generation device 10. The information about the override includes, for example, information about the acceleration / deceleration of the vehicle caused by the driver's operation during automatic driving control, whether or not deceleration occurs during a preparatory action, and whether or not steering occurs. Note that the preparatory action referred to here means an action of the driver slowing down to stop before a white line, an intersection, a railroad crossing, etc. during automatic driving control. The driver may override and decelerate the vehicle when he / she wants to decelerate faster than the deceleration caused by the automatic driving control, or when he / she wants to decelerate even more than the deceleration caused by the automatic driving control.

[0014] Furthermore, the vehicle control state input unit 20 inputs information indicating the state of the vehicle height to the dialogue generation device 10. The vehicle height is selected by the driver to be a height suitable for driving on "off-road" unpaved roads, or a height suitable for driving on "on-road" paved roads.

[0015] The voice input unit 30 receives voice information uttered by the driver in the vehicle using a sound collection device such as a microphone installed in the vehicle. The sound collection device is not limited to a microphone, and may be any sensor that can input voice. The voice input unit 30 inputs the voice information in the vehicle input from the sound collection device to the dialogue generation device 10.

[0016] The audio output unit 40 uses a speaker installed in the vehicle to output the dialogue data generated by the dialogue generation device 10 as voice within the vehicle.

[0017] The dialogue generation device 10 acquires voice information uttered by the driver of a vehicle, who is the dialogue partner, via the voice input unit 30, and generates dialogue data for asking a question or replying to the driver. The dialogue generation device 10 outputs the generated dialogue data as voice via the voice output unit 40. The dialogue generation device 10 includes a controller 11 and a dialogue data storage unit 12.

[0018] The controller 11 is a general-purpose computer including a CPU (Central Processing Unit), a memory, a storage device, an input / output unit, etc. The controller 11 is connected to a vehicle control state input unit 20, an audio input unit 30, and an audio output unit .

[0019] A computer program for causing the controller 11 to function as the dialogue generation device 10 is installed in the controller 11. By executing the computer program, the controller 11 functions as a plurality of information processing circuits included in the dialogue generation device 10.

[0020] Although an example is shown here in which the multiple information processing circuits of the dialogue generation device 10 are realized by software, it is of course possible to configure the information processing circuits by preparing dedicated hardware for executing each piece of information processing described below. Also, the multiple information processing circuits may be configured by individual hardware.

[0021] The controller 11 includes a vehicle control state acquisition unit 111, a priority setting unit 112, a driver mood estimation unit 113, a dialogue response method setting unit 114, a dialogue generation unit 115, and an output control unit 116 as a plurality of information processing circuits.

[0022] The vehicle control state acquisition unit 111 acquires information on the vehicle control state that can be set by the driver's operation from the vehicle control state input unit 20. The vehicle control state acquisition unit 111 may acquire one type of information on the vehicle control state, or may acquire a plurality of pieces of information on the vehicle control state. The number and types of information on the vehicle control state acquired by the vehicle control state acquisition unit 111 are not particularly limited.

[0023] The vehicle control state acquisition unit 111 acquires, as information on the control state of the vehicle, for example, information indicating a driving mode that determines the responsiveness of the drive actuator to the required driving force of the vehicle.

[0024] In addition, the vehicle control state acquisition unit 111 acquires automatic driving control information indicating at least one of the following during automatic driving control of the vehicle: vehicle distance, acceleration, speed, frequency of lane changes, or driving mode, as information regarding the control state of the vehicle.

[0025] Furthermore, the vehicle control state acquisition unit 111 acquires, as information on the control state of the vehicle, for example, information on the driver's override of the automatic driving control of the vehicle.

[0026] Furthermore, the vehicle control state acquisition unit 111 acquires, for example, information indicating the state of the vehicle height as information relating to the control state of the vehicle.

[0027] When the vehicle control state acquisition unit 111 acquires a plurality of pieces of information related to the vehicle control state, the priority setting unit 112 sets priorities of the plurality of pieces of information based on the degree of influence of the driver's mood on the vehicle control state. The driver mood estimation unit 113, which will be described later, estimates the mood of the vehicle driver based on the acquired information related to the vehicle control state and the priority.

