JP2003106846A - Agent apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make the meaning or the content of the operation of an agent easily intuited in response to the cultural background of a user's birthplace or the like, his taste or the like. SOLUTION: The operation to be conducted by an agent is not uniformly specified but is done to meet the user's cultural information such as, for example, user's birthplace, taste or the like for the content or the meaning of the communication with a driver. For example, when the agent greets, if the user can specify Japanese from the user information, the 'bowing' operation is conducted, or if the user can be specified a Western user, a 'right hand salute' operation is performed. Thus, the cultural information including the user's taste and favorite is interpreted from the user information, the agent performs the operation corresponding to the cultural background, and hence the user can easily intuit the meaning or the content used at present by the agent from the operation.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an agent device and, for example, to an agent device which has a conversation with a passenger of a vehicle and operates a device.

[0002]

2. Description of the Related Art An agent device for causing an agent to appear in a vehicle, judging the state of the vehicle, and positively communicating with the user, such as a conversation, is disclosed in JP-A-11-37.
No. 766 is proposed. According to this agent device, for example, the remaining amount of fuel such as gasoline and light oil is detected, and the detected values of the past five times immediately before refueling are stored, and when the average value is reached, the agent is the refueling time. It is supposed to let you know. In such an agent device, the agent is caused to perform an action of expressing a specific intention or meaning. For example, an agent is supposed to bow to express a greeting or thank-you. Then, as the operation of expressing such a specific meaning, an operation that allows the user to intuition as much as possible is adopted. The action of the agent expressing a specific meaning is expressed by one action.

[0003]

However, for the user who communicates with the agent,
Even though the meanings are the same, the current situation is that the operations performed differ depending on the cultural background, hobbies, preferences, etc. of the user. For example, a greeting action for a Japanese user is a "bow", whereas a greeting action for a Western European user is a "raise hand" action. Therefore, even if the agent is made to uniformly perform a "bowing" action with respect to the meaning content of the greeting, it is difficult for a Japanese user to understand, but it is difficult for a Western European user to understand the meaning of the action. Similarly, for a user who rarely watches soccer, even if an agent gives out a yellow card or a red card, the meaning is often difficult to understand, so the meaning and content of the operation (the state the agent wants to express) The user must learn.

If the agent does not have to be a human as an appearance, and the octopus and the desk lamp can communicate in the same movement as a human, some users may feel more familiar and interesting. Even for an agent whose appearance is non-human, as in the case of a human agent, it is in the sense to make the user intuitively understand the meaning of the operation by making the operation correspond to the cultural background of the user. preferable.

Therefore, it is a first object of the present invention to provide an agent device in which the meaning and contents of the operation of the agent can be easily understood depending on the cultural background and hobbies of the user. A second object of the present invention is to provide an agent device which makes it easy to intuitively understand the meaning and content of its operation even in the case of a non-human type agent.

[0006]

According to a first aspect of the present invention, agent appearance means for causing an artificial artificial life object to appear in a vehicle compartment, and driver information storage means for sequentially storing various kinds of information about the driver. A situation detecting means for detecting at least one of the situation of the vehicle and the situation of the driver, and an expression state determining means for determining a state desired by the pseudo-life form corresponding to the detection result of the situation detecting means. A state in which the pseudo life form is desired to be expressed by the expression state decision means, a motion decision means for deciding a motion of the pseudo life object based on the stored driver information, and an image of the decided motion. The display means for displaying is provided in the agent device to achieve the above object. According to a second aspect of the present invention, in the agent device according to the first aspect, a human-type action storage unit that defines an action when the pseudo life form is a human type, and the pseudo life form is a non-human type. And a non-human-type action storage unit that defines the action in the case of the above-mentioned action, wherein the action determining unit is based on the action defined in the human-type action storage unit or the non-human-type action storage unit. It is characterized by determining the behavior of living things. In the invention according to claim 3, agent appearance means for causing an artificial artificial life object to appear in the vehicle interior, situation detection means for detecting at least one of the situation of the vehicle and the situation of the driver, and the pseudo. A motion storage unit that defines a motion when the life form is a human type, an expression state determination unit that determines a state desired to be expressed by the pseudo life form in accordance with a detection result by the situation detection unit, and the expression state. A state desired to be expressed by the pseudo life determined by the determination means,
A motion determining means for determining a motion of the pseudo-living body based on the motion storing means; and a display means for displaying the determined motion as an image. When a body is selected, the action is determined by associating each part of the non-human type pseudo life form with the constituent part of the human type pseudo life form. According to a fourth aspect of the invention, in the agent device according to the third aspect, there is provided driver information storage means for sequentially storing various kinds of information regarding the driver, and the operation determining means is the stored driver information. The motion of the pseudo-life is determined based on According to a fifth aspect of the invention, in the agent device according to the first or fourth aspect, the driver information storage means has a driver's origin area, a driver's hobby, occupation, a driver's habit, and academic background. At least one of the above.

[0007]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, preferred embodiments of an agent device of the present invention will be described in detail with reference to FIGS. (1) Outline of Embodiment In the agent device according to this embodiment, an anthropomorphized agent is made to appear in the vehicle by an image (a three-dimensional image such as a planar image or holography), and this agent is connected to the driver. Let them perform various communications and operations on their behalf. Instead of stipulating uniformly the actions that the agent performs with respect to the content and meaning of communication with the driver, perform actions that match the cultural background of the user, for example, the place of origin or hobbies. . For example, when the agent greets, if the user information identifies the user as a Japanese person, the agent performs a "bow" action, and if the agent identifies as a Western user, the agent raises a "right hand" action. In this way, by interpreting the cultural background including the user's hobbies and preferences from the user information and the agent performing an action corresponding to the cultural background, the user can understand the meaning and content that the agent is trying to express. It becomes easy to intuitively understand from the operation. Also, in the case of an agent with a non-human appearance,
The body part of the non-human agent is made to correspond to the human part, and is operated in the same manner as the human agent according to the cultural background of the user. In this case, when the appearance (body structure) of the non-human agent is different from that of human, only the parts that are important and necessary for expression are applied to the parts of the non-human agent.

The user information expressing the cultural background of the user is appropriately input by each user at the time of initial setting. Further, when there is user information that has not been input, when the load on the driving operation of the driver is low (when the steering operation at a certain angle or more is not being performed, when the device such as the audio is not being operated,
(Eg, waiting for a signal), the agent asks a question to obtain user information that has not been input, and collects and saves the user information from the answers to the questions.

