Information notification control device, information notification control program, driver assistance system, AI model generation method, and AI model generation program

Abstract

This technology provides a way to reduce the size of the AI ​​model used for controlling information notifications. [Solution] The exemplary information notification control device is an information notification control device that controls the notification of information, and obtains the value of a group variable given to each group determined according to the content of the notification of the acquired information, estimates notification parameters by executing an AI model based on the value of the group variable and the values ​​of other input variables, and controls the notification based on the notification parameters.

Description

Technical Field

[0001] The present invention relates to a technique for performing notification control of information using an AI (Artificial Intelligence) model.

Background Art

[0002] Conventionally, a notification system that gives notifications for purposes such as driving support to a driver (user) has been mounted on a vehicle (see, for example, Patent Document 1). As disclosed in Patent Document 1, the means for information notification in a vehicle are diverse. For example, display, sound, and vibration are used.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In a vehicle, information notification has diverse notification means (display, sound, vibration, etc.) as described above, and also diverse types of information to be notified. For this reason, the control of information notification is not easy, and it is even more difficult to perform notification control that suits the preferences of users such as drivers. In view of such circumstances, it is conceivable to apply an AI model to the control of information notification.

[0005] However, when applying an AI model to a notification system mounted on a vehicle, in consideration of the performance of the processor and memory for arithmetic processing mounted on the vehicle, etc., it is desirable that the capacity of the AI model be made as small as possible.

[0006] In view of the above points, an object of the present invention is to provide a technique capable of reducing the capacity of an AI model used for notification control of information.

Means for Solving the Problems

[0007] An exemplary information notification control device of the present invention is an information notification control device that controls the notification of information, and obtains the value of a group variable given to each group determined according to the content of the notification of the acquired information, estimates notification parameters by executing an AI model based on the value of the group variable and the values ​​of other input variables, and controls the notification based on the notification parameters. [Effects of the Invention]

[0008] In this exemplary invention, the input variables related to the content of notifications (notification content) input to the AI ​​model are assigned not to each notification content, but to groups determined according to the notification content. Therefore, according to this exemplary invention, the number of possible values ​​for the input variables related to the notification content input to the AI ​​model can be reduced. As a result, the capacity of the AI ​​model used for controlling information notifications can be reduced. [Brief explanation of the drawing]

[0009] [Figure 1] Diagram showing the general configuration of the driver assistance system. [Figure 2] Block diagram showing the configuration of the functional unit of the controller of the information notification control device. [Figure 3] Figure showing an example of a group variable conversion table. [Figure 4] Schematic diagram to explain the HMI control model. [Figure 5] This figure shows specific examples of inputs and outputs in an HMI control model. [Figure 6] A flowchart illustrating the flow of notification control processing performed by the information notification control device. [Figure 7] A schematic diagram illustrating the processing flow in the HMI control model. [Figure 8] Diagram illustrating the HMI control model related to the modified version. [Figure 9] Block diagram showing the schematic configuration of the AI ​​model generation device. [Figure 10] A diagram showing an example of data with correct labels. [Figure 11] A flowchart illustrating the process of generating an AI model. [Figure 12] A flowchart illustrating the group classification process. [Modes for carrying out the invention]

[0010] Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the drawings. In the drawings, the same or corresponding parts will be denoted by the same reference numerals and will not be repeated in the description.

[0011] <1. Driver assistance systems> Figure 1 shows a schematic configuration of a driver assistance system 100 according to an embodiment of the present invention. The driver assistance system 100 is a system that assists in the driving of a vehicle VE. The driver assistance system 100 can also be described as an information notification system that notifies the driver of the vehicle VE of information related to driving. The information notified may include not only information related to driving, but also entertainment information such as music and movies, and in such a configuration, the system indicated by reference numeral 100 may be understood as an information provision system. When configured as an information provision system, the provision (notification) of information related to driving is not essential, and the information provision system is not limited to vehicles, but can be applied to various uses such as homes, factories, or offices.

[0012] As shown in Figure 1, the driver assistance system 100 comprises an information notification control device 1, an information source device 2, and an information notification device 3. In this embodiment, the information notification control device 1, the information source device 2, and the information notification device 3 are mounted on the vehicle VE and are configured to communicate with each other, for example, by CAN (Controller Area Network) communication. Note that communication methods other than CAN communication may be used, and wireless communication may be used instead of wired communication.

[0013] The information notification control device 1 performs control related to the notification of information. The information notification control device 1 determines in what notification mode to notify the information to be notified to the notification target (a detailed example is a driver) received from the information source device 2. Further, the information notification control device 1 controls the information notification device 3 so that the notification is performed in the determined notification mode. The determination of the notification mode includes, for example, the determination of notification means, notification location, and notification parameters such as notification intensity. As described above, the information notification control device 1 performs notification control on the information notification device 3 by acquiring information from the information source device 2. In the present embodiment, the information notification control device 1 uses an AI model at least for the determination of notification parameters. Details of the information notification control device 1 configured as such will be described later.

[0014] Note that in the present embodiment, the information notification control device 1 is an in-vehicle device mounted on the vehicle VE, but this is an example. The information notification control device 1 may be composed of an in-vehicle device and a server device provided communicably via a communication network such as the Internet with the in-vehicle device. Further, the information notification control device 1 mounted on the vehicle VE may be configured as a part of in-vehicle devices such as a navigation device or a display audio.

[0015] The information source device 2 is a device that serves as an information source of the information to be notified to the notification target, and provides notification information to the information notification control device 1. Specifically, the information source device 2 is a device that serves as an information source of the information to be notified to the driver. The information source device 2 outputs the information to be notified to the driver toward the information notification control device 1. The information source device 2 may be configured to, for example, read out information stored in a storage medium (not shown) and output it toward the information notification control device 1. Further, the information source device 2 may be configured to generate the information to be notified by itself and output it toward the information notification control device 1. Further, the information source device 2 may be configured to receive the information to be notified from another device such as a server device and output it toward the information notification control device 1.

[0016] In this embodiment, the information (notification information) output from the information source device 2 is information for assisting driving. That is, the information source device 2 is a device that serves as an information source for information for assisting driving. The information for assisting driving may widely include information related to driving, such as information indicating the state of the vehicle VE, information indicating the state of the driver, information notifying about the surrounding environment of the vehicle VE, information related to route guidance, and attention - calling information during driving. The information source device 2 that outputs such notification information may be constituted by, for example, a navigation device, a safety monitoring device, or the like.