[0028] For example, the driver's override operation for the automatic driving control is the most recent operation of the driver, and it is considered that the most recent driver's mood has the greatest influence on the control state of the vehicle. Therefore, the priority setting unit 112 sets the priority of the plurality of pieces of information such that, for example, the priority of the information on the driver's override is the highest among the plurality of pieces of information on the control state of the vehicle acquired. In addition, it is considered that the driver's mood also affects the driver's operation for the automatic driving control. Therefore, the priority setting unit 112 may set the priority of the plurality of pieces of information such that the priority of the automatic driving control information is the second. By setting the priority of the automatic driving control information to the second, the driver mood estimation unit 113 can estimate the driver's mood even when the driver has not overridden the automatic driving.

[0029] In addition, it is considered that the driver's mood also affects the operation related to the vehicle height state of the driver. Therefore, the priority setting unit 112 may set the priority of a plurality of pieces of information so that the priority of the information indicating the vehicle height state is the third. By setting the priority of the information indicating the vehicle height state to the third highest, the driver mood estimation unit 113 can estimate the driver's mood even when the automatic driving is not overridden and the automatic driving control is not performed.

[0030] The driver mood estimation unit 113 estimates the mood of the driver of the vehicle based on the acquired information on the control state of the vehicle. When the vehicle control state acquisition unit 111 acquires a plurality of pieces of information on the control state of the vehicle, the driver mood estimation unit 113 estimates the mood of the driver of the vehicle based on the plurality of pieces of information and a priority. For example, the driver mood estimation unit 113 estimates the mood of the driver of the vehicle based on the information with the highest priority among the plurality of pieces of information on the control state of the vehicle. The driver mood estimation unit 113 estimates whether the mood of the driver of the vehicle is "impatient (first mood)" or "relaxed (second mood)", which is a calmer mood than "impatient", based on the acquired information on the control state of the vehicle.

[0031] The driver mood estimation unit 113 may estimate the vehicle driver's mood based on the most recently changed information on the vehicle control state instead of estimating the vehicle driver's mood based on the acquired information on the vehicle control state and the priority. The driver mood estimation unit 113 may also calculate the degree to which the driver's mood is "impatient" or "relaxed" by weighting multiple pieces of information on the vehicle control state in descending order of priority.

[0032] Now, with reference to FIG. 2, a specific example in which the driver's mood estimation unit 113 estimates the driver's mood based on information relating to the control state of the vehicle will be described.

[0033] The driver mood estimation unit 113 estimates the driver's mood based on, for example, information indicating the driving mode. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the driving mode is "sport mode" and that the driver's mood is "relaxed" when the driving mode is "eco mode."

[0034] The driver mood estimation unit 113 may also estimate the driver's mood based on, for example, automatic driving control information indicating the distance between vehicles during automatic driving control. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the distance between vehicles is "short distance" and that the driver's mood is "relaxed" when the distance between vehicles is "long distance".

[0035] The driver mood estimation unit 113 may estimate the driver's mood based on, for example, automatic driving control information indicating acceleration during automatic driving control. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the acceleration is greater than a first threshold, and that the driver's mood is "relaxed" when the acceleration is less than the first threshold.

[0036] The driver mood estimation unit 113 may estimate the driver's mood based on, for example, automatic driving control information indicating the speed during automatic driving control. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the speed is faster than the second threshold, and that the driver's mood is "relaxed" when the speed is slower than the second threshold.

[0037] The driver mood estimation unit 113 may estimate the driver's mood based on, for example, automatic driving control information indicating the frequency of lane changes during automatic driving control. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" if the frequency of lane changes is greater than a third threshold, and that the driver's mood is "relaxed" if the acceleration is less than the third threshold.

[0038] The driver mood estimation unit 113 may estimate the driver's mood, for example, based on information regarding acceleration and deceleration during driver override. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the driver accelerates and decelerates faster than the automatic driving control, and estimates that the driver's mood is "relaxed" when the driver accelerates and decelerates slower than the automatic driving control. The driver mood estimation unit 113 estimates that the driver's mood is "relaxed" when the driver performs an operation to decelerate faster than the automatic driving during a preparatory operation. The driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the driver steers during automatic driving.

