Judgment system, judgment program, judgment device, and judgment method

Abstract

To provide a determination system, determination program, determination device, and determination method that can appropriately determine the discrepancy between the actions a subject should take and the actions the subject actually takes. [Solution] The system comprises: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is the action the subject should take in response to the surrounding circumstances, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information to identify the state of the subject's actions; a second identification means for identifying the subject's actions, which is the actions of the subject, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a discrepancy-related determination, which is a determination of the discrepancy between the standard action identified by the first identification means and the subject's actions identified by the second identification means.

Description

Technical Field

[0001] The present invention relates to a determination system, a determination program, a determination device, and a determination method.

Background Art

[0002] Conventionally, a technique for monitoring the safety of a user has been known (for example, Patent Document 1).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] By the way, a technique for determining the deviation between the action that a target person should take and the action of the target person has been desired.

[0005] The present invention has been made in view of the above, and an object thereof is to provide a determination system, a determination program, a determination device, and a determination method capable of appropriately determining the deviation between the action that a target person should take and the action of the target person.

Means for Solving the Problems

[0006] To solve the above-mentioned problems and achieve the objective, the determination system described in claim 1 comprises: a first acquisition means for acquiring ambient situation information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding circumstances of the subject and is related to movement, based on the ambient situation information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information for identifying the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action of the subject and is related to movement, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination regarding the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means, wherein the first identification means is the subject Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Based on the above, one of the candidate standard actions predetermined in a plurality of candidate standard actions is identified as the standard action, and the second identification means, Level of danger Based on the criteria, the determination means identifies one of the candidate target behaviors predetermined from a plurality of candidate target behaviors as the target behavior, and the determination means Level of danger Based on this, the deviation relationship determination is performed, and the determination means determines the corresponding standard action identified by the first identification means. Level of danger The first Level of danger And, before the target person's actions identified by the second identification means Level of danger The second Level of danger The determination of whether or not there is a discrepancy between them is performed as the aforementioned discrepancy relationship determination.

[0007] The determination system according to claim 2 comprises: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding circumstances of the subject and is related to movement, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information for identifying the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action of the subject and is related to movement, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination regarding the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means, wherein the first identification means determines the subject's Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Based on the above, one of the candidate standard actions predetermined in a plurality of candidate standard actions is identified as the standard action, and the second identification means, Level of danger Based on the criteria, the determination means identifies one of the candidate target behaviors predetermined from a plurality of candidate target behaviors as the target behavior, and the determination means Level of danger Based on this, the deviation relationship determination is performed, and the determination means determines the corresponding standard action identified by the first identification means. Level of danger The first Level of danger The second identifying means corresponds to the action of the target person identified by the second identifying means. Level of danger The second Level of danger The degree of discrepancy between these two is determined as the aforementioned discrepancy-related determination.

[0008] The determination system according to claim 3 comprises: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding circumstances of the subject and is related to movement, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information for identifying the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action of the subject and is related to movement, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination of the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means, wherein the first identification means determines the subject's Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Based on the above, one of the candidate standard actions predetermined in a plurality of candidate standard actions is identified as the standard action, and the second identification means, Level of danger Based on the criteria, the determination means identifies one of the candidate target behaviors predetermined from a plurality of candidate target behaviors as the target behavior, and the determination means Level of danger Based on this, the deviation relationship determination is performed, and the determination means determines the corresponding standard action identified by the first identification means. Level of danger The first Level of danger The second identifying means corresponds to the action of the target person identified by the second identifying means. Level of danger The second Level of danger In the case where the two are diverging from each other, the first Level of danger and the second Level of danger The aforementioned between each other Level of danger The determination of the relative magnitudes is performed as the aforementioned deviation relationship determination.

[0009] The determination program described in claim 4 causes the computer to function as: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding circumstances of the subject and is related to movement, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information to identify the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action of the subject and is related to movement, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a discrepancy-related determination, which is a determination regarding the discrepancy between the standard action identified by the first identification means and the subject-side action identified by the second identification means, wherein the first identification means determines the subject's Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Based on the above, one of the candidate standard actions predetermined in a plurality of candidate standard actions is identified as the standard action, and the second identification means, Level of danger Based on the criteria, the determination means identifies one of the candidate target behaviors predetermined from a plurality of candidate target behaviors as the target behavior, and the determination means Level of danger Based on this, the deviation relationship determination is performed, and the determination means determines the corresponding standard action identified by the first identification means. Level of danger The first Level of danger The second identifying means corresponds to the action of the target person identified by the second identifying means. Level of danger The second Level of danger The determination of whether or not there is a discrepancy between them is performed as the aforementioned discrepancy relationship determination.

[0010] The determination program according to claim 5 causes a computer to function as: a first acquisition means for acquiring ambient situation information indicating the situation around the target person; a first specifying means for specifying a reference action, which is an action to be taken by the target person corresponding to the situation around the target person and is related to movement, based on the ambient situation information acquired by the first acquisition means; a second acquisition means for acquiring action state information for specifying the state of the target person's action; a second specifying means for specifying a target person's action related to movement, which is the target person's action based on the action state information acquired by the second acquisition means; and a determination means for performing a divergence-related determination, which is a determination regarding the divergence between the reference action specified by the first specifying means and the target person's action specified by the second specifying means. The first specifying means specifies any one of a plurality of candidate reference actions predetermined based on Avoiding danger related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. as the reference action, and the second specifying means specifies any one of a plurality of candidate target person's actions predetermined based on Level of danger as the target person's action. The determination means performs the divergence-related determination based on Level of danger . The determination means makes a determination regarding the degree of divergence between the first Level of danger corresponding to the reference action specified by the first specifying means and the second Level of danger corresponding to the target person's action specified by the second specifying means as the divergence-related determination.