[0017] Note that the information source device 2 may be constituted by only one device or may be constituted by a plurality of devices. Also, in this embodiment, the information source device 2 is an in - vehicle device, but at least some of its functions may be constituted by a server device that is communicable with the in - vehicle device. Further, when the notification information may also include information other than information related to driving, the notification information may include entertainment information, Internet information such as news, and the like.

[0018] The information notification device 3 performs information notification according to a command from the information notification control device 1. Specifically, the information notification device 3 performs information notification by using at least one of the driver's vision, hearing, and touch according to a command from the information notification control device 1. In this embodiment, the information notification device 3 performs notification of information for assisting driving. The information notification device 3 is constituted by a plurality of types of devices. The plurality of types of devices are a display device 31, an audio output device 32, and a vibration generating device 33.

[0019] The display device 31 is a means of notifying information using the driver's vision. The display device 31 is positioned in a location where the driver of the vehicle VE to which the driver assistance system 100 is applied can easily see the displayed content. For example, the display device 31 is a liquid crystal display or an organic EL display placed on the dashboard of the vehicle VE. Alternatively, the display device 31 is a head-up display (HUD) that projects information onto the windshield of the vehicle VE. Alternatively, the display device 31 is an indicator light placed on the meter panel, rearview mirror, or side mirror. The display device 31 may also be part of a stationary or portable navigation system or safety monitoring device mounted in the vehicle. The display device 31 may also be a display operation device with an operation function such as a touch panel. There may be one or more display devices 31. If there are multiple, there may be multiple types of display devices 31.

[0020] The voice output device 32 is a means of notifying information using the driver's hearing. The voice output device 32 is, for example, a speaker positioned in a location where the driver of a vehicle VE to which the driver assistance system 100 is applied can easily hear the voice. The voice output device 32 may also be part of a stationary or portable navigation system or safety monitoring system installed in the vehicle. There may be one or more voice output devices 32.

[0021] The vibration generator 33 is a means of notifying the driver of information using their sense of touch. The vibration generator 33 includes, for example, a vibrator positioned in a location where the driver of the vehicle can feel the vibration. The vibrator is positioned, for example, in the driver's seat. More specifically, the vibrator is positioned in the seat that supports the driver's buttocks, or in the backrest that supports the driver's back when their buttocks are on the seat. The vibrator may be, for example, a vibrator with an electrical-magnetic circuit configuration in which the diaphragm of a speaker (having a structure suitable for acoustic conversion) is replaced with a diaphragm suitable for vibration transmission, or a vibrator with a configuration utilizing a piezoelectric element.

[0022] In this embodiment, the system utilizes three senses—sight, hearing, and touch—when notifying information, but this is merely an example. The types of senses that can be used when notifying information may be one or two, for example. For instance, the types of senses that can be used when notifying information may be two, such as sight and hearing. Furthermore, the information notification device 3 may consist of one type of device or multiple devices other than three.

[0023] <2. Configuration of Information Notification Control Device> Next, we will describe in detail the configuration of the information notification control device 1, which controls the notification of information. As shown in Figure 1, the information notification control device 1 comprises a controller 11 and a memory 12. The information notification control device 1 is a so-called computer device and, in addition to the controller 11 and memory 12, also comprises an input / output unit (not shown).

[0024] The controller 11 is configured to include an arithmetic circuit that performs calculations. The arithmetic circuit is more specifically composed of a processor. The processor is composed of, for example, a CPU (Central Processing Unit). The controller 11 may consist of one processor or multiple processors. If it consists of multiple processors, those processors should be provided to communicate with each other.

[0025] Memory 12 consists of volatile memory and non-volatile memory. The volatile memory is specifically RAM (Random Access Memory). The non-volatile memory is specifically ROM (Read Only Memory). The non-volatile memory may also be flash memory or a hard disk drive, etc. The non-volatile memory stores a program 121 and data that can be read by the computer.

[0026] The program 121 stored in memory 12 may be provided, for example, on a computer-readable non-volatile recording medium. The non-volatile recording medium may be, for example, an optical recording medium (e.g., an optical disc), a magneto-optical recording medium (e.g., a magneto-optical disc), a USB memory, or an SD card, in addition to the non-volatile memory described above. As another example, the program 121 may be provided from a program provision server via a communication network such as the Internet (a configuration provided by so-called download).

[0027] The functions of the controller 11 are realized by the processor executing arithmetic processing according to the program 121 stored in memory 12. The number of programs 121 that realize the functions of the controller 11 may be one or more. The functions of the controller 11 may be realized by the execution of arithmetic processing according to the program 121 by an arithmetic circuit, i.e., by software, but may also be realized by other methods. At least some of the functions of the controller 11 may be realized using, for example, an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array). In other words, at least some of the functions of the controller 11 may be realized by hardware using a dedicated IC or the like. Furthermore, at least some of the functions of the controller 11 may be realized by using a combination of software and hardware.

[0028] Figure 2 is a block diagram showing the configuration of the functional units of the controller 11 of the information notification control device 1. As shown in Figure 2, the controller 11 comprises an information acquisition unit 111, a pre-processing unit 112, a group value acquisition unit 113, an HMI (Human Machine Interface) control model execution unit 114, and an output control unit 115 as its functional units. Note that each functional unit 111 to 115 is a conceptual component. The function performed by one component may be distributed among multiple components. Alternatively, the functions of multiple components may be integrated into a single component.

[0029] The information acquisition unit 111 acquires information from the memory 12 and from devices (in-vehicle devices or external devices) or sensors that are configured to communicate with the information notification control device 1. If the information notification control device 1 is configured to communicate with external devices, it includes a communication unit with an interface circuit for connecting to a communication network (not shown), such as the Internet. The information acquired by the information acquisition unit 111 includes notification information output from the information source device 2. Furthermore, the information acquired by the information acquisition unit 111 includes various types of information for input into the HMI control model, which will be described in detail later.

[0030] The preprocessing unit 112 performs various preprocessing steps necessary for providing information notification when it becomes necessary to notify the driver, such as by obtaining notification information from the information source device 2. These preprocessing steps include, for example, determining the notification priority and deciding on the notification format.

[0031] The priority of a notification is determined based on the urgency of the information to be notified. For example, if the information to be notified is warning of danger, the urgency will be high and the priority will be determined to be high. For example, if the information to be notified is directions such as turning left or right, the urgency will not necessarily be high and the priority will be determined to be low. The priority of a notification may be determined by creating a table in advance that associates the notification content (notification information) with the priority, storing it in memory 12, and using that table. Alternatively, as another example, the priority of a notification may be obtained using an AI model that takes the notification content and various information such as the occupant (driver) profile as input and outputs the priority. Such an AI model may be generated by supervised learning performed on a large amount of data with correct labels that associate the notification content and occupant profile with the correct value of the priority.