[0039] The driver mood estimation unit 113 may estimate the driver's mood based on information about deceleration during a preparatory operation when the driver overrides the vehicle. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "relaxed" if the driver performs an operation to decelerate the vehicle more than the automatic driving during the preparatory operation.

[0040] The driver mood estimation unit 113 may estimate the driver's mood based on, for example, information about steering during driver override. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the driver steers during automatic driving.

[0041] The driver mood estimation unit 113 may estimate the driver's mood based on information indicating the vehicle height state, for example. In this case, the driver mood estimation unit 113 estimates that the driver's mood is "impatient" when the vehicle height is set to a state suitable for "off-road" driving, and estimates that the driver's mood is "relaxed" when the vehicle height is set to a state suitable for "on-road" driving.

[0042] The dialogue response method setting unit 114 sets the manner of replying to the driver according to the estimated driver's mood.

[0043] Now, with reference to FIG. 3, a specific example in which the dialogue response method setting section 114 sets the response method based on the driver's mood will be described.

[0044] The dialogue response method setting unit 114 sets the manner of response so that the time for replying to the driver is shorter when the estimated driver's mood is "impatient" and the time for replying to the driver is longer when the estimated driver's mood is "relaxed". The dialogue response method setting unit 114 sets the time for replying to the driver to "short" when the estimated driver's mood is "impatient" and sets the time for replying to the driver to "long" when the estimated driver's mood is "relaxed".

[0045] Furthermore, the dialogue response method setting unit 114 may set the response method so that, for example, when the estimated driver's mood is "impatient," the driver responds by adding, at the start of the response, a word that suggests a hasty dialogue response (first predetermined word), such as "um," "ah," etc. Furthermore, the dialogue response method setting unit 114 may set the response method so that, for example, when the estimated driver's mood is "relaxed," the driver responds by adding, at the start of the response, a word that suggests a thoughtful dialogue response (second predetermined word), such as "hmm," "you're right," etc.

[0046] Furthermore, the dialogue response method setting section 114 may set the response method so that, for example, when the estimated driver's mood is "relaxed," the driver responds by dividing the response into words.

[0047] Furthermore, the dialogue response method setting section 114 may set the response method to be a noun-finished reply when the estimated driver's mood is "impatient", for example.

[0048] The dialogue response method setting unit 114 may change at least one of the speed, accent, and intonation of the response according to the estimated driver's mood. For example, when the estimated driver's mood is "impatient", the dialogue response method setting unit 114 sets the response method so that the speed of the response is faster, the accent is stronger, and the intonation is stronger, and when the estimated driver's mood is "relaxed", the dialogue response method setting unit 114 sets the response method so that the speed of the response is slower, the accent is weaker, and the intonation is weaker. The dialogue response method setting unit 114 may set only one of the above-mentioned multiple response methods, or may combine multiple response methods. The type and number of response methods to be set are not particularly limited. When the driver mood estimation unit 113 calculates the degree to which the driver's mood is "impatient" and the degree to which the driver's mood is "relaxed", the dialogue response method setting unit 114 may set the response method in stages based on the degree of "impatient" and "relaxed".

[0049] The dialogue generation unit 115 recognizes the voice information input from the voice input unit 30, and acquires dialogue data corresponding to the recognized voice information from the dialogue data storage unit 12. The dialogue generation unit 115 changes the acquired dialogue data based on the reply method set by the dialogue reply method setting unit 114, and generates dialogue data for output.

[0050] For example, when the dialogue response method setting unit 114 sets the reply method to add impatient or thoughtful words to reply, the dialogue generation unit 115 generates dialogue data for output by adding voice data corresponding to the words to be added to the beginning of the dialogue data. Also, when the dialogue response method setting unit 114 sets the reply method to be a noun-finished reply, the dialogue generation unit 115 generates dialogue data for output such that the dialogue data ends with a noun.

[0051] The output control unit 116 controls the dialogue data for output generated by the dialogue generation unit 115 based on the manner of response set by the dialogue response method setting unit 114, and outputs the dialogue data as voice via the voice output unit 40. For example, when the dialogue response method setting unit 114 sets the response time to "short", the output control unit 116 starts outputting the dialogue data for output 0.4 seconds after the driver's voice acquired from the voice input unit 30 ends. Also, when the dialogue response method setting unit 114 sets the response time to "long", the output control unit 116 starts outputting the dialogue data for output 1.4 seconds after the driver's voice acquired from the voice input unit 30 ends.