[0011] ​​​​The determination program according to claim 6 causes a computer to function as: a first acquisition means for acquiring ambient situation information indicating the situation around the target person; a first specification means for specifying a reference action, which is an action to be taken by the target person corresponding to the situation around the target person and is related to movement, based on the ambient situation information acquired by the first acquisition means; a second acquisition means for acquiring action state information for specifying the state of the target person's action; a second specification means for specifying a target person's action related to movement, which is the target person's action, based on the action state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination regarding the deviation between the reference action specified by the first specification means and the target person's action specified by the second specification means. The first specification means specifies, as the reference action, any one of a plurality of candidate reference actions predetermined based on Avoiding danger related to achieving the goal The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Among the plurality of candidate reference actions, any one of the candidate reference actions is specified as the reference action. The second specification means specifies, as the target person's action, any one of a plurality of candidate target person's actions predetermined based on Level of danger Among the plurality of candidate target person's actions, any one of the candidate target person's actions is specified as the target person's action. The determination means performs the deviation-related determination based on Level of danger Based on this, the determination means determines the magnitude relationship between the first Level of danger corresponding to the reference action specified by the first specification means Level of danger and the second Level of danger corresponding to the target person's action specified by the second specification means Level of danger When they deviate from each other, the determination means makes the determination regarding the magnitude relationship between the first Level of danger and the second Level of danger as the deviation-related determination. Level of danger

[0012] ​The determination device according to claim 7 comprises: a first acquisition means for acquiring ambient information indicating the surrounding conditions of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding conditions of the subject and is related to movement, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information for identifying the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action of the subject and is related to movement, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination of the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means, wherein the first identification means determines the subject's Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Based on the above, one of the candidate standard actions predetermined in a plurality of candidate standard actions is identified as the standard action, and the second identification means, Level of danger Based on the criteria, the determination means identifies one of the candidate target behaviors predetermined from a plurality of candidate target behaviors as the target behavior, and the determination means Level of danger Based on this, the deviation relationship determination is performed, and the determination means determines the corresponding standard action identified by the first identification means. Level of danger The first Level of danger The second identifying means corresponds to the action of the target person identified by the second identifying means. Level of danger The second Level of danger The determination of whether or not there is a discrepancy between them is performed as the aforementioned discrepancy relationship determination.

[0013] The determination device according to claim 8 comprises: a first acquisition means for acquiring ambient information indicating the surrounding conditions of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding conditions of the subject and is related to movement, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information for identifying the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action of the subject and is related to movement, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination of the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means, wherein the first identification means determines the subject's Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Based on the above, one of the candidate standard actions predetermined in a plurality of candidate standard actions is identified as the standard action, and the second identification means, Level of danger Based on the criteria, the determination means identifies one of the candidate target behaviors predetermined from a plurality of candidate target behaviors as the target behavior, and the determination means Level of danger Based on this, the deviation relationship determination is performed, and the determination means determines the corresponding standard action identified by the first identification means. Level of danger The first Level of danger The second identifying means corresponds to the action of the target person identified by the second identifying means. Level of danger The second Level of danger The degree of discrepancy between these two is determined as the aforementioned discrepancy-related determination.

[0014] The determination device according to claim 9 comprises: a first acquisition means for acquiring ambient information indicating the surrounding conditions of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding conditions of the subject and is related to movement, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information for identifying the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action of the subject and is related to movement, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination of the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means, wherein the first identification means determines the subject's Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. Based on the above, one of the candidate standard actions predetermined in a plurality of candidate standard actions is identified as the standard action, and the second identification means, Level of danger Based on the criteria, the determination means identifies one of the candidate target behaviors predetermined from a plurality of candidate target behaviors as the target behavior, and the determination means Level of danger Based on this, the deviation relationship determination is performed, and the determination means determines the corresponding standard action identified by the first identification means. Level of danger The first Level of danger The second identifying means corresponds to the action of the target person identified by the second identifying means. Level of danger The second Level of danger In the case where the two are diverging from each other, the first Level of danger and the second Level of danger The aforementioned between each other Level of danger The determination of the relative magnitudes is performed as the aforementioned deviation relationship determination.