[0032] Determining the notification format includes determining the means and location of information notification. This determination process determines, for example, which of text, graphics, sound, and vibration will be used as the means of information notification. It also determines where the display, sound generation, or vibration generation will occur. Note that if there is only one notification means for each category such as display, the notification location is automatically determined by the determination of the notification means, so in such a configuration, there is no need to perform a notification location determination process. The means of information notification, etc., may be determined by a pre-prepared rule base. This rule base may use, for example, notification priority information or occupant (driver) profiles. The occupant profile is, for example, stored in memory 12 (see Figure 1) in advance. For example, in the case of a driver with poor hearing, it is decided that voice will not be used as a notification means, and other means (display or vibration) will be used instead. Also, if the notification priority is high, it is decided that multiple notification means will be used.

[0033] The group value acquisition unit 113 acquires the value of a group variable assigned to each group, which is determined according to the content of the notification information acquired by the information acquisition unit 111. In other words, the information notification control device 1 acquires the value of a group variable assigned to each group, which is determined according to the content of the notification (notification content), by acquiring information from the information source device 2.

[0034] In this embodiment, a group is formed by gathering together notification content that is presumed to be similar from a specific viewpoint. The information notification control device 1 has result information of the grouping of notification content that is similar from a specific viewpoint into the same group, and utilizes this result information. In other words, in this embodiment, the information notification control device 1 does not perform the grouping of notification content itself, but rather uses the results of grouping that has already been performed to obtain group values.

[0035] The state of similarity from the specific viewpoint described above refers to a state in which the same type of input variable is included in the input variables that primarily influence the estimation result of the notification parameters used for notification control of the information notification device 3 (the estimation result using the AI ​​model). From this viewpoint, each group (each group) is comprised of notification contents that contain the same type of input variable in the input variables that primarily influence the estimation result of the notification parameters. With this configuration, notification contents with similar output tendencies to the input of the AI ​​model can be grouped together. For this reason, even if the number of possible values ​​that the input variables related to the notification contents input to the AI ​​model can be taken into account is reduced by grouping, it is possible to suppress a decrease in the estimation (inference) performance of the AI ​​model. Note that a group usually contains multiple types of notification contents (notification information), but in the grouping process, it may not be possible to find any items that belong to the same group, so a group may contain only one type of notification content.

[0036] The pre-classified groups include, for example, a group whose input variable type primarily affects the notification parameter is collision time to collision (TTC; Time to Collision) (TTC-dependent group). The TTC-dependent group includes, for example, notifications related to "pause," "collision," and "merging." The pre-classified groups also include, for example, a group whose input variable type primarily affects the notification parameter is driver state information (driver state-dependent group). The driver state-dependent group includes, for example, notifications related to "driver drowsiness" and "distraction." These classification examples are merely illustrative, and different classifications may be used.

[0037] The number of "types of input variables that primarily have an influence" mentioned above may be singular or multiple. Furthermore, each group may contain a mix of notification content that is estimated to be similar from perspectives other than the specific viewpoint mentioned above. Detailed examples of grouping based on notification content will be discussed later.

[0038] The group variable value (group value) is a value assigned to each group for the purpose of identifying each group included in multiple groups. Group variable information, including information on the group value, is stored in memory 12. In detail, the group value acquisition unit 113 (i.e., the information notification control device 1) acquires the group variable value (group value) using conversion information that associates the relationship between the notification content and the group variable value (group value). The conversion information is, for example, a table or a conversion formula. By using this conversion information, the process of acquiring the group variable value can be easily performed.

[0039] Figure 3 shows an example of the group variable conversion table 200. The group variable conversion table 200 is an example of the conversion information described above. The group variable conversion table 200 is a table used to obtain the group variable value (group value) from the notification content of the acquired notification information. As shown in Figure 3, the information items in the group variable conversion table 200 include the notification content, group name, and group value.

[0040] The "Notification Content" field stores the content of the notification information obtained from the information source device 2. The group variable conversion table 200 generates a data record for each notification content. The data record stores the "group name" and "group value," which are the information corresponding to the notification content.

[0041] The "Group Name" field stores the group name pre-assigned to the notification content. The "Group Value" field stores the value of the group variable pre-assigned to the notification content. In this embodiment, since the purpose of using the group variable conversion table 200 is to obtain the value of the group variable (group value) from the notification content, the "Group Name" field may be omitted.

[0042] In the example shown in Figure 3, if the notification content is "Caution at merging" or "Caution regarding collisions," the group value acquisition unit 113 acquires "1" as the group value. If the notification content is "Caution regarding falling asleep," the group value acquisition unit 113 acquires "2" as the group value.

[0043] The HMI control model execution unit 114 (see Figure 2) reads the model information 122 (see Figure 1) stored in the memory 12 and executes processing by the HMI control model. The model information 122 is the model information of the HMI control model, and in detail includes the structure and parameters of the HMI control model, as well as code instructions for executing processing by the HMI control model.

[0044] Figure 4 is a schematic diagram illustrating the HMI control model 4. The HMI control model 4 is a trained AI model that performs inference processing in response to information input and outputs the inference results. Specifically, the HMI control model 4 receives priority information and notification format obtained by the preprocessing unit 112, as well as various other information, and outputs notification parameters for providing information notifications that are estimated to be comfortable for the driver, as a result of inference on the input. The HMI control model 4, which has such functions, may consist of a single AI model or multiple AI models. In this embodiment, as will be described later, the HMI control model 4 is composed of multiple AI models. Because it is composed of multiple AI models, the model information 122 stored in the memory 12 contains model information for each of the multiple AI models.

[0045] Other types of information input to the HMI control model 4 may include, for example, group information, scene information, DMS (Driver Monitoring System) information, occupant profiles, and environmental information. This information is obtained, for example, from devices (in-vehicle or external devices) or sensors that are communicatively connected to the information notification control device 1, or from the memory 12 provided by the information notification control device 1.

[0046] Group information is information about a group determined according to the notification content, and in detail, it is the value of the group variable described above. Scene information is information such as ADAS (Advanced driver-assistance systems) operation scenes and surrounding vehicle approach scenes. ADAS operation scenes include automatic braking scenes. DMS information is driver monitoring information (status information) from the in-vehicle camera, such as driver drowsiness, concentration level, or posture. Occupant profile is occupant information such as the driver, such as age, gender, or driving history. Environmental information is environmental information inside and outside the vehicle VE, such as brightness information inside and outside the vehicle, noise information inside the vehicle, vibration information inside the vehicle, or weather information.