[0052] In addition, for example, when the dialogue response method setting unit 114 sets the response method so that the response is made by dividing the response into words, the output control unit 116 outputs the dialogue data for output generated by the dialogue generation unit 115 by dividing the data into words.

[0053] Furthermore, the output control section 116 outputs the dialogue data for output generated by the dialogue generation section 115 with the speed, accent and intonation of the response changed by the dialogue response method setting section 114 .

[0054] [Processing performed by the voice dialogue device] Next, an example of processing executed by the dialogue generation device 10 will be described with reference to the flowchart of Fig. 4. The dialogue generation device 10 starts the processing of Fig. 4 when voice information is input from the voice input unit 30.

[0055] In step S11, the vehicle control state acquisition unit 111 acquires a plurality of pieces of information related to the control state of the vehicle. For example, the vehicle control state acquisition unit 111 acquires information related to the driver's override of the automatic driving control of the vehicle, automatic driving control information of the vehicle, and information indicating the state of the vehicle height.

[0056] The process proceeds to step S12, where the priority setting unit 112 sets priorities of the multiple pieces of information based on the degree of influence of the driver's mood on the control state of the vehicle. For example, the priority setting unit 112 sets priorities of the multiple pieces of information such that, among the multiple pieces of information related to the control state of the vehicle acquired in step S11, information related to the override of the driver has the highest priority. The priority setting unit 112 also sets priorities of the multiple pieces of information such that the automatic driving control information has the second priority and the information indicating the vehicle height state has the third priority.

[0057] The process proceeds to step S13, where the driver mood estimation unit 113 estimates the mood of the driver of the vehicle based on the information on the control state of the vehicle acquired in step S11 and the priority set in step S12. The driver mood estimation unit 113 estimates the mood of the driver of the vehicle based on the information with the highest priority among multiple pieces of information on the control state of the vehicle. The driver mood estimation unit 113 estimates whether the mood of the driver of the vehicle is "impatient (first mood)" or "relaxed (second mood)", which is a calmer mood than "impatient", based on the acquired information on the control state of the vehicle.

[0058] The process proceeds to step S14, where the dialogue response method setting unit 114 sets the manner of replying to the driver according to the driver's mood estimated in step S13. The dialogue response method setting unit 114 sets the manner of replying so that when the estimated driver's mood is "impatient", the time for replying to the driver is shorter, and when the estimated driver's mood is "relaxed", the time for replying to the driver is longer.

[0059] Furthermore, when the driver's mood estimated in step S13 is "impatient", the dialogue response method setting unit 114 sets the response method so that the driver adds, at the start of the response, words that suggest a hasty dialogue response (first predetermined words), such as "umm," "ah," etc. Furthermore, when the driver's mood estimated in step S14 is "relaxed," the dialogue response method setting unit 114 sets the response method so that the driver adds, at the start of the response, words that suggest a thoughtful dialogue response (second predetermined words), such as "hmm," "you're right," etc.

[0060] The process proceeds to step S15, where the dialogue generation unit 115 recognizes the voice information input from the voice input unit 30, and acquires dialogue data corresponding to the recognized voice information from the dialogue data storage unit 12. The dialogue generation unit 115 modifies the acquired dialogue data based on the manner of reply set in step S14, and generates dialogue data for output. Based on the manner of reply set in step S14, the dialogue generation unit 115 adds voice data corresponding to the words to be added to the beginning of the dialogue data, and generates dialogue data for output.

[0061] The process proceeds to step S16, where the output control unit 116 controls the dialogue data for output generated in step S15 based on the manner of response set in step S14, and outputs the dialogue data for output to the voice output unit 40. If the dialogue response method setting unit 114 sets the response time to "short" in step S14, the output control unit 116 starts outputting the dialogue data for output 0.4 seconds after the driver's voice acquired from the voice input unit 30 ends. If the dialogue response method setting unit 114 sets the response time to "long" in step S14, the output control unit 116 starts outputting the dialogue data for output 1.4 seconds after the driver's voice acquired from the voice input unit 30 ends. Thereafter, the dialogue generation device 10 ends the process of FIG. 4.