The agent in the present embodiment is an artificial pseudo-lifeform, and its appearance (planar image, stereoscopic image such as holography) appears in the vehicle by the image display device. As the processing of this agent, the situation of the vehicle including the vehicle itself, passengers, oncoming vehicles, etc. is judged and learned (including not only the learning of the situation but also the response and reaction of the driver), and the vehicle situation at each time point. Based on the learning results up to that point, the agent responds to the driver and the vehicle with various variations (action = action and voice). As a result, the driver can freely call out a plurality of agents in the vehicle and have a relationship (communication) with each other, thereby making the environment in the vehicle comfortable. Here, the artificial pseudo life form (agent) in the present embodiment has the same identity of a subject such as a specific human being, living thing, or cartoon character, and the pseudo life form having the same identity is the identity of the subject. -Output (action, response by voice) is performed so as to maintain continuity. Further, the identity / continuity is expressed as having a unique individuality, and the agent of the present embodiment that appears in the vehicle interior can recognize the voices that are uttered according to the past learning content even in the same vehicle situation. Images are different.

(2) Details of Embodiment FIG. 1 is a block diagram showing the configuration of an agent device in this embodiment. In this embodiment, the overall processing unit 1 that controls the entire communication function is provided.
The overall processing unit is a navigation processing unit 10 that searches for a route to a set destination and guides by voice or image display, an agent processing unit 11, an external I / F unit 1 for the navigation processing unit 10 and the agent processing unit 11.
2. Image processing unit 13 that processes image output and input images of agent images and map images, voice control unit 14 that controls voice output and input voice of agent voices and route guidance voices, and relates to vehicles and drivers. It has a situation information processing unit 15 that processes detection data of various situations, an input control unit 16, and a storage device control unit 17.

The navigation processing unit 10 and the agent processing unit 11 include a CPU (central processing unit) for controlling data processing and operation of each unit, and a ROM connected to the CPU via a bus line such as a data bus or a control bus. RAM,
Equipped with a timer. Both processing units 10 and 11 are connected to a network (in-vehicle LAN (Local Area Net).
work)), and mutually processed data can be acquired. The ROM is a read-only memory in which various data and programs for controlling by the CPU are stored in advance, and the RAM is a random access memory used as a working memory by the CPU.

In the navigation processing unit 10 and the agent processing unit 11 of this embodiment, the CPU reads various programs stored in the ROM and executes various processes. Note that the CPU is the recording medium drive device 23.
The computer program is read from the external recording medium set in, and stored (installed) in the agent data 30, the navigation data 31, and other storage devices such as a hard disk (not shown), and the necessary programs and the like are stored in the RAM from the storage device. It may be read (loaded) and executed. Also, the necessary programs and the like may be directly read from the recording medium driving device 23 into the RAM and executed.

In this embodiment, the agent processing unit 1
1 creates various communication operations of the agent including conversation with the driver and processing operations performed by the navigation processing unit 10, assuming various situations (scenes) of the vehicle and the driver, results of the driver information, and the like. According to the scenario, In other words, various situations such as vehicle speed, time, running area, temperature, gasoline level, psychological distance, driver information, etc. are set as scenario activation conditions and scene branching conditions, and the behavior of the agent in each situation is set as a scenario for each situation. It is prescribed. Each scenario is composed of a plurality of consecutive scenes. A scene is a scene in a scenario, where an agent asks a question or provides information. Each scene is organized by title, list, balloon, background, and other small units (parts). Each scene proceeds sequentially according to the scenario. Depending on the scenario, there are a plurality of scenes to be selected depending on the driver's answer to a question asked in a specific scene, the situation of the vehicle, and the like. That is, there is a scenario in which a scene branches according to an answer on the way. The data of the scenario including the scene is stored in the scenario data file 302 described later. Information (scenario activation condition) that defines when and where to execute a scenario, what kind of screen configuration is used when it is executed, what kind of operation or conversation the agent has, and the modules such as the navigation processing unit 10 Data that defines what kind of instruction is given and what to do next when an event is received (which scene) is stored as scenario data 1 to n in the scenario data file 302 for each scene.

FIG. 2 shows the configuration of the agent processing section 11. As shown in FIG. 2, the agent processing unit 11 includes a situation determination unit 111, a scenario execution request unit 112, a scenario drive unit 113, a scenario data reading unit 114, a drawing processing unit 115, a voice recognition management unit 116, and a user input. The reception unit 117 is provided. Each of these units 111 to 117 is realized by a corresponding engine (software) stored in a CPU, a ROM, and the like included in the agent processing unit 11. The engine that constitutes each unit such as the situation determination unit 111 issues a corresponding event when transmitting data to another engine. The events issued by the respective units 111 to 117 include an event for determining whether or not to branch the scene.

The status judgment section 111 is an API (Applicat) for notifying a change in status such as time, place and various inputs.
Ion Programming Interface)
Is equipped with. The situation determination unit 111 (destination setting, etc.)
The scenario execution request determination unit 112 responds to changes in the situation.
Issue an event to. Scenario execution request determination unit 112
Determines whether or not to execute the scenario by referring to the situation determination result with the API function. When executing a scenario, a scenario activation request event is issued to the scenario driver 113.

The scenario drive unit 113 determines whether or not to start the scenario activation request event issued by the scenario execution request unit 112. When it is determined that the scenario data is to be activated, the scenario driving unit 113 requests the scenario data reading unit 114 to read the scenario data. In addition, drawing request according to the scenario data,
The drawing processing unit 115 is requested to request a motion with the character. It also notifies the voice recognition management unit 116 of the dictionary used for voice recognition according to the scenario data. Furthermore,
The scenario drive unit 113 drives a scenario (manages scene transitions) based on the result of voice recognition or user input.

The scenario data reading unit 114 reads the corresponding scenario data from the scenario data file 302 based on the event request from the scenario driving unit 113. Further, the image data is read in response to an event request from the scenario drive unit 113. The drawing processing unit 115 draws a corresponding scene or a character (agent) based on an event request from the scenario driving unit 113. The voice recognition management unit 116 controls instructions to the voice recognition unit 142. For example, the usage dictionary notified from the scenario drive unit 113 is designated. The voice recognition unit 116, when the voice recognition unit 142 produces a voice recognition result based on the specified use dictionary, outputs an event corresponding to the recognition result to the scenario drive unit 11.
Issue to 3. The user input reception unit 117 issues an event corresponding to the input content to the scenario drive unit 113 when there is an input from the user. The input from the user is determined by the information from the input control unit 16. It should be noted that although the case where an event is issued and the case where an API function is used has been described in each part of the agent processing unit 11, all may be used to issue an event, or all may use the API function.