[0047] The notification parameters output by the HMI control model 4 are, in detail, notification parameters corresponding to the notification format determined by the preprocessor 112. For example, if it is determined that text, graphics, voice, and vibration will be used as notification formats, parameters corresponding to these will be output. For example, for text and graphics, parameter information such as size, placement, and color scheme will be output. For voice, parameter information such as volume and sound quality will be output. For vibration, parameter information such as vibration intensity and vibration frequency will be output.

[0048] Figure 5 shows specific examples of inputs and outputs in HMI control model 4. The "input values" shown in Figure 5 are specific examples of values ​​input to HMI control model 4. In the example shown in Figure 5, the input values ​​input to HMI control model 4 are "1" for the input item "Group", "5" for the input item "Priority", "1" for the input item "Gender", "3" for the input item "Age", and "3" for the input item "In-vehicle vibration". Note that the input item corresponds to the type of input variable.

[0049] Note that the input items shown in Figure 5 represent only a portion of the total input items, and input values ​​are provided similarly for input items other than those shown in Figure 5. Furthermore, the form of the values ​​for each input item (priority, gender, age, in-car vibration, etc.) is defined using a value conversion table or the like to ensure that the values ​​are in a form suitable for handling in the AI ​​model. For example, for gender, male is defined as "1", female as "2", and others as "3". Similarly, for age, for example, under 20 years old is defined as "1", 21 to 25 years old as "2", 26 to 35 years old as "3", 36 to 59 years old as "4", 60 to 69 years old as "5", 70 to 79 years old as "6", and 80 years old and over as "7". The input value entered in the input item "Group" is the group value obtained by the group value acquisition unit 113 described above.

[0050] The "output values" shown in Figure 5 are specific examples of values ​​output from the HMI control model 4. The output values ​​shown in Figure 5 are obtained as a result of inputting all the input values ​​for each input item shown in Figure 5 into the HMI control model 4. In the example shown in Figure 5, it is assumed that the vibration generator 33 is used as a notification means, and the parameter values ​​of "vibration intensity" and "vibration frequency" are output as notification parameters. As with the input values ​​mentioned above, the format of the output parameter values ​​is defined to be a numerical format suitable for handling in the AI ​​model.

[0051] The output control unit 115 (see Figure 2) controls the information notification device 3 based on a notification signal indicating the content of the notification information (notification content) and notification parameters output by the HMI control model 4. In detail, the output control unit 115 controls the display device 31, the audio output device 32, and the vibration generator 33 (all see Figure 1) according to the notification parameters output by the HMI control model 4. As a result of this control, the driver receives information notification from at least one of the display device 31, the audio output device 32, and the vibration generator 33.

[0052] <3. Operation of the Information Notification Control Device> Next, the operation of the information notification control device 1 will be described.

[0053] Figure 6 is a flowchart illustrating the flow of notification control processing executed by the information notification control device 1. This flowchart shows the technical content of a computer program (information notification control program) that enables a computer to implement the information notification control method of this embodiment. This computer program can be stored and provided (sold, distributed, etc.) on various non-volatile recording media readable by a computer. This computer program can also be provided from a program provision server via a communication network such as the Internet. This computer program may consist of only one program, or it may consist of multiple programs working together.

[0054] The process shown in Figure 6 is executed as appropriate when the vehicle VE is powered on and the information notification control device 1 becomes capable of executing notification control processing.

[0055] In step S1, the controller 11 (information acquisition unit 111) acquires notification information from the information source device 2. As described above, the notification information is information related to the operation of the vehicle VE, such as information warning of danger or information urging caution. Once the notification information is acquired, the process proceeds to the next step S2.

[0056] In step S2, the controller 11 (preprocessing unit 112) performs various preprocessing steps necessary for notification control using the HMI control model 4. Specifically, as described above, this includes determining the notification priority and deciding on the notification format. Once the preprocessing is complete, the process proceeds to the next step S3.

[0057] In step S3, the controller 11 (group value acquisition unit 113) acquires the group value based on the content of the notification information (notification content) acquired in step S1 and the group variable conversion table 200 (see Figure 3). Once the group value is acquired, the process proceeds to step S4. Note that the order of the processes in step S2 and step S3 may be reversed.

[0058] In step S4, the controller 11 (HMI control model execution unit 114) executes processing by the HMI control model 4. That is, the controller 11 estimates the notification parameters for controlling notifications on the information notification device 3. In estimating the notification parameters, the group value and the values ​​of other input variables are input to the HMI control model 4. Information regarding the other input variables is acquired as appropriate by the information acquisition unit 111. The process of acquiring this information regarding the other input variables may be included in the above preprocessing (step S2).

[0059] Other input variables include at least one of the following: variables relating to the recipient of the notification (in a detailed example, the driver), variables relating to the surrounding information of the recipient of the notification, and variables relating to the priority of the notification. By inputting the values ​​of these variables into the HMI control model 4, it is possible to estimate notification parameters for providing notifications that the driver (user) finds comfortable. In this embodiment, other input variables include variables relating to the driver, variables relating to the driver's surrounding information, and variables relating to the priority of the notification.

[0060] Other specific examples of input variables include those other than "Group" shown in Figure 5. To explain in more detail, specific examples of variables related to the recipient of the notification include variables obtained from the DMS information and occupant profiles mentioned above, and more specifically, gaze variables, clothing variables, emotional variables, visual acuity variables, hearing variables, age variables, gender variables, driving history variables, etc. Specific examples of variables related to the surrounding information of the recipient of the notification include variables obtained from the scene information and environmental information mentioned above, and more specifically, TTC variables, external visibility variables, interior brightness variables, interior noise variables, vehicle vibration (interior vibration) variables, etc. Variables related to the priority of the notification include variables obtained from the priority information obtained in the preprocessing unit 112 (priority variables).

[0061] Figure 7 schematically illustrates the processing flow in HMI control model 4. In Figure 7, "input information" is a collective term for the group values ​​and other input variables input to HMI control model 4. "Output information" refers to the output values ​​(notification parameters) output from HMI control model 4. In Figure 7, n is a natural number greater than or equal to 3, but this is an example, and in other examples, n may be "2".

[0062] As shown in Figure 7, in this embodiment, the HMI control model 4 is composed of multiple AI models 40. More specifically, the multiple AI models 40 are composed of group-specific AI models (trained models) separately provided for each group determined according to the notification content. That is, the HMI control model 4 includes multiple group-specific AI models 40 separately provided for each group determined according to the notification content. Each group-specific AI model 40 is trained according to the characteristics of each group and outputs notification parameters according to the characteristics of the group upon input of input information. The training of the group-specific AI models 40 will be described later.