[0015] The determination method according to claim 10 includes: a first acquisition step of acquiring ambient information indicating the surrounding circumstances of a subject; a first identification step of identifying a standard action related to movement, which is an action that the subject should take in response to the surrounding circumstances of the subject, based on the ambient information acquired in the first acquisition step; a second acquisition step of acquiring behavioral state information for identifying the state of the subject's actions; a second identification step of identifying a subject-side action related to movement, which is an action of the subject, based on the behavioral state information acquired in the second acquisition step; and a determination step of performing a deviation-related determination, which is a determination of the deviation between the standard action identified in the first identification step and the subject-side action identified in the second identification step, wherein the first identification step involves the subject Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. In the second identification step, one of the candidate standard actions predetermined based on the criteria is identified as the standard action, and in the second identification step, Level of danger Based on the criteria, one of the candidate target behaviors predetermined is identified as the target behavior, and in the determination step, Level of danger Based on the above, the deviation relationship determination is performed, and in the determination step, the criteria behavior identified in the first identification step corresponds to the above Level of danger The first Level of danger and the subject actions identified in the second identification step correspond to the aforementioned actions of the subject. Level of danger The second Level of danger The determination of whether or not there is a discrepancy between them is performed as the aforementioned discrepancy relationship determination. The determination method according to claim 11 includes: a first acquisition step of acquiring ambient information indicating the surrounding circumstances of a subject; a first identification step of identifying a standard action related to movement, which is an action that the subject should take in response to the surrounding circumstances of the subject, based on the ambient information acquired in the first acquisition step; a second acquisition step of acquiring behavioral state information for identifying the state relating to the subject's actions; a second identification step of identifying a subject-side action related to movement, which is an action of the subject, based on the behavioral state information acquired in the second acquisition step; and a determination step of performing a deviation-related determination, which is a determination of the deviation between the standard action identified in the first identification step and the subject-side action identified in the second identification step, wherein the first identification step involves the subject Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. In the second identification step, one of the candidate standard actions predetermined based on the criteria is identified as the standard action, and in the second identification step, Level of danger Based on the criteria, one of the candidate target behaviors predetermined is identified as the target behavior, and in the determination step, Level of danger Based on the above, the deviation relationship determination is performed, and in the determination step, the criteria behavior identified in the first identification step corresponds to the above Level of danger The first Level of danger and the subject actions identified in the second identification step correspond to the aforementioned actions of the subject. Level of danger The second Level of danger The degree of discrepancy between these two is determined as the aforementioned discrepancy-related determination. The determination method according to claim 12 includes: a first acquisition step of acquiring ambient information indicating the surrounding circumstances of a subject; a first identification step of identifying a standard action that corresponds to the surrounding circumstances of the subject and is related to movement, based on the ambient information acquired in the first acquisition step; a second acquisition step of acquiring behavioral state information for identifying the state of the subject's actions; a second identification step of identifying a subject-side action that corresponds to the subject's actions and is related to movement, based on the behavioral state information acquired in the second acquisition step; and a determination step of performing a deviation-related determination, which is a determination of the deviation between the standard action identified in the first identification step and the subject-side action identified in the second identification step, wherein the first identification step involves the subject Avoiding danger Related to achieving the objective The degree of risk, which indicates the degree of the risk that the subject will be involved in a traffic accident. In the second identification step, one of the candidate standard actions predetermined based on the criteria is identified as the standard action, and in the second identification step, Level of danger Based on the criteria, one of the candidate target behaviors predetermined is identified as the target behavior, and in the determination step, Level of danger Based on the above, the deviation relationship determination is performed, and in the determination step, the criteria behavior identified in the first identification step corresponds to the above Level of danger The first Level of danger and the subject actions identified in the second identification step correspond to the aforementioned actions of the subject. Level of danger The second Level of danger In the case where the two are diverging from each other, the first Level of danger and the second Level of danger The aforementioned between each other Level of danger The determination of the relative magnitudes is performed as the aforementioned deviation relationship determination. The determination system according to claim 13 further comprises a notification means for providing notification regarding the subject based on the determination result of the determination means, in the determination system according to any one of claims 1 to 3. [Effects of the Invention]

[0016] According to the determination system described in claim 1, the determination program described in claim 4, the determination device described in claim 7, and the determination method described in claim 10, by performing a deviation-related determination, which is a determination regarding the deviation between the standard behavior and the behavior of the subject, it becomes possible to appropriately determine, for example, the deviation between the behavior that the subject should take and the behavior of the subject. Furthermore, by identifying one of the multiple candidate standard behaviors as the standard behavior, and one of the multiple candidate target behaviors as the target behavior, for example, the standard behavior and the target behavior can be appropriately identified, making it possible to appropriately determine the discrepancy between the behavior the target should take and the behavior of the target. 。

[0017] According to the determination system described in claim 2, the determination program described in claim 5, the determination device described in claim 8, and the determination method described in claim 11, by performing a deviation-related determination, which is a determination regarding the deviation between the standard behavior and the behavior of the subject, it becomes possible to appropriately determine, for example, the deviation between the behavior that the subject should take and the behavior of the subject. Furthermore, by identifying one of the multiple candidate standard behaviors as the standard behavior, and one of the multiple candidate target behaviors as the target behavior, for example, the standard behavior and the target behavior can be appropriately identified, making it possible to appropriately determine the discrepancy between the behavior the target should take and the behavior of the target. 。

[0018] According to the determination system described in claim 3, the determination program described in claim 6, the determination device described in claim 9, and the determination method described in claim 12, by performing a deviation-related determination, which is a determination regarding the deviation between the standard behavior and the behavior of the subject, it becomes possible to appropriately determine, for example, the deviation between the behavior that the subject should take and the behavior of the subject. Furthermore, by identifying one of the multiple candidate standard behaviors as the standard behavior, and one of the multiple candidate target behaviors as the target behavior, for example, the standard behavior and the target behavior can be appropriately identified, making it possible to appropriately determine the discrepancy between the behavior the target should take and the behavior of the target. 。

[0019] According to the determination system described in claim 13, by providing notification regarding the subject based on the determination result of the determination means, it becomes possible to provide useful information regarding the discrepancy between the action the subject should take and the action the subject is actually taking. [Brief explanation of the drawing]

[0022] [Figure 1] This is a block diagram of a terminal device according to an embodiment of the present invention. [Figure 2] This is a flowchart for the process of identifying standard behaviors. [Figure 3] This is a plan view of the intersection. [Figure 4] This is a diagram illustrating information about the surrounding environment. [Figure 5] This is an explanatory diagram of candidate criteria behavioral information. [Figure 6] This is a flowchart for identifying the actions of the target individual. [Figure 7] This is an explanatory diagram of the behavioral information of the candidate. [Figure 8] This is a flowchart of the decision-making process. [Figure 9] This is a diagram explaining each judgment. [Modes for carrying out the invention]

[0023] Hereinafter, embodiments of the determination system, determination program, determination device, and determination method according to the present invention will be described in detail with reference to the drawings. However, the present invention is not limited by these embodiments. Here, the basic concepts and terminology will be explained, followed by a description of specific embodiments.