In FIG. 1, a storage medium driving device 23 and a communication control unit 24 are connected to the external I / F unit 12, a display device 27 and an image pickup device 28 are connected to the image processing unit 13, and a voice control unit 14 is connected. A voice output device 25 and a microphone (voice acquisition means) 26 are connected to the situation information processing unit 15
Various status detection devices 40 are connected to the input control unit 16
An input device 22 is connected to the.

The various situation detecting device 40 comprises a current position detecting device 41 and a situation detecting section 42. The current position detection device 41 is for detecting an absolute position (based on latitude and longitude) of the vehicle, and measures the position of the vehicle using an artificial satellite GPS (Global Positioni).
ng System) receiver 411 and orientation sensor 412
A steering angle sensor 413, a distance sensor 414, a beacon receiving device 415 that receives position information from beacons arranged on the road, and the like are used. GPS receiver 411
The beacon receiving device 415 and the beacon receiving device 415 can measure the position independently, but the GPS receiving device 411 and the beacon receiving device 415
In a place where the reception by the vehicle is impossible, the current position is detected by dead reckoning using the azimuth sensor 412, the distance sensor 414 and the steering angle sensor. Direction sensor 41
Reference numeral 2 denotes, for example, a geomagnetic sensor that detects the geomagnetism to determine the direction of the vehicle, a gyro such as a gas rate gyro or an optical fiber gyro that detects the rotational angular velocity of the vehicle and integrates the angular velocity to determine the direction of the vehicle, and left and right wheels. A wheel sensor or the like in which a sensor is arranged and the amount of displacement of the azimuth is calculated by detecting the turning of the vehicle based on the output pulse difference (difference in moving distance) is used. The rudder angle sensor 413 detects the steering angle α using an optical rotation sensor attached to the rotating portion of the steering wheel, a rotation resistance volume, or the like. As the distance sensor 414, various methods such as a method of detecting and counting the number of rotations of the wheel or a method of detecting acceleration and integrating twice are used. The distance sensor 414 and the steering angle sensor 413 also function as a driving operation status detecting means.

The status detecting unit 42 functions as a driving operation status detecting means for detecting the status of driving operation, and includes a brake sensor 421, a vehicle speed sensor 422, and a direction indicator detector 42.
3, a shift lever sensor 424, and a side brake (parking brake) sensor 425. The situation detection unit 42 also includes an air conditioner detector 427, a wiper detector 428, and an audio detector 429 that function as a device operation situation detection unit that detects the situation of equipment operation.

The brake sensor 421 detects whether or not the foot brake is depressed. Vehicle speed sensor 422
Detects the vehicle speed. The turn signal detector 423 detects whether the driver is operating the turn signal and whether the turn signal is blinking. Shift lever sensor 424
Detects whether the driver is operating the shift lever and the shift lever position. The side brake (parking brake) sensor 425 detects whether the driver is operating the side brake and the state of the side brake (on or off). The air conditioner detector 427 detects whether the driver is operating various switches of the air conditioner. The wiper detector 428 detects whether the driver is operating the wiper. The audio detector 429 detects whether the driver is operating an audio device such as a radio, a CD player, and a cassette poker, and whether or not a sound is being output by the audio device. In addition, the status detection unit 42 also serves as device operation status detection means, such as a light detection sensor that detects the operation status of lamps such as headlamps and room lamps, a seatbelt detection sensor that detects a driver's seatbelt removal operation. It is equipped with a sensor.

The input device 22 is used for the driver to input driver information and for the driver to respond to all other inquiries from the agent according to the present embodiment.
It is also one means. The input device 22 is a predetermined traveling environment (transmission condition) of a vehicle that wants to transmit a request for information such as traffic jam information to the information providing station, the current location (departure point) and destination (arrival point) at the start of traveling in the navigation processing, It is also for inputting the type of mobile phone used in the car. As the input device 22, various devices such as a touch panel (functioning as a switch), a keyboard, a mouse, a light pen, a joystick, a remote controller using infrared rays, a voice recognition device, and the like can be used. Also,
A remote controller using infrared rays and the like, and a receiving unit for receiving various signals transmitted from the remote controller may be provided. On the remote control, in addition to the joystick for moving the cursor displayed on the screen, menu specifying keys (buttons),
Various keys such as a numeric keypad are arranged. Input control unit 16
Detects data corresponding to the contents input by the input device 22 and detects the data by the agent processing unit 11 or the navigation processing unit 1.
Supply to 0. The input control unit 16 functions as a device operation status detection unit by detecting whether or not the driver is performing an input operation.

The recording medium driving device 23 is a driving device used for reading a computer program for the navigation processing unit 10 and the agent processing unit 11 to perform various processes from an external recording medium. The computer program recorded on the recording medium includes various programs and data. Here, the recording medium means a recording medium on which a computer program is recorded, and specifically, a magnetic recording medium such as a floppy disk, a hard disk, a magnetic tape, a semiconductor recording medium such as a memory chip or an IC card, a CD- A recording medium such as a ROM, MO, PD (Phase Change Rewritable Optical Disk) from which information is optically read, a recording medium using a paper card, a paper tape, or the like, and a recording medium in which a computer program is recorded by various methods are included. .

The recording medium driving device 23 reads the computer program from these various recording media, and when the recording medium is a writable recording medium such as a floppy disk or an IC card, the navigation processing unit 10 The RAM of the agent processing unit 11, the data of the storage device 29, and the like can be written in the recording medium. For example, the IC card is made to store data such as learning contents (learning item data, response data) about the agent function and driver information, and the data is read from the IC card storing these even when driving another vehicle. By using it, it becomes possible to communicate with the agent in the learned state according to his / her past response situation. As a result, not the agent for each vehicle but the agent having the learning content unique to each driver can appear in the vehicle. The scenario data file may be stored in the IC card.
This makes it possible to create an original scenario unique to each user.

The communication control unit 24 is adapted to be connected to a mobile phone including various wireless communication devices. The communication control unit 24 communicates with an information providing station that provides data relating to traffic information such as road congestion and traffic regulation in addition to telephone calls over telephone lines, and karaoke data used for in-car communication karaoke. It is possible to communicate with the information providing station that provides the. In addition, learning data regarding the agent function,
It is also possible to send and receive scenario data files and the like.