[0063] The input information provided to the HMI control model 4 includes a group value. Depending on the input group value, it is determined which group-specific AI model 40 to use. This determination process may be included in the processing of the HMI control model 4, or it may be performed before the input information is provided to the HMI control model 4, and not included in the processing of the HMI control model 4. In the example shown in Figure 7, the group value is "1", and the first AI model M1 corresponding to the group value "1" is used to estimate the notification parameter. Specifically, the input information is input to the first AI model M1 and not to the other AI models M2...Mn. For this reason, the output information of the HMI control model 4 is the notification parameter estimated by the first AI model M1.

[0064] In the example shown in Figure 7, if the group value included in the input information is "2", the second AI model M2 is used, and the output value output from the second AI model M2 becomes the notification parameter estimated by the HMI control model 4. If the group value included in the input information is "n", the nth AI model Mn is used, and the output value output from the nth AI model Mn becomes the notification parameter estimated by the HMI control model 4.

[0065] Furthermore, in the example shown in Figure 7, regardless of which of the multiple group-specific AI models 40 is selected, the input items (input variables) of the input information input to the selected group-specific AI model 40 remain the same. However, this is merely an example, and the configuration may be such that the input items of the input information input to the AI ​​model change when the selected group-specific AI model 40 (in other words, the group) changes.

[0066] Returning to Figure 6, once the process in step S4 is completed, the process proceeds to step S5.

[0067] In step S5, the controller 11 (output control unit 115) controls the information notification device 3 based on the notification parameters output from the HMI control model 4. In the example shown in Figure 7, notification parameters (vibration intensity and vibration frequency) for controlling the vibration generator 33 are output, and the vibration generator 33 is controlled accordingly. As a result, the driver receives information notification using the vibration generator 33 (tactile). Note that in the example shown in Figure 7, the HMI control model 4 is configured to output only tactile notification parameters, but this is just an example. Depending on the information input to the HMI control model 4, notification parameters for the display device 31 (visual) or the audio output device 32 (auditory) may also be output. In addition, depending on the information input to the HMI control model 4, at least two of the visual, auditory, and tactile notification parameters may be output.

[0068] As explained above, the information notification control device 1 estimates notification parameters by executing the HMI control model 4 based on the values ​​of the group variable and the values ​​of other input variables, and controls notifications based on the notification parameters obtained through estimation. In this configuration, the values ​​of the input variables related to the notification content that are input to the HMI control model 4 for estimating notification parameters are assigned not to each notification content, but to groups determined according to the notification content. This reduces the number of possible values ​​for the input variables related to the notification content that are input to the HMI control model 4. As a result, the capacity of the AI ​​model used for information notification control can be reduced.

[0069] Furthermore, as described above, the grouping based on notification content is performed so that notifications with similar relationships between the model's input and output are grouped together. In this configuration, the possibility of the HMI control model 4 making inappropriate estimations of notification parameters due to the influence of grouping can be kept to a minimum. As a result, it is possible to reduce the capacity of the AI ​​model while suppressing a decrease in the notification parameter estimation ability.

[0070] Furthermore, in this embodiment, notification parameters are estimated using a group-specific AI model 40 selected from multiple group-specific AI models 40 according to the value of the group variable. With this configuration, compared to the case where notification parameters are estimated for all notification content using a single AI model, it is easier to train the AI ​​model to produce the desired output, and the likelihood of appropriately estimating the notification parameters is increased. On the other hand, since the total number of AI models can be reduced compared to the case where an AI model is created individually for each notification content, the total capacity of the AI ​​model can be reduced. In other words, in this embodiment, it is possible to achieve both ensuring the performance of notification parameter estimation by the AI ​​model (HMI control model 40) and reducing the capacity of the AI ​​model.

[0071] In the above example, a configuration with an AI model for each group was used, but this is merely an example. As shown in Figure 8, the HMI control model 4A may be composed of a single AI model instead of multiple AI models based on grouping. Figure 8 is a diagram illustrating an modified version of the HMI control model 4A. In Figure 8, the input and output information are the same as in Figure 7.

[0072] In the modified example shown in Figure 8, the HMI control model 4 is a shared AI model used by all groups determined according to the notification content. As shown in Figure 8, in this modified example, as in the embodiment described above, the input information includes group values. The configuration of the modified example is very effective when the notification parameters can be appropriately estimated using a single AI model instead of providing an AI model for each group. In this configuration, the HMI control model 4A is composed of a single AI model rather than multiple AI models, thus reducing the capacity of the AI ​​model. Furthermore, since the input variables related to the notification content input to the HMI control model 4A are assigned by group rather than by notification content, the number of possible values ​​for the variables can be reduced, thereby reducing the capacity of the AI ​​model. Note that in the configuration shown in Figure 7, the group values ​​are basically not input into the AI ​​model but are used for selecting the AI ​​model. In contrast, in the configuration shown in Figure 8, the group values ​​are input into the AI ​​model and used for inference processing in the AI ​​model.

[0073] <4. Generating the HMI control model> Next, we will explain matters related to the generation of the HMI control model 4 used in the information notification control device 1.

[0074] [4-1. AI Model Generation Device] Figure 9 is a block diagram illustrating the schematic configuration of an AI model generation device 5 according to an embodiment of the present invention. The AI ​​model generation device 5 is a device that generates an AI model for estimating notification parameters used for controlling information notification, and more specifically, a device that generates an HMI control model 4. The AI ​​model generation device 5 may be an in-vehicle device mounted on a vehicle VE, but in this embodiment, it is an external device. The AI ​​model generation device 5 configured as an external device may be, for example, a server device such as a cloud server. The AI ​​model generation device 5 may also be composed of multiple devices that are provided to communicate with each other.

[0075] As shown in Figure 9, the AI ​​model generation device 5 comprises a controller 51 and a memory 52. ​​The AI ​​model generation device 5 is a so-called computer device.

[0076] The controller 51 is configured to include an arithmetic circuit that performs calculations. The arithmetic circuit is more specifically composed of a processor. The processor is composed of, for example, a CPU. The controller 51 may be composed of one processor or multiple processors. If it is composed of multiple processors, those processors should be provided to be able to communicate with each other.

[0077] Memory 52 is composed of volatile memory (such as RAM) and non-volatile memory (such as ROM). The non-volatile memory stores a program 521 and data that can be read by a computer. The program 521 stored in memory 52 may be provided, for example, by a computer-readable non-volatile recording medium. The non-volatile recording medium may be, for example, an optical recording medium, a magneto-optical recording medium, a USB memory, or an SD card, in addition to the non-volatile memory described above. As another example, the program 521, etc., may be provided from a program provision server via a communication network such as the Internet (a configuration provided by so-called download).