[0024] (Basic concept) First, let me explain the basic concept. The judgment system according to the present invention is a system that performs discrepancy-related judgment, which is a judgment regarding the discrepancy between standard behavior and the subject's behavior, and the concept includes, for example, a system used to help the subject achieve their goals.

[0025] "Target entity" is a concept that refers to persons, animals, and robots to which the judgment system applies, and includes, for example, humans (adults, children), pets (dogs, cats, etc.), and designated robots.

[0026] "Standard behavior" is a concept that indicates the actions that a subject should take, and specifically, it refers to actions that correspond to the circumstances surrounding the subject.

[0027] "Subject behavior" is a concept that describes the actions of the subject, and includes, for example, the subject's current actions, past actions, and future actions.

[0028] In the embodiments described below, for example, the "target person" is a child (specifically, an elementary school student), and the objective is to avoid danger to the target person.

[0029] (composition) First, the terminal device according to this embodiment will be described. Figure 1 is a block diagram of the terminal device according to an embodiment of the present invention.

[0030] The terminal device 100 is a judgment system and judgment device, and is, for example, a device carried by the target user (elementary school student).

[0031] The physical configuration of the terminal device 100 is arbitrary, but for example, it can be attached to a predetermined location on the user. The terminal device 100 can be attached, for example, to a predetermined part (e.g., the shoulder strap) of a school bag that the user carries when going to and from school. When the terminal device 100 is attached to the shoulder strap of the school bag while the user is wearing and holding the school bag, the terminal device 100 will be attached to the predetermined location on the user via the shoulder strap, facing forward (the direction in which the user is moving). In other words, the terminal device 100 will be held by the user. Note that "forward side" may also be referred to as "front side".

[0032] The configuration described here is merely an example, and other configurations may be adopted (for example, a configuration in which the device hangs around the user's neck, a configuration attached to the user's clothing or shoes, a configuration stored inside a pocket or school bag, a configuration in which the terminal device moves in accordance with the body at a location away from the body (a state in which it is held or in contact with the body), or a configuration that utilizes information from fixed infrastructure such as surveillance cameras).

[0033] The terminal device 100 in Figure 1 includes, for example, a camera 11, a sensor 12, a speaker 13, a recording unit 14, and a control unit 15.

[0034] (Configuration - Camera) Camera 11 is a means of capturing images of the surroundings of the user (hereinafter also simply referred to as "user") holding the terminal device 100. The specific type and configuration of camera 11 are arbitrary, but for example, the explanation will use a camera with a fisheye lens configured to capture video of the area in front of the user as an example.

[0035] (Configuration - Sensor) Sensor 12 is a detection means for detecting information to identify the state of the user's behavior. The specific type and configuration of sensor 12 are arbitrary, but for example, an example using an accelerometer to detect the user's acceleration will be explained.

[0036] (Configuration - Speakers) Speaker 13 is an acoustic means that outputs sound (including various sounds such as alarm sounds). The specific type and configuration of speaker 13 are arbitrary, but for example, the explanation will be given using a known speaker that outputs various sounds.

[0037] (Composition - Record Section) The recording unit 14 is a recording means (storage means) for recording programs and various data necessary for the operation of the terminal device 1, and is configured, for example, using flash memory (not shown) as an external recording device. However, any other recording medium can be used instead of flash memory.

[0038] (Configuration - Control Unit) The control unit 15 in Figure 1 is a control means for controlling the terminal device 1, and is specifically a computer comprising a CPU, various programs interpreted and executed on the CPU (including basic control programs such as an OS, and application programs launched on the OS to realize specific functions), and internal memory such as RAM for storing programs and various data. In particular, the program according to this embodiment is installed on the terminal device 1 via any recording medium or network, and substantially constitutes each part of the control unit 15.

[0039] Functionally, the control unit 15 includes, for example, a first acquisition means, a first identification means, a second acquisition means, a second identification means, a determination means, and a notification means.

[0040] ===First acquisition method=== The first acquisition method is a means of acquiring ambient information that shows the surrounding circumstances of the subject.

[0041] "Surrounding information" refers to information that indicates the circumstances surrounding a subject, and is a concept that includes, for example, image information (videos and still images), audio information, vibration information, pressure information, wind speed information, location information (location information that can be determined by GPS, etc.), and map information that indicate the circumstances surrounding the user who is the subject. Furthermore, location information may be interpreted to include not only absolute location information, but also relative location information in relation to other people or objects besides the subject.

[0042] In this embodiment, we will explain the case where "surrounding information" is image information (video).