The voice output device 25 is composed of a plurality of speakers arranged in the vehicle. The voice controlled by the voice control unit 14, for example, a guidance voice for voice route guidance and a driver by an agent. A voice or sound is output according to a normal conversation for communication or a question for acquiring driver information of the present embodiment.
The audio output device 25 may also be used as an audio speaker. The microphone 26 is the voice control unit 1.
4 functions as a voice input unit for inputting and outputting a voice to be voice-recognized, for example, an input voice of a destination or the like in navigation processing, a driver's conversation with an agent (including a response by the driver), and the like. This microphone 2
6, a dedicated microphone having directivity is used in order to accurately collect the driver's voice. The voice output device 25 and the microphone 26 may form a hands-free unit so that telephone communication can be performed without using a mobile phone. The microphone 26 and the voice recognition unit 142 function as a conversation detection unit that detects whether or not the driver is in conversation with the fellow passenger. In this case, the microphone 26 and the voice recognition unit 142.
Functions as a situation detection unit that detects the situation of the driver.

The display device 27 displays a road map and various image information for route guidance by the processing of the navigation processing unit 10, and various parts (parts) of the agent action (moving image) screen configuration by the agent processing unit 11. )
Is displayed. Further, the image inside and outside the vehicle captured by the image capturing device 28 is also displayed after being processed by the image processing unit 13. Display device 2
As the display device 7, various display devices such as a liquid crystal display device and a CRT are used. The display device 27 may have a function as the input device 22, such as a touch panel.

The image pickup device 28 is a C for picking up an image.
It is composed of a camera equipped with a CD (charge coupled device), and in addition to an in-vehicle camera for imaging the driver, each out-of-vehicle camera for imaging the front, rear, right side and left side of the vehicle is arranged. The image captured by each camera of the image capturing device 28 is supplied to the image processing unit 13 and subjected to processing such as image recognition, and each recognition result (presence / absence of passenger, recognition of driver, etc.) is communicated to the agent. It is supposed to be reflected.

The storage device 29 is connected to the storage device control section 17 so that data and programs can be read and written under the control of the storage control device 17. The storage device 29 stores agent data 30 and navigation data 31 as various data (including programs) necessary for realizing various agent functions and navigation functions according to this embodiment. The storage device 29 includes various recording media such as a floppy disk, a hard disk, a CD-ROM, an optical disk, a magnetic tape, an IC card, an optical card, and the like.
The drive is used. In this case, for example, the learning item data 304, the response data 305, the driver information 30.
7. The scenario data file 302 is composed of a plurality of different recording media and drive devices, such as an IC card or a floppy disk that is easy to carry, and other data such as a hard disk. A storage medium may be used.

The agent data 30 includes an agent program 301, a scenario data file 302,
Voice data 303, learning item data 304, response data 305 composed of voice data 303, image data 306 for displaying the appearance and behavior of the agent, driver information 307, semantic classification table 308, profile table 309, agent Correspondence table 310,
Stores various data required for processing by other agents.

The agent program 301 stores an agent processing program for realizing the agent function.

Learning item data 304 and response data 30
5 is data that stores the result learned by the agent according to the driving operation and response of the driver. Therefore, the learning item data 304 and the response data 305 are stored / updated (learned) for each driver. The learning item data 304 stores, for example, items to be learned by the agent, such as the total number of times the ignition is turned on, the number of times the vehicle is turned on per day, and the data of the remaining fuel amount at the time of refueling the previous five times. According to the learning content stored in the learning item data 304, for example, the greeting content at the time of appearance of the agent changes according to the number of times the ignition is turned on, or the fuel remaining amount becomes less than or equal to the average value of the fuel remaining amounts of the previous five times. In some cases, the agent makes a refueling proposal. The response data 305 stores a history of user responses to agent actions for each predetermined scenario. In the response data 305, the response date and time and the response content are stored a predetermined number of times for each response item. Ignored as the response content,
Each of the cases of rejected and accepted (accepted) is judged from the voice recognition and the input result of the input device 22 and stored.

The scenario data file 302 stores scenario data in which the actions of the agent in each situation and scene are defined. FIG. 3 conceptually shows the data structure and data contents of the scenario data file 302. The scenario data file 302 is
A plurality of scenario data defining each scene are collected to form one file. The scenario data file 302 includes a file header, a scenario management table, an automatic start condition table, a list of manually startable scenarios, each scenario data 1 to n, and additional resource data.

The file header is the header of the scenario file and stores information about the file. The scenario management table stores data for managing the scenario data recorded in the scenario file, for example, data such as the priority order when executing each scenario data.

In the automatic start condition table, condition data for automatically starting the scenario data recorded in the scenario file is defined and stored for each scenario data. As the automatic start condition, for example, a specific point and distance data regarding a specific scenario to be executed when a point at a predetermined distance from the specific point A, or an ignition on condition for executing a scenario in which an agent greets at startup is turned on. Various automatic start conditions such as, are stored. Further, as the automatic start prohibition condition for prohibiting the automatic start, for example, (A) manual start is prioritized, (B) automatic start is not performed while another scenario is being executed, (C) execution is performed within 30 minutes, for example. Some scenarios don't start,
(D) Does not start when listening to music, (E) Does not start while turning the curve right or left (during steering operation),
(F) Does not start during sudden deceleration or sudden acceleration, (G)
It does not start while talking to a passenger, (H) It does not start while operating various devices such as a navigation device, audio, air conditioner, shift lever, side brake, wiper, turn signal, (I) Audio Automatic start prohibition conditions such as not starting during use (condition for not disturbing listening to music or broadcasting: audio is not on), (J) not starting during guidance guidance of the travel route by the navigation processing unit 10, etc. Is also stored. However, even if these automatic start prohibition conditions are satisfied, the priority order is defined in the scenario management table so as to be automatically started in an emergency (a scenario of an emergency contact to a fire department etc.). The individual automatic start prohibition conditions include a condition commonly applied to each scenario and a condition applied to an individual scenario.

In this embodiment, the condition for activating the question scenario when the agent asks a question in order to collect the driver information is a condition for judging that the load condition of the driver is a low load condition. The following conditions and other conditions are specified. (A) Do not execute a scenario for asking questions within one hour (conditions for not asking too many questions) (b) 15 minutes or more have passed since ignition was turned on (immediately after starting driving (C) There is no conversation in the vehicle (by collecting sound from the microphone 26) (d) There is no speed change for a predetermined time on a straight road (detected from the navigation data 31 and the current position). (E) Waiting for a signal (navigation data 3
1. Judgment from current position and vehicle speed) These question activation conditions are determined as to whether any one of the indispensable conditions for activation should be satisfied. For example, limited to the above conditions (a) to (e),
When all of (a) to (c) are satisfied and (d) or (e) is satisfied, the activation condition of the question scenario is satisfied, and the question scenario is activated. Even in this case, if the automatic activation prohibition conditions (A) to are satisfied, activation of the question scenario is prohibited. The start determination routine based on the automatic start condition including the question start condition is, for example, processed every minute or when a predetermined event is received. The events designated in advance include destination setting, mail arrival, and the like.