[0078] The functions of the controller 51 are realized by the processor executing arithmetic processing according to a program stored in memory 52. ​​The number of programs that realize the functions of the controller 51 may be one or more. The functions of the controller 51 may be realized by software, but may also be realized by other methods. At least some of the functions of the controller 51 may be realized by hardware, for example, by using an ASIC or FPGA. Furthermore, at least some of the functions of the controller 51 may be realized by using a combination of software and hardware.

[0079] As shown in Figure 9, the controller 51 comprises a classification unit 511, a learning unit 512, and an aggregation unit 513 as its functional units. Note that each of the functional units 511 to 513 is a conceptual component. The function performed by one component may be distributed among multiple components. Alternatively, the functions of multiple components may be integrated into a single component.

[0080] The classification unit 511 groups multiple types of notification information, which have been prepared in advance for information notification using the information notification device 3, based on the content of the notification. In this embodiment, the information notification device 3 is included in the driver assistance system 100, and the content of the notification information (notification content) is related to driver assistance. Details of the group classification process performed by the classification unit 511 will be described later.

[0081] The learning unit 512 performs two types of learning: learning to obtain a provisional model, which is an AI model prepared experimentally for use in processing by the classification unit 511, and learning to obtain the final model, which is an AI model to be used as a product. Both of these learning processes are performed using a learning dataset that collects a large amount of data with correct labels. In this embodiment, the HMI control model 4 used in the information notification control device 1 corresponds to the AI ​​model to be used as a product.

[0082] Figure 10 shows an example of data with correct labels. As shown in Figure 10, data with correct labels has correct labels used to compare the input values ​​input to the AI ​​model with the output values ​​output from the AI ​​model. The input values ​​of the AI ​​model include input values ​​for multiple items as described above. The correct labels are generated, for example, based on the feedback of collaborators who help in obtaining the correct labels. In detail, the correct labels are generated based on the feedback of collaborators obtained by having them experience notifications from the information notification device 3 when the notification parameters are varied for each of the various input conditions obtained by changing the input values. The notification parameters that the collaborator found comfortable become the correct labels. By using a large number of collaborators, a large amount of data with correct labels can be obtained.

[0083] Learning using data with correct labels is, in detail, supervised learning. Supervised learning can be performed using known methods, and during learning, the model parameters are repeatedly adjusted so that the error between the model's output value and the correct label is minimized. Upon completion of supervised learning, a trained AI model is obtained. The HMI control model 4 used in the information notification control device 1 is a trained AI model that has undergone such learning.

[0084] As described above, in this embodiment, the HMI control model 4 is composed of multiple group-specific AI models 40 (see Figure 7). Therefore, in detail, training the above model (HMI control model 4) means training each of the multiple group-specific AI models.

[0085] The aggregation unit 513 aggregates multiple group-specific AI models 40, which have been trained separately, into one, forming an HMI control model 4 composed of multiple group-specific AI models 40. The aggregation referred to here may simply be a formal aggregation for the purpose of combining multiple group-specific AI models 40 into a single package of information. Alternatively, if the HMI control model 4 has a decision processing function that determines which group-specific AI model 40 to use according to the group value, the aggregation referred to here may be performed by adding such a function to combine multiple group-specific AI models 40 into one.

[0086] [4-2. AI Model Generation Method] Next, we will explain how to generate an AI model that estimates notification parameters used to control information notifications. This AI model generation method is performed using the AI ​​model generation device 5 described above.

[0087] Figure 11 is a flowchart illustrating the flow of the AI ​​model generation method of this embodiment. This flowchart shows the technical content of a computer program (AI model generation program) that enables a computer to implement the AI ​​model generation method of this embodiment. This computer program can be stored on various non-volatile recording media readable by a computer and provided (sold, distributed, etc.). This computer program may consist of only one program, or it may consist of multiple programs working together. Furthermore, the process shown in Figure 11 is, in detail, a method for generating the HMI control model 4, and is executed as appropriate at the timing when the generation of the HMI control model 4 is required.

[0088] In step S11, the controller 51 (classification unit 511) groups the multiple types of notification information that have been prepared in advance for information notification using the information notification device 3, based on the content of the notification. In other words, in the method for generating the AI ​​model, grouping is performed based on the content of the information notification. In this grouping, notification contents that are estimated to have a similar relationship between the input to the AI ​​model and its output are grouped together. By performing grouping in this way, it is possible to suppress a decrease in the accuracy (performance) of the AI ​​model (specifically the HMI control model 4) due to the influence of grouping. A detailed example of group classification (grouping) will be explained with reference to Figure 12.

[0089] Figure 12 is a flowchart illustrating the group classification process. Note that Figure 12 is a diagram illustrating the detailed processing of step S11 in Figure 11.

[0090] In step S111, the controller 51 (classification unit 511) calculates the contribution of each of the pre-prepared types of notification information (notification content). Qualitatively, the contribution is a parameter that represents the degree of influence each input variable has on the estimation results of the notification parameters using the AI ​​model; a larger value indicates a greater degree of influence. The calculation of the contribution is performed using the hypothetical model described above. The contribution is determined for each notification content, and more specifically, for each input variable in each notification content.

[0091] More specifically, the contribution of each input variable to each notification is determined using a hypothetical model prepared individually for each notification (notification content). The contribution of each input variable is given by a linear regression coefficient derived from the relationship between multiple input and output values ​​obtained by changing the input values ​​multiple times and obtaining multiple output values ​​from the hypothetical model. Examples of input variables include the TTC variable, in-vehicle noise variable, gaze variable, visual acuity variable, hearing variable, age variable, etc.

[0092] The provisional models prepared for each notification are pre-trained AI models that have been trained using training datasets collected for each notification. For example, if the notification is "Caution: Merging," the provisional model is trained using a training dataset consisting of a large amount of data with correct labels collected under circumstances where "Caution: Merging" is issued. Similarly, if the notification is "Caution: Entanglement," the provisional model is trained using a training dataset consisting of a large amount of data with correct labels collected under circumstances where "Caution: Entanglement" is issued. The provisional models are merely designed to understand the trends regarding the input to the model and its output, and therefore can be simple models that are sufficient to achieve this objective. For this reason, the amount of data prepared for training the provisional models can be less than the amount of data prepared for generating the main models described above.