[0043] ===First identification means=== The first identification means is a means for identifying a standard action, which is the action that the subject should take in response to the circumstances surrounding the subject, based on the surrounding circumstances information acquired by the first acquisition means. For example, the first identification means identifies one of several candidate standard actions predetermined based on the degree of relevance to achieving the subject's objective as the standard action.

[0044] "Degree of relevance to achieving the objective" is a concept that indicates the degree to which something is relevant to the achievement of the target person's objective. For example, it includes the degree of risk (degree of danger) relevant to the achievement of the objective of a user (elementary school student or blind person, etc.) avoiding danger, and the degree of relevance to the achievement of the objective of a user (elderly person) preventing wandering or monitoring pets (degree of distance from a designated location, for example).

[0045] In this embodiment, the "degree of relevance to achieving the objective" is explained by illustrating the case where the degree of risk is related to the user's (elementary school student's) objective of avoiding danger (for example, avoiding dangers related to traffic).

[0046] A "candidate standard behavior" is a behavior that is designated as a standard behavior, and is predetermined based on criteria such as its relevance to achieving the objective.

[0047] ===Second acquisition means=== The second acquisition method is a means of acquiring behavioral state information to identify the state of the subject's actions.

[0048] "Behavioral status information" refers to information used to identify the state of a subject's behavior. This concept includes, for example, information indicating the subject's acceleration, angular velocity (measured by an inertial sensor), vibration, direction (the direction the subject is facing), location information (location information that can be determined by GPS, etc.), vital information such as heart rate and electroencephalogram, gaze, head position, and limb movements. It should be noted that location information may be interpreted to include not only absolute location information but also relative location information in relation to other people or objects.

[0049] In this embodiment, we will explain the case where the "behavioral state information" is information indicating the subject's acceleration.

[0050] ===Second identification means=== The second identification means is a means for identifying a subject's behavior, which is the subject's behavior, based on the behavioral state information acquired by the second acquisition means. The second identification means is, for example, a means for identifying one of several candidate subject's behaviors predetermined based on the degree of relevance to achieving the objective as the subject's behavior.

[0051] "Candidate target behaviors" are behaviors that are identified as candidate behaviors on the target side, and are predetermined based on criteria such as their degree of relevance to achieving the objective.

[0052] ===Judgment means=== The determination means is a means for performing a deviation-related determination, which is a determination of the deviation between the standard behavior identified by the first identification means and the target person's behavior identified by the second identification means. The determination means performs the deviation-related determination based, for example, on the degree of relevance to achieving the objective.

[0053] The determination means, for example, performs a determination of whether or not there is a discrepancy between the first degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the standard behavior identified by the first identification means, and the second degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the target person's behavior identified by the second identification means (hereinafter also referred to as "discrepancy determination"), as a discrepancy relevance determination.

[0054] The determination means, for example, performs a determination of the degree of discrepancy between the first degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the standard behavior identified by the first identification means, and the second degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the target person's behavior identified by the second identification means (hereinafter also referred to as "determination of degree of discrepancy"), as a determination of discrepancy relevance.

[0055] The determination means performs a determination of the relative magnitude of the degree of relevance to achieving the objective between the first degree of relevance to achieving the objective and the second degree of relevance to achieving the objective, when there is a discrepancy between the first degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the standard behavior identified by the first identification means, and the second degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the target person's behavior identified by the second identification means (hereinafter also referred to as "discrepancy magnitude determination"), as a discrepancy relationship determination.

[0056] ===Notification means=== The notification means is a means of providing notification regarding the subject based on the determination result of the determination means. The recipient of the notification by this notification means is arbitrary and includes, for example, the subject themselves, people around the subject, or people at a distance (people located far from the subject). Furthermore, the recipient of the notification by the notification means may be interpreted as not being limited to people, but may also include, for example, animals or machines such as robots.

[0057] "Notification concerning the target person" is a concept that refers to any notification related to the target person, and includes, for example, warnings, alerts, cautions, suggestions, recommendations, inquiries, and various messages.

[0058] The processing performed by the control unit 15 configured in this way will be described later.

[0059] (process) Next, we will describe the processes performed by the terminal device 100 configured in this way, such as the standard behavior identification process, the target person's behavior identification process, and the judgment process.

[0060] (Processing - Standard Action Identification Process) First, let's explain the standard behavior identification process. Figure 2 is a flowchart of the standard behavior identification process (hereinafter, each step will be referred to as "S"). The standard behavior identification process is the process of identifying a standard behavior, and is, for example, the process executed by the terminal device 100 in Figure 1.

[0061] The timing of executing this standard behavior identification process is arbitrary, but for example, it will be repeatedly started after the power of the terminal device 100 is turned on, and the explanation will begin from the point when the execution has started (the same applies to the subject-side behavior identification process and judgment process described later).

[0062] Figure 3 is a plan view of the intersection. Here, we will illustrate and explain the case where the user is facing an intersection with poor visibility due to obstructions, as shown in Figure 3.

[0063] ===SA1=== In SA1 in Figure 2, the control unit 15 acquires ambient information. Specifically, although it is arbitrary, for example, video images captured by camera 11 are collected in units of a predetermined time (for example, 3 to 5 seconds), and these video images are collected as ambient information for each unit.

[0064] Figure 4 is an example of surrounding situation information. Here, for example, if the user is facing the direction of the intersection shown in Figure 3, a video image (a video image grouped by each unit) including the image in Figure 4 that shows the area in front of the user (the direction of travel) is acquired as surrounding situation information.