Of the above conditions, condition (E),
(F), (H), (J), (d), (e) determines the load on the driving operation, and conditions (C), (D),
From (G), (I), (J), (a), (b) and (c), the degree of annoyance felt by the driver is determined.

The manual startable scenario list stores list data of scenarios that can be manually started in the scenario data recorded in the scenario file. The additional resource data stores voice data and image data to be added to the scenario data for expanding the storage device, cultural background as driver information, and the like.

Each scenario data 1 to n is composed of a scenario header, a voice recognition dictionary DB, an image data management table, a scene management table, each scene data, and actual image data. Each scenario consists of multiple scenes.
The scenario header is a header section of scenario data and stores information about scenarios. The voice recognition dictionary DB is a voice recognition dictionary DB used in the scenario.
Thus, the voice recognition dictionary data is stored. The image data management table stores data for designating an image used in a scenario. The image of the agent is stored in the image data 306, and the image unique to the scenario is stored in the actual image data at the end of the scenario data. The scene data management table stores data for managing the scene data recorded in the scenario.

Each scene data 1 to m is a scene header,
It is composed of screen configuration data, character action instruction data, various processing instruction data, and a scene branch table.
The scene header is a header section of scene data, and stores information about the scene and data for managing each data section belonging to the scene data. The screen configuration data is a section that defines the screen configuration displayed in the scene, and stores data of each part of the screen configuration displayed on the display device 27. The character action instruction data stores instruction data relating to the action instruction of the character in the scene and instruction content (data instructing the voice data 303 to be used). As the various processing instruction data, data for requesting a processing instruction to another module (for example, navigation) in the scene is stored. The scene branch table is a section that stores branch information from the scene to the next scene, and stores data that defines the next scene when an event is received in the certain scene.

FIG. 5 shows an example of a scene screen displayed on the display device 27 based on the scene data. The scene screen shown in FIG. 5 is a scene screen of a question scenario in which a driver asks a question in order to acquire a hobby preference (meal) which is uninput driver information (scene number 0).
x0001). As shown in FIG. 5, the scene screen includes an agent display screen 51 on which an agent image (still image, moving image) is displayed, a balloon screen 52 on which characters corresponding to the agent's voice are displayed, a title screen 53, and , A scene display screen 54 on which image data specific to each scene (image of actual image data, answer selection button, etc.) is displayed.

When the hobby / preference (meal) question scenario is activated, the agent processing section 11 first reads out the screen configuration data of the scene designated by the scene header from the scenario data file 302 and displays the scene screen 2
7 is displayed and a question voice corresponding to the question sentence is output from the voice output device 25. Figure 5 (a)
In the scene screen of the question scenario of, "Which genre of meal do you like?" Is displayed on the balloon screen 52.
The voice output device 25 outputs a voice corresponding to the display of the balloon screen 52. Further, in the scene display screen 54 in the scene screen of FIG. 5 (a), "Japanese food", "Western food", "Chinese food", and "None" are displayed on the four answer selection buttons 54a.

A plurality of scenes corresponding to the driver's answer branch off and follow the scene of the question to the driver according to the scene management table. That is,
When the driver selects the answer selection button “Japanese food” on the scene screen (scene number 0x0001) of FIG. 5A, the agent processing unit 11 is branched and displayed on the scene screen (b) corresponding to the answer. In this scene screen (b), the selected “Japanese food” is displayed on the title screen 53, and the balloon screen displays “I like Japanese food, isn't it?” In addition, on the scene screen of Japanese food after branching An actual image of Japanese food is read from the scenario data and displayed on the scene display screen 54. Then, the agent processing unit 11 stores the driver's answer, for example, “Japanese food” as driver information in the hobby / preference data of the driver information 307. In this way, each scene image and sound specified in the scenario are continuously displayed and output until the last scene, whereby the action of the agent in one scenario is completed.

Voice data 303 in the storage device 29 (FIG. 1)
Stores voice data for the agent to have a conversation with the driver according to the scene of the selected scenario. The voice data of conversation by the agent also stores voice data for the agent to ask a question for collecting driver information according to the present embodiment.
Each data of the voice data 303 is designated by the character motion instruction data of the scene data.

The image data 306 includes a still image showing the state of the agent used in each scene designated by the scenario, a moving image (animation) showing the action, and 3
D model data, motion data, etc. are stored. For example, a moving image in which the agent bows, a nodding moving image, a moving image in which the right hand is raised, 3D model data that configures the appearance of the agent and motion data that defines its behavior are stored. An image code is attached to each of these still images and moving images. Image data 3
The appearance of the agent stored in 06 does not have to be a human (male, female) appearance. For example, octopus, chick, dog,
It may be the appearance of animals such as cats, frogs, rats, etc., or the appearance of humanly designed (illustrated) animals, and even robotic appearance, such as floor stands and trees, and specific appearances. It may be the appearance of the character. In addition, the age of the agent does not have to be constant, and as the learning function of the agent, the appearance of the child is initially set, and the appearance grows and changes over time (the appearance of an adult changes, and It will change to your appearance). The image data 306 stores images of the appearances of these various agents, which can be selected from the input device 22 or the like according to the driver's preference.

The driver information 307 is information about the driver and is used to make the action by the agent more suitable to the driver's wishes, hobbies, and preferences. FIG. 5 conceptually shows the configuration of the driver information 307.
As illustrated in FIG. 5, the driver information 307 includes
Driver's ID (identification information), name, nationality, age, gender,
Marriage (married or unmarried), driver basic data including presence / absence of children, number of people, and age, and hobby / preference data are stored. In addition to this, data that shows what kind of education has been received, for example, the educational background of the university where the student came from, the past residence, the religion, etc. are stored. The hobby and preference data is composed of major items such as sports, eating and drinking, and travel, and detailed items included in the concept of these major items. For example, the large-item sport stores detailed data such as a favorite soccer team, a favorite baseball team, and interest in golf. In the present embodiment, priorities are determined for each of these pieces of driver information, and the agent asks the driver questions in descending order of priority among the driver information that has not yet been stored. . The driver basic data has a higher priority than the hobby and preference data.