[0062] [Effects] As described above, the voice dialogue device 1 according to one embodiment includes the vehicle control state acquisition unit 111, the estimation unit (driver mood estimation unit) 113, and the dialogue response method setting unit 114. The vehicle control state acquisition unit 111 acquires information related to the control state of the vehicle. The estimation unit 113 estimates the mood of the driver of the vehicle based on the acquired information related to the control state of the vehicle. The dialogue response method setting unit 114 sets a manner of replying to the driver according to the estimated driver's mood. The vehicle control state acquisition unit 111 acquires information related to the control state of the vehicle that can be set by the driver's operation.

[0063] Since the control state of the vehicle can be controlled by the driver, the information on the control state of the vehicle reflects the driver's mood. Therefore, the estimation unit 113 can estimate the driver's mood based on the information on the control state of the vehicle. The dialogue response method setting unit 114 can set a response method that matches the estimated driver's mood. The dialogue generation unit 115 can generate dialogue data based on the set response method, and the output control unit 116 can control and output the dialogue data based on the set response method. This makes it possible for the voice dialogue device 1 to enhance a sense of unity between the driver of the vehicle and the voice dialogue device 1 when the driver of the vehicle dialogues with the voice dialogue device 1.

[0064] Furthermore, the vehicle control state acquisition unit 111 of the voice dialogue device 1 according to an embodiment may acquire, as information on the control state of the vehicle, information indicating a driving mode that determines the responsiveness of the drive actuator to the required driving force of the vehicle.

[0065] By acquiring information indicating the vehicle driving mode set by the driver, the estimation unit 113 can estimate the driver's mood based on the driving mode.

[0066] In addition, the vehicle control state acquisition unit 111 may acquire automatic driving control information indicating at least one of the following distance, acceleration, speed, frequency of lane changes, and driving mode during automatic driving control of the vehicle as information on the control state of the vehicle. By acquiring the automatic driving control information set by the driver, the estimation unit 113 can estimate the driver's mood based on the automatic driving control information.

[0067] Furthermore, the vehicle control state acquisition unit 111 may acquire information on the driver's override of the automatic driving control of the vehicle as the information on the control state of the vehicle. The estimation unit 113 can estimate the driver's mood based on the information on the driver's override.

[0068] In addition, the vehicle control state acquisition unit 111 may acquire information indicating the state of the vehicle height as the information regarding the control state of the vehicle. By acquiring the information indicating the state of the vehicle height, the estimation unit 113 can estimate the driver's mood based on the state of the vehicle height.

[0069] Furthermore, the estimation unit 113 estimates whether the mood of the driver of the vehicle is a first mood or a second mood that is calmer than the first mood, based on the acquired information about the control state of the vehicle. The dialogue response method setting unit 114 sets a response method so that when the estimated driver's mood is the first mood, the time to reply to the driver is shorter than when the estimated driver's mood is the second mood.

[0070] The dialogue response method setting unit 114 changes the interval for replying to the driver in accordance with the driver's mood, so that the output control unit 116 can output dialogue data in an interval that suits the driver's mood at that time. This allows the driver to dialogue with the voice dialogue device 1 in an interval that suits his / her mood, and reduces the sense of discomfort that the voice dialogue device 1 gives to the driver during dialogue. As a result, the voice dialogue device 1 can enhance the sense of unity between the driver and the voice dialogue device when the driver of the vehicle dialogues with the voice dialogue device 1.

[0071] Moreover, the dialogue response method setting unit 114 may set the manner of replying such that, when the estimated driver's mood is a first mood, a first predetermined word is added at the start of replying. Moreover, the dialogue response method setting unit 114 may set the manner of replying such that, when the estimated driver's mood is a second mood, a second predetermined word is added at the start of replying. This allows the dialogue generation unit 115 to generate dialogue data to which a predetermined word matching the driver's mood is added at the start of replying to the driver, and the output control unit 116 to output the generated dialogue data. This allows the voice dialogue device 1 to further enhance the sense of unity between the driver and the voice dialogue device 1.