[0065] ===SA2=== In SA2 shown in Figure 2, the control unit 15 identifies a reference action based on the surrounding situation information acquired by SA1. Specifically, although arbitrary, it is identified using, for example, a first trained model.

[0066] ==First Trained Model== The "first pre-trained model" is a model used to identify a standard behavior based on surrounding context information. For example, when a video image is input as surrounding context information, this model outputs one candidate standard behavior from a predetermined set of candidate standard behaviors that corresponds to the input video image.

[0067] Figure 5 is an explanatory diagram of candidate criterion behavior information. "Candidate criterion behavior information" refers to information that indicates the aforementioned candidate criterion behaviors, and is information determined based on the degree of risk.

[0068] "Risk level" refers to the degree of danger related to traffic, specifically the likelihood of being involved in a traffic accident. For example, a higher numerical value indicates a higher degree of danger.

[0069] In this embodiment, as shown in Figure 5, for example, "standing still" is defined as the candidate criterion behavior information corresponding to "3," which has the highest risk level; "walking slowly" is defined as the candidate criterion behavior information corresponding to "2," which has the second highest risk level; and "walking" is defined as the candidate criterion behavior information corresponding to "1," which has the lowest risk level.

[0070] In other words, we will explain the case where the first trained model is configured to output one of the three candidate criterion behaviors shown in Figure 5 ("walking", "walking slowly", "standing still").

[0071] The training method for such a first pre-trained model is arbitrary, but for example, it may be generated by training with supervised machine learning. Specifically, a large number of training data (e.g., several hundred to several thousand) consisting of video images as ambient situation information and candidate criterion action information corresponding to those video images may be generated, and the model may be generated by training with machine learning using the generated training data.

[0072] Furthermore, the aforementioned training data may reflect the knowledge of adults who can accurately avoid traffic hazards. For example, for video footage of areas near intersections with poor visibility, the danger is extremely high, so training data may be generated that is associated with "stopped," which corresponds to a danger level of "3." Training data including other types of video footage may be generated from a similar perspective.

[0073] ==Processing== For the processing, it is assumed that the first trained model described above is recorded in the recording unit 14. Then, the surrounding situation information acquired by SA1 is input to the first trained model, one candidate standard action information is obtained from the first trained model, and the candidate standard action indicated by the obtained candidate standard action information is identified as the user's standard action.

[0074] Here, for example, ambient information, which is a video image including the image in Figure 4, is input to the first trained model. Since the ambient information in Figure 4 shows an intersection with poor visibility, the first trained model outputs "Candidate Criterion Action Information" = "Stopping" which corresponds to "Danger Level" = "3" in Figure 5. In this case, "Stopping" is identified as the user's criterion action.

[0075] Then, by repeatedly executing processes SA1 to SA2, the standard behavior will be identified repeatedly. This concludes the explanation of the standard behavior identification process.

[0076] (Processing - Identifying the actions of the target person) Next, we will explain the process of identifying the target person's actions. Figure 6 is a flowchart of the process of identifying the target person's actions. The process of identifying the target person's actions is the process of identifying the target person's actions, and is performed, for example, by the terminal device 100 in Figure 1.

[0077] This process for identifying the subject's behavior is executed repeatedly in parallel with the standard behavior identification process shown in Figure 2.

[0078] ===SB1=== In SB1 in Figure 6, the control unit 15 acquires behavioral state information. Specifically, although arbitrary, for example, it acquires time-series data of the user's acceleration detected by the sensor 12 in Figure 1 as behavioral state information. Note that "time-series data of acceleration" may be interpreted as referring to data corresponding to a set of accelerations detected at predetermined time intervals.

[0079] Here, for example, if a user starts walking slowly towards the intersection shown in Figure 3, time-series data of acceleration corresponding to walking slowly is obtained.

[0080] ===SB2=== In SB2 in Figure 6, the control unit 15 identifies the subject's behavior based on the behavioral state information acquired in SB1. Specifically, although arbitrary, it is identified using, for example, a second trained model.

[0081] ==Second pre-trained model== The "second pre-trained model" is a model used to identify subject behaviors based on behavioral state information. For example, when time-series data of acceleration is input as behavioral state information, this model outputs one candidate subject behavior information corresponding to the input time-series data of acceleration from a predetermined set of candidate subject behavior information.

[0082] Figure 7 is an explanatory diagram of the candidate subject's behavioral information. "Candidate subject's behavioral information" refers to the information indicating the candidate subject's behavior as described above, and is determined based on the degree of risk. The degree of risk is as explained in Figure 5.

[0083] In this embodiment, as shown in Figure 7, for example, the candidate subject's behavior information corresponding to the highest risk level "3" is defined as "standing still," the candidate subject's behavior information corresponding to the second highest risk level "2" is defined as "walking slowly," and the candidate subject's behavior information corresponding to the lowest risk level "1" is defined as "walking."

[0084] In other words, we will explain the case where the second trained model is configured to output one of the three candidate subject behavior information items shown in Figure 7 ("walking", "walking slowly", "standing still").

[0085] The training method for such a second pre-trained model is arbitrary, but for example, it may be generated by training with supervised machine learning. Specifically, a large number of training data (e.g., several hundred to several thousand) consisting of combinations of time-series data of acceleration as behavioral state information and candidate subject behavior information corresponding to the said acceleration time-series data may be generated, and the model may be generated by training with machine learning using the generated training data.