[0093] Once the contribution of each input variable is calculated for each of the pre-prepared types of notification content, the process proceeds to the next step, S112.

[0094] In step S112, the controller 51 (classification unit 511) classifies the notification content based on the contribution of each input variable for each acquired notification content. Classification is performed, for example, using the following equation (1). ||X1|-|Y1||+||X2|-|Y2||+····≦ α (1) X: Contribution of input variable A in notification content X1, X2, etc., indicate that they are different types of input variables. Y: Contribution of input variable B in notification content Y1, Y2, etc., indicate that the input variables are of different types. α: Pre-set classification threshold

[0095] In equation (1), if the numbers following "X" and "Y" are the same, it means that they represent the contribution of the same input variable. For example, X1 is the contribution of the TTC variable in notification content A, and Y1 is the contribution of the TTC variable in notification content B. Also, for example, X2 is the contribution of the gaze variable in notification content A, and Y2 is the contribution of the gaze variable in notification content B.

[0096] In detail, first, two notification contents are extracted from several types of notification contents that have been prepared in advance. Of the two extracted notification contents, one is assigned to notification contents A and the other to notification contents B. According to this assignment, the contribution of each input variable for each notification contents, which was determined earlier, is applied to the left side of equation (1) to find the sum of the left side of equation (1). If the calculated sum is less than or equal to the classification threshold α on the right side of equation (1), notification contents A and notification contents B are considered to be closely related, and notification contents B is placed in the same group as notification contents A. On the other hand, if the calculated sum is greater than the classification threshold α, notification contents A and notification contents B are not considered to be closely related, and notification contents B is placed in a different group from notification contents A.

[0097] After this, the notification content assigned to notification content A in the previous process remains assigned to notification content A, and for the other notification content that has not been processed using formula (1), a determination process is performed in order to determine whether or not it belongs to the same group as notification content A, using formula (1). If, after performing the above determination process for all notification content, there is one or fewer notification content that does not belong to the same group as notification content A, the classification process in step S112 is terminated.

[0098] On the other hand, if there are multiple notifications that remain ungrouped, it is determined whether there are any notifications among the remaining notifications that have not yet been designated as Notification A. If there are no notifications that have not yet been designated as Notification A, the classification process in step S112 is terminated. If there are notifications among the ungrouped notifications that have not yet been designated as Notification A, one such notification is selected and designated as Notification A. Then, Notification B, which will be compared with Notification A, is selected one by one from the remaining ungrouped notifications, and in each case, a determination process is performed using formula (1) to determine whether or not it belongs to the same group as Notification A. The above process is repeated until there is one or fewer notifications remaining that have not been grouped, or when there are no more notifications among the remaining ungrouped notifications that have not yet been designated as Notification A, the classification process in step S112 is terminated. Note that some groups formed in this way may contain only one notification. Once the process in step S112 is completed, the process proceeds to the next step, S113.

[0099] The classification method using equation (1) described above is merely an example, and other classification methods may be used. For example, instead of using equation (1), one could use a configuration in which the determination of whether or not to group is based on whether all of the difference parts on the left side of equation (1) (||X1|-|Y1||, ||X2|-|Y2||, etc.) are less than or equal to the classification threshold β.

[0100] In step S113, the controller 51 (classification unit 511) determines whether the number of groups obtained in the classification in step S112 is less than or equal to a preset value T. The preset value T is determined, for example, according to the capacity of the memory 12 provided in the information notification control device 1 that executes the HMI control model 4. If the number of groups is less than or equal to the preset value T (Yes in step S113), the group classification shown in Figure 12 is completed, and the process proceeds to step S12 in Figure 11. If the number of groups is greater than the preset value T (No in step S113), the process proceeds to step S114.

[0101] In step S114, the controller 51 (classification unit 511) increases the classification threshold (α in equation (1)) used in the classification based on contribution in step S112 and performs reclassification processing. The extent to which the classification threshold is increased can be determined as appropriate, but there is a limit to how much it can be increased, as increasing it too much may degrade the performance of the final AI model (HMI control model 4). The reclassification processing may be the same as the processing in step S112. However, it may also be configured to determine whether notification content included in groups with a small number of constituent notification content can be incorporated into an already existing group. Once the reclassification processing is complete, the process proceeds to the next step S115.

[0102] In step S115, the controller 51 (classification unit 511) determines whether the number of groups obtained in the reclassification in step S114 is less than or equal to a preset value T. The preset value T is the same as the preset value T in step S113. If the number of groups is less than or equal to the preset value T (Yes in step S115), the group classification shown in Figure 12 is completed, and the process proceeds to step S12 in Figure 11. If the number of groups is greater than the preset value T (No in step S115), the process proceeds to step S116.

[0103] In step S116, the controller 51 (classification unit 511) implements other classification methods to further reduce the number of groups. In the processing of step S116, for example, it is determined whether a notification belonging to a group with only one notification can be incorporated into another group using a different method than in step S114. For example, even if incorporating it into a group may lower the accuracy rate of the final AI model (HMI control model 4), it is determined that the impact of the decrease in accuracy rate is small for notification content that is estimated to be able to be incorporated into a group. If such a determination is made, the process of incorporating the notification content into one of the groups is performed. Examples of notifications that are less affected by the decrease in accuracy rate include notifications with low priority and frequency given in advance at the design stage. Another example is that the designer's know-how may be used to determine whether it is possible to incorporate it into a group. In this case, the incorporation decision itself may be made by the AI ​​model generation device 5 asking the designer, and the designer making the decision. Another example is that if it is determined that notifications have the same or similar sensors or detection methods related to the notification, they may be incorporated into a group. Once step S116 is completed, the group classification shown in Figure 12 is completed, and the process proceeds to step S12 in Figure 11.

[0104] Alternatively, steps S114 and S115 may be removed, and the system may be configured so that if the result in step 113 is determined to be "No", the process in step S116 is performed. Furthermore, if it is determined that the number of groups obtained after the process in step S116 is still large, the system may return to step S114 and repeat the processes from step S114 onward.

[0105] Returning to Figure 11, in step S12, the controller 51 (learning unit 512) trains the group-specific AI models prepared for each group obtained in the group classification in step S11. Through this training, multiple trained group-specific AI models 40 are obtained as shown in Figure 7. Note that the group-specific AI models before training may be newly prepared, or they may use a previously generated temporary model (the AI ​​model used for group classification) of one of the notification contents that constitute the same group.

[0106] The training of multiple group-based AI models is performed for each group separately, using a large amount of ground-correct labeled data (training dataset) prepared separately for each group. The training method is supervised learning, as described above. The training of multiple group-based AI models may be performed one by one in sequence, or it may be performed simultaneously in parallel using multiple devices (processors). Once the training of multiple group-based AI models is complete, the process proceeds to the next step S13.