[0072] Furthermore, the dialogue response method setting unit 114 may set the manner of replying so that the driver responds by dividing the response into words when the estimated driver's mood is the second mood. This allows the dialogue generation unit 115 to generate dialogue data in which the driver responds by dividing the response into words, and the output control unit 116 to output the generated dialogue data. This allows the dialogue generation unit 115 and the output control unit 116 to output dialogue data with a dialogue tendency that matches the driver's mood at that time. The voice dialogue device 1 can further enhance the sense of unity between the driver and the voice dialogue device 1.

[0073] Furthermore, the dialogue response method setting unit 114 may set the manner of reply so that the driver responds with a noun-ending sentence when the estimated driver's mood is the first mood. This allows the dialogue generation unit 115 to generate dialogue data in which the driver responds with a noun-ending sentence, and the output control unit 116 to output the generated dialogue data. This allows the dialogue generation unit 115 and the output control unit 116 to output dialogue data with a dialogue tendency that matches the driver's mood at that time, thereby making it possible to further enhance the sense of unity between the driver and the voice dialogue device 1.

[0074] Furthermore, the dialogue response method setting unit 114 may change at least one of the speed, accent, and intonation of the response according to the estimated driver's mood. This allows the output control unit 116 to output dialogue data at a speed, accent, and intonation that match the driver's mood. The driver can dialogue with the voice dialogue device at a speed, accent, and intonation that match his or her mood, and the sense of discomfort that the voice dialogue device gives to the driver during dialogue can be reduced. As a result, the voice dialogue device can enhance the sense of unity between the driver of the vehicle and the voice dialogue device 1 when the driver dialogues with the voice dialogue device 1.

[0075] Moreover, the voice dialogue device according to an embodiment further includes a priority setting unit 112. The vehicle control state acquisition unit 111 acquires a plurality of pieces of information related to the control state of the vehicle that can be operated by the driver. The priority setting unit 112 sets the priority of the plurality of pieces of information based on the magnitude of the influence of the driver's mood on the control state of the vehicle. The estimation unit 113 estimates the mood of the driver of the vehicle based on the plurality of pieces of information and the priority. By estimating the mood of the driver of the vehicle based on the plurality of pieces of information related to the control state of the vehicle and the priority, the estimation unit 113 can estimate the driver's mood more accurately by taking into account the magnitude of the influence of the driver's mood on the control state of the vehicle.

[0076] Furthermore, the priority setting unit 112 may set priorities of a plurality of pieces of information such that the information on the driver's override has the highest priority. By setting the priority of the information on the driver's override to the highest priority, the estimation unit 113 can estimate the driver's mood from the most recent behavior of the driver, and can estimate the driver's mood more accurately.

[0077] Furthermore, the priority setting unit 112 may set the priorities of a plurality of pieces of information so that the priority of the autonomous driving control information is the second highest. By setting the priority of the autonomous driving control information to the second highest, the estimation unit 113 can estimate the driver's mood from the driver's behavior even when the autonomous driving is not overridden.

[0078] Furthermore, the priority setting unit 112 may set the priorities of a plurality of pieces of information such that the priority of the information indicating the state of the vehicle height of the vehicle is the third highest. By setting the information indicating the state of the vehicle height to the third highest, the estimation unit 113 can estimate the driver's mood from the driver's behavior even when the automatic driving is not overridden and the automatic driving control is not being performed.

[0079] Furthermore, the estimation unit 113 may estimate the mood of the driver of the vehicle based on information related to the most recently changed control state of the vehicle. This allows the estimation unit 113 to estimate the driver's mood based on information reflecting the most recent driver's mood, thereby making it possible to more accurately estimate the current driver's mood.

[0080] As described above, the embodiment of the present invention has been described, but the description and drawings forming a part of this disclosure should not be understood as limiting this invention. From this disclosure, various alternative embodiments, examples and operating techniques will become apparent to those skilled in the art. [Explanation of symbols]

[0081] 1 Speech dialogue device 111 Vehicle control status acquisition unit 113 Driver mood estimation unit (estimation unit) 114 Dialogue response method setting section

Claims

1. a vehicle control state acquisition unit that acquires information regarding a control state of the vehicle; an estimation unit that estimates a mood of a driver of the vehicle based on the acquired information about a control state of the vehicle; a dialogue response method setting unit that sets a manner of replying to the driver in accordance with the estimated mood of the driver; A voice dialogue device comprising:

2. The vehicle control state acquisition unit acquires information regarding a control state of the vehicle that can be set by an operation of the driver.