[0086] Furthermore, the aforementioned training data may be generated by experimentally obtaining time-series data of acceleration while many elementary school students are performing various actions ("walking," "walking slowly," and "standing still") by attaching acceleration sensors to them, or by identifying and obtaining all or part of the data through simulation.

[0087] ==Processing== For the processing, it is assumed that the second pre-trained model described above is recorded in the recording unit 14. Then, the behavioral state information acquired by SB1 is input to the second pre-trained model, one candidate target behavior information output from the second pre-trained model is acquired, and the candidate target behavior indicated by the acquired candidate target behavior information is identified as the user's target behavior.

[0088] Here, for example, behavioral state information, which is time-series data of acceleration corresponding to walking slowly, is input to the second trained model. The second trained model then outputs "Candidate Criterion Behavior Information" = "Walking slowly" which corresponds to "Risk Level" = "2" in Figure 7. In this case, "Walking slowly" is identified as the target behavior of the user.

[0089] Then, by repeatedly executing processes SB1 and SB2, the subject's actions will be identified repeatedly. This concludes the explanation of the subject's actions identification process.

[0090] (Processing - Judgment Processing) Next, the judgment process will be explained. Figure 8 is a flowchart of the judgment process. The judgment process is a process that performs a discrepancy-related judgment, and is, for example, a process executed by the terminal device 100 in Figure 1.

[0091] This determination process is repeatedly executed using the processing results of the standard behavior process shown in Figure 2 and the subject-side behavior identification process shown in Figure 6, which are executed repeatedly.

[0092] ===SC1=== In SC1 in Figure 8, the control unit 15 performs a deviation-related determination. A deviation-related determination is, for example, a determination regarding the deviation between the standard behavior and the behavior of the target person. The specific content of the deviation-related determination is arbitrary, but in this embodiment, we will explain the case in which the deviation-related determination is performed as a determination of whether or not there is a deviation in the degree of risk, a determination of the degree of deviation in the degree of risk, and a determination of the magnitude of the deviation in the degree of risk.

[0093] ==Danger Level Discrepancy Assessment== "Determination of deviation in risk level" refers to the aforementioned "determination of deviation level," and in this embodiment, it is a concept that indicates the determination of whether or not there is a deviation between the first risk level, which is the risk level corresponding to the standard behavior identified in SA2 in Figure 2, and the second risk level, which is the risk level corresponding to the subject's behavior identified in SB2 in Figure 6.

[0094] Furthermore, "the first risk level corresponding to the standard behavior identified in SA2 in Figure 2" (hereinafter also simply referred to as "the first risk level") refers to the risk level in Figure 5 that is associated with the candidate standard behavior information corresponding to that standard behavior. It is also possible to interpret "the first risk level" as corresponding to "the degree of relevance to achieving the first objective."

[0095] In other words, for example, if "walking" is identified in SA2 in Figure 2, the first risk level becomes "1", if "walking slowly" is identified in SA2 in Figure 2, the first risk level becomes "2", and if "standing still" is identified in SA2 in Figure 2, the first risk level becomes "3".

[0096] Furthermore, the "second risk level, which corresponds to the target person behavior identified in SB2 in Figure 6" (hereinafter also simply referred to as "second risk level") refers to the risk level in Figure 7 that is associated with the candidate target person behavior information corresponding to that target person behavior. It should also be interpreted that the "second risk level" corresponds to the "degree of relevance to achieving the second objective."

[0097] In other words, for example, if "walking" is identified in SB2 in Figure 6, the second risk level becomes "1", if "walking slowly" is identified in SB2 in Figure 6, the second risk level becomes "2", and if "standing still" is identified in SB2 in Figure 6, the second risk level becomes "3".

[0098] ==Assessment of the degree of deviation from the degree of risk== "Risk level deviation determination" refers to the aforementioned "deviation degree determination," and in this embodiment, it is a concept that indicates the determination of the degree of deviation between the first risk level and the second risk level.

[0099] ==Risk Level Discrepancy Assessment== "Risk degree deviation magnitude determination" is the same as the aforementioned "deviation magnitude determination," and in this embodiment, it is a concept that indicates the determination of the relative magnitude of the first risk degree and the second risk degree when the first risk degree and the second risk degree are in a state of mutual deviation.

[0100] ==Specific Judgment Method== The specific determination methods for the three determinations related to the above-mentioned deviation relationship are arbitrary, and they may be determined one by one. However, in this embodiment, we will focus on the fact that both the first risk level (i.e., the "risk level" information in Figure 5) and the second risk level (i.e., the "risk level" information in Figure 7) are numerical information, and explain the case in which we perform an operation to find the difference between the first risk level and the second risk level (i.e., the operation of "first risk level" - "second risk level"), and then make the three determinations based on the result of the operation.

[0101] Figure 9 is an explanatory diagram for each judgment. The "Calculation Result" column in Figure 9 shows the calculation result of "First Risk Level" - "Second Risk Level," and the "Risk Level Discrepancy Determination," "Risk Level Discrepancy Degree Determination," and "Risk Level Discrepancy Magnitude Determination" columns show examples of the judgment results for each judgment.

[0102] For example, if the result of the calculation is "0", it will be determined that there is no discrepancy, as shown in the top row of Figure 9.

[0103] Furthermore, for example, if the calculation result is "+1", it will be determined that there is a discrepancy, the degree of the discrepancy is small, and the first risk level is greater than the second risk level, as shown in the second row of Figure 9.