The driver information 307 is created for each driver when there are a plurality of drivers who drive the vehicle. Then, the driver is specified and the corresponding driver information is used. The driver is specified by, for example, when the ignition is turned on, an agent common to all the drivers appears, inquires of the driver, and specifies from the answer. For the driver's inquiry, a button for selecting the name of the driver who has already entered and other information is displayed on the display device, and a voice prompting the selection is output. If other is selected, the new user registration screen is displayed. In addition, weight, fixed position of driver's seat (front and rear position, angle of backrest), angle of rearview mirror,
At least one driver-specific information such as the height of the line of sight, the data obtained by digitizing the face photograph, and the characteristic parameters of the voice is stored in the driver information 307, and the driver is specified from these information. Good.

FIG. 6 conceptually shows the contents of the meaning classification table 308 and the profile table 309. FIG. 6A conceptually shows the contents of the meaning classification table 308. This meaning classification table 308
Is a table in which unique meanings of output contents such as greetings of greetings, affirmations, and gratitudes (animation meanings, expressions, states that the agent wants to express) that can be associated with the actions of the agent are held. FIG. 6A shows an example of simplified table contents, and the actual table is defined in more detail. A semantic code is individually attached to each meaning of each agent operation.

6B and 6C conceptually show the contents of the profile table 309. It is a table that defines which animation is selected when outputting each item of the meaning of the meaning classification table 308. In each profile table 309, the action code corresponding to the meaning code and the image data 3 corresponding to the action code are stored.
An image code for specifying a moving image (animation) or still image (hereinafter simply referred to as moving image) to be selected from 06 is stored. In FIG. 6, in order to facilitate understanding, not the operation code but the content of the agent operation corresponding to each operation code is displayed. In the profile table 309, a plurality of tables are defined according to various cultural backgrounds of the user, and FIG. 6 “Human character animation for Japanese” (b) and “Human character animation for Western” are shown. (C) is illustrated.
Which profile table 309 is used is
Interpreted and determined from user information. If the user information is not sufficiently accumulated, the default value is, for example, "human type character animation for Japanese people". The default value may be, for example, a value according to the country in which the agent device is used, in addition to the value for Japanese. For example, when the agent greets an encounter, if the user is Japanese, FIG.
The profile table 309 is selected and the video to bow is started, but the profile table 309 of (c) is selected for another user, an American, and the video of the motion of raising hands is started. become. As described above, by preparing various profile tables 309 according to the cultural background of the user, the agent can perform an operation that the user can intuitively understand.

The profile table 309 is stored for a humanoid agent as an operation of the agent according to the cultural background. On the other hand, there are cases where it is desired to use a non-human character that does not necessarily have a part similar to a human as an agent. Even in such a case, if all profile tables are created, not only will the number of tables become enormous, but many related tables will be modified when a specific operation is modified (such as changing the corresponding image code). I need it. Therefore, in the present embodiment, by registering the action of the non-human character corresponding to each action defined for the human character in the agent correspondence table 310, various non-human characters can be registered. I have to.

FIG. 7 shows an agent correspondence table 310.
Is a conceptual representation of the contents of. The agent correspondence table 10 defines the action content of the non-human character corresponding to the action content (action code) of the human character. That is, the motion code and the image code indicating the motion content of the non-human character are stored corresponding to the motion code of the human character. The moving image corresponding to the image code is stored in the image data 306 of the agent data 30. In FIG. 7, as an example of the non-human character, the agent correspondence table (motion correspondence table) for the floor stand type character shown in FIG. 10 is displayed with specific motion contents. The agent correspondence table is defined for each non-human character.

Agent correspondence table 31 shown in FIG.
With 0, the motion of shaking the head of the human character is a motion of shaking the lamp shade of the non-human character. As described above, with the agent correspondence table 310 shown in FIG. 7, even if there is no profile table for non-human characters, each movement for Japanese people shown in FIG. Converted to type character behavior. Further, the operation for Westerners shown in FIG. 6C is converted into each operation shown in FIG.

The navigation data 31 (see FIG. 1) includes various data files used for route guidance, such as communication area data files, drawing map data files, intersection data files, node data files, road data files, and search data. Various data such as files and photo data files are stored.

In the communication area data file, a mobile phone which is connected to the communication control unit 24 or is not connected and is used in the car displays the communication area on the display device 27 or routes the communication area. Communication area data to be used for searching is stored for each type of mobile phone. The drawing map data file contains the display device 2
7 stores drawing map data to be drawn. This drawing map data stores a hierarchical map, for example, map data for each layer such as Japan, Kanto region, Tokyo, Kanda from the highest layer. A map code is attached to the map data of each layer.

In the intersection data file, the intersection number for identifying each intersection, the name of the intersection, the coordinates (latitude and longitude) of the intersection, the number of the road where the intersection is the starting point or the ending point, and the presence or absence of a signal are included. It is stored as data. The node data file stores node data including information such as latitude and longitude that specify the coordinates of each point on each road. That is, this node data is data relating to one point on the road, and when connecting the nodes is called an arc, the road is represented by connecting each of the plurality of node strings by an arc. In the road data file, the road number that identifies each road, the intersection number that is the start point and the end point, the number of the road that has the same start point and the end point, the thickness of the road, the prohibition information such as entry prohibition, the photograph of the photo data described later The number etc. are stored. Intersection data stored in the intersection data file, node data file, and road data file,
Road network data composed of node data and road data is used for route search.

The search data file stores intersection string data, node string data, and the like that make up the route generated by the route search. The intersection row data is composed of information such as an intersection name, an intersection number, a photograph number showing a characteristic landscape of the intersection, a corner, and a distance. In addition, the node string data includes information such as east longitude and north latitude indicating the position of the node. The photograph data file stores photographs of each intersection and characteristic landscapes that can be seen while going straight ahead, in the form of digital, analog, or silver halide photographs corresponding to the photograph numbers.

Next, the operation of the agent device thus configured will be described. FIG. 8 shows the operation when the scenario is executed by the agent device. The agent processing unit 11 acquires the current status and the status change from various data supplied from the status information processing unit 15,
It is determined whether or not the scenario activation condition stored in the automatic activation condition table of the scenario data file 302 is satisfied, that is, whether or not a scenario that can be activated exists in the acquired situation. When the scenario activation condition other than the inquiry scenario activation condition is satisfied, the agent processing unit 11 executes the scenario corresponding to the satisfied scenario activation condition.