[0107] Furthermore, the training datasets prepared separately for each group may be composed of data for each of the multiple types of notification content that belong to a group. However, alternatively, the training datasets prepared separately for each group may be composed of data for only some of the multiple types of notification content that belong to a group, such as just one of them. Even in this case, since notification content with similar output tendencies to the AI ​​model's inputs belongs to the same group, it is unlikely that the performance of the group-specific AI models 40 obtained after training will deteriorate. Moreover, reducing the number of types of notification content used for training in this way may reduce the preparation burden when collecting training data.

[0108] Furthermore, since the trained group-specific AI models 40 are the main models that constitute the HMI control model 4, it is preferable to prepare a larger amount of training data during training in step S12 compared to the case of training a provisional model. Even in this case, the number of AI models that need to be generated by grouping is reduced compared to the case where an AI model is generated individually for each notification content, which helps to suppress the increased effort required to prepare the training data.

[0109] In step S13, the controller 51 (aggregation unit 513) aggregates the multiple group-specific AI models 40, which have been trained separately, into one. This generates an HMI control model 4 composed of the multiple group-specific AI models 40.

[0110] As described above, in the AI ​​model generation method of this embodiment, each group-specific AI model prepared for each group classified by group division is separately trained using the ground truth labeled data collected for each group to generate an AI model (HMI control model 4). This makes it possible to generate an AI model with a smaller overall size compared to the case where an AI model is constructed by combining multiple individual AI models set up for each notification content.

[0111] The generated, trained HMI control model 4 (specifically, model information) is stored, for example, on various non-volatile recording media readable by a computer, or made available for download via a communication network such as the Internet, so that it can be used by the information notification control device 1.

[0112] Furthermore, after generating the AI ​​model (HMI control model 4) according to the process shown in Figure 11, new notification content may be added. In such cases, the contribution of each input variable to the new notification content can be determined using a provisional model, as in the case described above, and the system can be configured to determine whether or not it can be incorporated into one of the existing groups using equation (1) above. If it is determined that it can be incorporated into any of the groups, it is not necessary to generate a new trained group-specific AI model. If the group classification using equation (1) above determines that it cannot be incorporated into any of the existing groups, it may be incorporated into one of the groups using one of the other classification methods described above. Alternatively, an AI model with natural language processing capabilities can be used to detect a group from among the existing groups that is determined to be similar to the language constituting the new notification content, and the new notification content can be automatically incorporated into the detected group. If it is determined that it cannot be incorporated into any of the existing groups, a group-specific AI model corresponding to the new notification content should be generated.

[0113] Furthermore, although the above describes the case where the HMI control model 4 is composed of multiple group-specific AI models 40, as mentioned above, the HMI control model 4 may also be composed of a single AI model. When generating an AI model with such a configuration, group values ​​will be added to the input information input to the AI ​​model during training.

[0114] <5. Things to keep in mind> The various technical features disclosed in the embodiments for carrying out the invention as described herein can be modified in various ways without departing from the spirit of the technical creation. Furthermore, the multiple embodiments and modifications disclosed in the embodiments for carrying out the invention as described herein may be combined to the extent possible. [Explanation of Symbols]

[0115] 1. Information notification control device 2... Information source device 3... Information notification device 4. 4A...HMI control model (AI model) 20. Group-specific AI models 121... Program (Information Notification Control Program) 100...Driving assistance system 200...Group Variable Conversion Table (Conversion Information) 521... Program (AI Model Generation Program) VE...vehicle

Claims

1. An information notification control device that controls the notification of information, Obtain the value of the group variable assigned to each group, which is determined according to the content of the notification of the acquired information. The notification parameters are estimated by running an AI model based on the values ​​of the group variables and other input variables. An information notification control device that controls the notification based on the notification parameters.

2. The AI ​​model includes a plurality of group-specific AI models, each separately provided for each group. The information notification control device according to claim 1, wherein the estimation of the notification parameters is performed using a group-specific AI model selected from the plurality of group-specific AI models according to the value of the group variable.

3. The information notification control device according to claim 1, wherein the AI ​​model is a shared AI model used by all of the aforementioned groups.

4. The information notification control device according to any one of claims 1 to 3, wherein each of the aforementioned groups includes a collection of notification contents that include the same type of input variable, with respect to the input variable that primarily affects the estimation result of the notification parameter among the other input variables.

5. The information notification control device according to any one of claims 1 to 3, wherein the other input variables include at least one of the following: a variable relating to the recipient of the notification, a variable relating to the surrounding information of the recipient of the notification, and a variable relating to the priority of the notification.

6. The information notification control device according to any one of claims 1 to 3, wherein the value of the group variable is obtained using conversion information that associates the relationship between the content of the notification and the value of the group variable.

7. A program that causes a computer to execute an information notification control method for controlling the notification of information, The aforementioned computer, Obtain the value of the group variable assigned to each group, which is determined according to the content of the notification of the acquired information, The notification parameters are estimated by running an AI model based on the values ​​of the group variables and the values ​​of other input variables. The notification is controlled based on the notification parameters, An information notification and control program that functions as a means to carry out [the action].

8. A driver assistance system that assists in driving a vehicle, An information notification device that provides information to support the aforementioned operation, A source information device that serves as the source of the aforementioned information, An information notification control device that performs notification control to the information notification device upon acquisition of the information from the information source device, Equipped with, The aforementioned information notification control device is Obtain the value of the group variable assigned to each group, which is determined according to the content of the notification of the acquired information. The notification parameters are estimated by running an AI model based on the values ​​of the group variables and other input variables. A driver assistance system that performs notification control based on the notification parameters.

9. A method for generating an AI model that estimates notification parameters used to control information notifications, Grouping will be performed based on the content of the aforementioned notification. An AI model generation method comprising generating the AI ​​model by separately training each of the group-specific AI models prepared for each of the groups classified according to the above group division, using data with correct labels collected for each group.

10. The AI ​​model generation method according to claim 9, wherein, in the grouping, the contents of the notifications that are presumed to have a similar relationship between the input to the AI ​​model and the output thereof are placed in the same group.

11. A program that causes a computer to execute a method for generating an AI model used to estimate notification parameters for controlling information notifications, The aforementioned computer, The AI ​​models are generated by separately training each of the group-specific AI models prepared for each group classified according to the content of the notification, using the properly labeled data collected for each group. An AI model generation program that functions as a means to execute [the specified action].

Images