2. The speech dialogue device according to claim 1.

3. The vehicle control state acquisition unit acquires information indicating a driving mode that determines a responsiveness of a drive actuator to a required driving force of the vehicle.

3. The speech dialogue device according to claim 2.

4. The vehicle control state acquisition unit acquires automatic driving control information indicating at least one of a vehicle distance, an acceleration, a speed, a frequency of lane changes, or a driving mode during automatic driving control of the vehicle.

3. The speech dialogue device according to claim 2.

5. The vehicle control state acquisition unit acquires information regarding an override of the driver for the automatic driving control of the vehicle.

3. The speech dialogue device according to claim 2.

6. The vehicle control state acquisition unit acquires information indicating a state of a vehicle height of the vehicle.

3. The speech dialogue device according to claim 2.

7. The estimation unit estimates whether a mood of the driver of the vehicle is a first mood or a second mood that is calmer than the first mood based on the acquired information about a control state of the vehicle, The dialogue response method setting unit sets the manner of response so that, when the estimated mood of the driver is the first mood, a time for replying to the driver is shorter than when the estimated mood of the driver is the second mood. The voice dialogue device according to any one of claims 1 to 6.

8. The dialogue response method setting unit sets the manner of response so that, when the estimated mood of the driver is the first mood, a first predetermined word is added at the start of the response. The speech dialogue device according to claim 7.

9. The dialogue response method setting unit sets the manner of response so that, when the estimated mood of the driver is the second mood, a second predetermined word is added at the start of the response. The speech dialogue device according to claim 7.

10. The dialogue response method setting unit sets the manner of response so that the driver responds by dividing the response into words when the estimated driver's mood is the second mood. The speech dialogue device according to claim 7.

11. The dialogue response method setting unit sets the manner of response so that the driver responds with a noun-finished sentence when the estimated driver's mood is the first mood. The speech dialogue device according to claim 7.

12. The dialogue response method setting unit changes at least one of the speed, accent, and intonation of the response according to the estimated mood of the driver. The speech dialogue device according to claim 7.

13. the vehicle control state acquisition unit acquires a plurality of pieces of information relating to a control state of the vehicle that can be operated by the driver, a priority setting unit that sets priorities of the plurality of pieces of information based on a degree of influence of a driver's mood on a control state of the vehicle; The estimation unit estimates a mood of the driver of the vehicle based on the plurality of pieces of information and the priority.

2. The speech dialogue device according to claim 1.

14. The vehicle control state acquisition unit acquires at least information regarding an override of the driver with respect to an automatic driving control of the vehicle, The priority setting unit sets the priorities of the plurality of pieces of information so that the information regarding the override of the driver has the highest priority. The speech dialogue device according to claim 13.

15. The vehicle control state acquisition unit further acquires autonomous driving control information indicating at least one of a vehicle distance, an acceleration, a speed, a frequency of lane changes, or a driving mode during autonomous driving control of the vehicle, The priority setting unit sets the priorities of the plurality of pieces of information so that the priority of the automatic driving control information is the second highest. The speech dialogue device according to claim 14.

16. The vehicle control state acquisition unit further acquires information indicating a vehicle height state of the vehicle, The priority setting unit sets the priorities of the plurality of pieces of information so that the information indicating the vehicle height state of the vehicle has a third priority. The speech dialogue device according to claim 15.

17. The estimation unit estimates the mood of the driver of the vehicle based on information regarding a most recently changed control state of the vehicle.

2. The speech dialogue device according to claim 1.

18. A method for controlling a speech dialogue device executed by a computer, comprising: Obtaining information regarding the control state of the vehicle; Estimating the mood of a driver of the vehicle based on the acquired information about a control state of the vehicle; A manner of replying to the driver is set according to the estimated mood of the driver. A method for controlling a voice dialogue device.

Images