That is, the scenario data reading unit 11
4 is the scenario data file 302 which is the character action instruction data of the scenario data satisfying the scenario activation condition.
Read from. Then, the scenario data reading unit 114
Reads the meaning code corresponding to the read character action instruction data from the meaning classification table 308 and outputs the meaning code to the drawing processing unit 115.

The drawing processing unit 115 performs the meaning code processing and the animation processing according to the received meaning code. That is, in the meaning code processing, the driver information 307
The data regarding the driver's cultural background (for example, nationality) is checked to identify the profile table 309. Then, the animation code (image code) corresponding to the meaning code output by the scenario drive unit 113 is read from the specified profile table 309. When the user specifies a non-human character as an agent (this character designation information is stored in the driver information 307), the action corresponding to the meaning code from the profile table 309 of the specified human character. The code is read, and the animation code (image code) corresponding to the operation code is read from the agent correspondence table 310 and output. The agent correspondence table 310 is specified from the character designation information stored in the driver information 307.

Then, in the animation processing, the drawing processing unit 115 reads out the animation data corresponding to the read animation code from the image data 306, outputs it to the image processing unit 13 as an image of each scene, and displays it on the display device 27. .

As described above, a plurality of profile tables 309 in which meanings and actions are associated with each other are prepared according to the cultural background of the user, the cultural background of the user is checked from the driver information 307, and the corresponding profile table 309 is checked.
To display the behavior of the agent. Agents behave differently depending on their cultural background even when they convey the same meaning. FIG. 9 shows the difference in the behavior of the agents depending on the cultural background, for example, in the case where the agent gives a "greeting for meeting". As shown in this figure, if the user is Japanese, the profile table 309 for Japanese is selected for the semantic classification “encounter greeting”, and a moving image “bowing” is displayed on the display device 27. It On the other hand, if the user is a Western user, the Western profile table 309 is selected,
A moving image of the "raising hands" operation is displayed on the display device 27. In this way, since the agent performs an action that matches the cultural background of the user, the user can intuitively understand the meaning of the action of the agent.

FIG. 10 and FIG. 11 show the operation of the semantic classification “encounter greeting” in the desk lamp type character which is a non-human type character. The action by the non-human character is performed by the agent correspondence table 310.
By motion conversion based on. FIG. 10 shows the conversion when the user is Japanese. This FIG. 10 (a)
As shown in FIG.
(FIG. 7), the bowing motion of the humanoid character is converted into a motion of tilting the upper part forward in the humanoid character, and as shown in (b), the desk lamp is corresponding to the “bowing” motion of the humanoid character. The type character displays a moving image in which the upper part including the lamp shade is tilted forward.

FIG. 11 shows the conversion when the user is a Westerner. As shown in this FIG.
From the agent correspondence table 310, as shown in (b), a motion of raising a hand, which is a greeting of a Westerner's encounter,
It is converted into a motion of raising the pedestal (hand of the desk lamp) corresponding to the hand, and the moving image is displayed.

Next, a second embodiment of the present invention will be described. In the first embodiment described above, the image data 3
A still image and a moving image (animation) corresponding to each action of the agent are stored in 06, and the animation is displayed. On the other hand, in the second embodiment, an image of the constituent parts (head, chest, upper arm, etc.) of each character is stored in the image data 306, and the relationship between the constituent parts and their movements are defined to define a profile table. It is stored in 309. Similar to the first embodiment, a plurality of profile tables 309 are defined according to cultural backgrounds such as those for Japanese people and those for Westerners based on human-type characters. FIG. 12 shows the contents of the profile table 309 in the second embodiment. 12
(B) is a profile table for Japanese users,
(C) shows a profile table for Western users. Both profile tables 309 are associated with the meaning code of the meaning classification table 308, as in the first embodiment (see FIG. 6).

In the profile table in the second embodiment, the reference parts 1, 2,
, 3 specify which constituent parts of the agent image to correspond to, and how to move each constituent part. By defining in this way,
Since it is not necessary to save the animation image for each operation individually, the amount of data can be reduced.

FIG. 13 conceptually shows the contents of the agent correspondence table 310 in the second embodiment. In the agent correspondence table 310 in the first embodiment, the motion of the non-human character corresponding to each motion of the human character that is the reference character is defined for each character. On the other hand, in the second embodiment, the correspondence between the constituent parts of the human character and the constituent parts of the non-human character is defined as shown in FIG. Which part of the human character as the reference character and which part of the non-human character corresponds to which configuration is separately defined as illustrated in FIG. 14, and is defined in the agent correspondence table 310. Then, according to the agent correspondence table 310 of FIG. 13, each operation of the profile table 309 of FIG. 12B is converted into each operation of FIG. 12D and displayed as the operation of the non-human agent. Similarly, the profile table 309 of FIG. 12C is converted into each operation of FIG. 12E and displayed.

Although the embodiment of the agent device of the present invention has been described above, the present invention is not limited to the described embodiment, and various modifications can be made within the scope of the claims. is there. In the described embodiment, the difference in motion recognition between the Japanese user and the Western user due to the difference in cultural background based on the region such as nationality and place of origin has been described, but based on other cultural backgrounds of the user. Profile table 309
May be defined.

For example, the profile table 309 may be defined according to the relationship between the agent and the user. FIG. 15 exemplifies the difference in the operation of the greeting of the encounter depending on whether the user and the agent are close or unfriendly. As shown in FIG. 15, the profile table 309 is defined such that the agent acts as a bow if the user and the agent are close to each other with respect to the semantic classification “meeting greeting”, and bows deeply if the user is not close. Note that, in FIG. 15, the example is for a Japanese user, and similarly for a Western user, the operation may be changed according to the level of familiarity.

Here, the degree of familiarity between the user and the agent can be determined by the psychological distance between the agent and the user. Here, the psychological distance between the agent and the driver is represented by parameters such as friendliness, confidence, obedience, and morality, and is changed by the driver's correspondence and the like. For example, if a word of disrespect is given to the agent's action,
The agent's psychological state should increase friendship. On the other hand, if the agent's suggestion is negatively answered or denounced, the agent's psychological state is made less friendly. In this case, the parameter indicating the psychological distance is stored in the agent data 30, and the value is updated according to the reaction of the user.

If there is an action that can be intuitively understood from the user information in the action used by the user's hobby or occupation, the correspondence between the meaning and the action may be defined in the profile table 309. . FIG. 16 shows a difference in operation of agents that notify the departure of the route when the vehicle deviates from the guidance route while the navigation route is being guided. That is, if the user has a hobby of golf, a video of the action of the agent making a circle with both hands is displayed in the meaning of being out of the route. Display a moving image (or still image) of the posture (especially the posture without gesture).

If the user is familiar enough, the action of fictitious culture may be used. For example, if the user is familiar with science fiction movies, a video of an agent performing a fictional salute action used in a play as a greeting expression is displayed.

Further, although the floor-standing type character has been described as an example of the non-human type agent (character) in the described embodiment, another non-human type character may be used. In this case, as shown in FIG. 17, the constituent parts of the operation performed by the agent may be changed depending on the cultural background. FIG. 17 shows the correspondence relationship between the constituent parts of the octopus character and the human parts as the non-human type agent for Japanese users and users other than Japanese users other than Japan. As shown in FIG. 17, in the case of a Japanese user, as shown in (a), an agent correspondence table 310 is defined in addition to recognizing an octopus as two parts of a head, an arm, and a state. On the other hand, in the case of Asian users other than Japan, as shown in FIG. 17B, the agent correspondence table 310 is defined in addition to the recognition of hair, face and torso arm separately.

[0073]

According to the present invention, the operation of the agent is changed according to the cultural background and hobbies of the user, so that the user can intuitively understand the meaning and content of the operation of the agent. Further, according to the present invention, even in the case of a non-human type agent, it becomes easy to intuitively understand the meaning and content of the operation.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an agent device according to an embodiment of the present invention.

FIG. 2 is a configuration diagram of an agent processing unit in the agent device.

FIG. 3 is an explanatory diagram conceptually showing the data structure and data contents of a scenario file in the agent device.

FIG. 4 is an explanatory diagram showing a scene screen displayed on a display device based on scene data of a scenario.

FIG. 5 is an explanatory diagram conceptually showing the configuration of driver information in the agent device.

FIG. 6 conceptually shows the contents of a meaning classification table 308 and a profile table 309.

FIG. 7 is data conceptually showing the contents of an agent correspondence table 310.

FIG. 8 shows an operation when a scenario is executed by the agent device.

FIG. 9 is a diagram showing a difference in agent behavior depending on cultural backgrounds, as an example of a case where an agent gives a “greeting for an encounter”.

FIG. 10 is an explanatory diagram of a conversion operation of a desk lamp character for a Japanese user in the agent device.

FIG. 11 is an explanatory diagram of a conversion operation of a desk lamp type character for a Japanese user in the agent device.

FIG. 12 is an explanatory diagram showing the contents of a profile table in the second embodiment of the agent device.

FIG. 13 is an explanatory diagram conceptually showing the contents of an agent correspondence table in the second embodiment of the agent device.

FIG. 14 is an explanatory diagram showing correspondence between constituent parts of a human character and a non-human character in the agent device.

FIG. 15 is an explanatory diagram showing the difference in the greeting operation of the agent due to the difference in the intimacy between the agent devices.

FIG. 16 is an explanatory diagram exemplifying an operation that can be intuitively understood by an operation used in a user's hobby or occupation in the agent device.

FIG. 17 is an explanatory diagram showing the correspondence between the constituent parts of the octopus character and the human parts in the agent device for Japanese users and Asian users other than Japan.

[Explanation of symbols]

1 Overall processing unit 10 Navigation processing unit 11 Agent processing unit 12 External I / F section 13 Image processing unit 14 Voice control unit 15 Situation Information Processing Department 16 Input control unit 17 Storage device control unit 22 Input device 24 Communication control unit 25 audio output device 26 microphone 27 Display 28 Imaging device 29 storage 30 agent data 31 Navigation data 40 Various status detectors 41 Current position detector 42 Status detector

Continued front page    (72) Inventor Hiroaki Kondo             2-19-12 Sotokanda, Chiyoda-ku, Tokyo Stock             Inside the company Equus Research (72) Inventor Manabu Matsuda             2-19-12 Sotokanda, Chiyoda-ku, Tokyo Stock             Inside the company Equus Research (72) Inventor Kazuhide Adachi             2-19-12 Sotokanda, Chiyoda-ku, Tokyo Stock             Inside the company Equus Research (72) Inventor Tadashi Hirano             2-19-12 Sotokanda, Chiyoda-ku, Tokyo Stock             Inside the company Equus Research F term (reference) 2C032 HB02 HB22 HB24 HB25 HC01                       HC08 HC14 HC15 HC16 HC27                       HC31 HD16 HD23                 2F029 AA02 AB01 AB07 AB13 AC02                       AC13 AC16 AC18                 5E501 AA13 AA23 AC03 AC34 BA13                       BA15 CA02 CB15 EA21 EB05                       EB18 FA14 FA46 FB25                 5H180 AA01 BB05 CC02 CC12 FF04                       FF05 FF12 FF13 FF22 FF25                       FF27 FF33 FF40

Claims (5)

[Claims] 1. An agent appearance means for causing an artificial artificial life object to appear in a vehicle compartment, a driver information storage means for sequentially storing various kinds of information about a driver, and a vehicle situation and a driver situation. Situation detecting means for detecting at least one; expression state determining means for determining a state desired to be expressed by the pseudo life object in response to a detection result by the situation detecting means; and the pseudo life determined by the expression state determining means. A state in which the body wants to be expressed, a motion determining unit that determines a motion of the pseudo-life form based on the stored driver information, and a display unit that displays the determined motion as an image. Agent device that does. 2. A human-type action storage unit that defines an action when the pseudo-life is a human type, and a non-human action memory that defines an action when the pseudo-life is a non-human type. Means, wherein the motion determining means determines the motion of the pseudo-living body based on the motion defined in the human-type motion storage means or the non-human-type motion storage means. 1
The agent device described in 1. 3. An agent appearance means for causing an artificial artificial life object to appear in a vehicle compartment, a situation detection means for detecting at least one of a vehicle situation and a driver's situation, and the artificial life object is a human being. A motion storing unit that defines a motion in the case of a type, an expression state determining unit that determines a state that the pseudo life object wants to express corresponding to a detection result by the situation detecting unit, and an expression state determining unit A motion deciding means for deciding a motion of the quasi-living object based on the motion storing means, and a display means for displaying the decided motion as an image. When the driver selects a non-human type pseudo life form, the operation is determined by associating each part of the non-human type pseudo life form with the constituent part of the human type pseudo life form. Agent device to collect. 4. A driver information storage means for sequentially storing various kinds of information about the driver, wherein the motion determining means determines a motion of the pseudo-life form based on the stored driver information. The agent device according to claim 3, wherein the agent device is provided. 5. The driver information storage means is at least one of a driver's area of origin, a driver's hobby, occupation, driver's habit, and educational background. The agent device according to claim 4.