[0104] Furthermore, for example, if the calculation result is "+2", it will be determined that there is a discrepancy, the degree of the discrepancy is large, and the first risk level is greater than the second risk level, as shown in the third row of Figure 9.

[0105] Furthermore, for example, if the calculation result is "-1", it will be determined that there is a discrepancy, the degree of the discrepancy is small, and the first risk level is smaller than the second risk level, as shown in the fourth row of Figure 9.

[0106] Furthermore, for example, if the calculation result is "-2", it will be determined that there is a discrepancy, the degree of the discrepancy is large, and the first risk level is smaller than the second risk level, as shown in the fifth row of Figure 9.

[0107] ==Processing== Regarding the processing, first, the first risk level, which corresponds to the standard behavior identified in SA2 in Figure 2, is identified. The method of identification is arbitrary, but for example, the first risk level corresponding to the most recently identified standard behavior may be identified. In detail, the table information shown in Figure 5 (table information in which risk levels and candidate standard behavior information are mutually associated) is recorded in the recording unit 14, and a method of identification by referring to this table information may be used. Alternatively, the risk levels associated with standard behaviors may be written in the program code, and a method of identification using this written description may be used.

[0108] Here, for example, if SA2 in Figure 2 is identified as "stopped," then "3" is assigned as the first level of danger.

[0109] Furthermore, a second risk level, which corresponds to the risk level associated with the subject behavior identified in SB2 of Figure 6, is identified. This identification method is arbitrary, but for example, the second risk level corresponding to the most recently identified subject behavior may be identified. Specifically, the table information shown in Figure 7 (table information in which risk levels and candidate subject behavior information are mutually associated) is recorded in the recording unit 14, and a method of identification by referring to this table information may be used. Alternatively, the risk level associated with the subject behavior is written in the program code, and a method of identification using this written description may be used.

[0110] Here, for example, if "walking slowly" is identified in SB2 of Figure 6, then "2" is assigned as the second level of risk.

[0111] Next, a calculation is performed to subtract the identified second risk level from the identified first risk level, and a judgment corresponding to the calculation result (the judgment shown in Figure 9 above) is made.

[0112] Here, for example, if we identify "3" as the first risk level and "2" as the second risk level, we perform the calculation "3-2" and make a judgment corresponding to the calculation result "+1". In other words, as illustrated in the second row of Figure 9, we determine that there is a discrepancy, the degree of the discrepancy is small, and the first risk level is greater than the second risk level.

[0113] ===SC2=== In SC2 shown in Figure 8, the control unit 15 provides notification regarding the target person, who is a user, based on the determination result of SC1.

[0114] ==If the calculation result is "0"== Specifically, although optional, for example, if the calculation result in SC1 is "0" (see the top row of Figure 9) and it is determined that there is no discrepancy, then the degree of risk corresponding to the user's action (action of the target person) matches the degree of risk of the standard action, and from the perspective of achieving the objective of risk avoidance, the user's action can be judged as appropriate, and therefore no notification will be issued. Alternatively, in this case, in order to notify a recommendation message affirming the current action and recommending that the current action be maintained, the system may be configured to output the recommendation message as an audio message from speaker 13 in Figure 1.

[0115] ==If the calculation result is "+1"== Furthermore, for example, if the calculation result in SC1 is "+1" (see the second row of Figure 9), and it is determined that there is a discrepancy, the degree of the discrepancy is small, and the first level of risk is greater than the second level of risk, then the level of risk corresponding to the standard action (first level of risk) is greater than the level of risk corresponding to the user's action (target person's action) (second level of risk). From the perspective of achieving the objective of risk avoidance, the user's action is judged to be inappropriate, and in order to notify the user of a warning to encourage risk avoidance, a warning sound or warning message is output from speaker 13 in Figure 1.

[0116] In this case, since the degree of discrepancy was determined to be small, the audio will be output at a relatively low volume.

[0117] ==If the calculation result is "+2"== Furthermore, if, for example, the calculation result in SC1 is "+2" (see the third row of Figure 9), and it is determined that there is a discrepancy, the degree of the discrepancy is large, and the first level of danger is greater than the second level of danger, then, similar to the case of "+1", a warning sound or warning message will be output from speaker 13 in Figure 1 to notify a warning to encourage danger avoidance.

[0118] In this case, since it was determined that the degree of discrepancy was large, the audio will be output at a relatively high volume (higher than the volume when the calculation result is "+1" as described above).

[0119] ==If the calculation result is "-1"== Furthermore, for example, if the calculation result in SC1 is "-1" (see the fourth row of Figure 9), and it is determined that there is a discrepancy, the degree of the discrepancy is small, and the first risk level is smaller than the second risk level, then the risk level corresponding to the standard action (first risk level) is smaller than the risk level corresponding to the user's action (target person's action) (second risk level). From the perspective of achieving the objective of risk avoidance, the user's action can be judged to be appropriate. However, since the risk levels do not match, the speaker 13 in Figure 1 outputs an inquiry message as audio to notify the user of the reason for this discrepancy and to provide information (inquiry, etc.) to encourage the user to take action that corresponds to the situation around them.

[0120] In this case, since the degree of discrepancy was determined to be small, the audio will be output at a relatively low volume.

[0121] ==If the calculation result is "-2"== Furthermore, for example, if the calculation result in SC1 is "-2" (see the 5th row in Figure 9), and it is determined that there is a discrepancy, the degree of the discrepancy is large, and the first risk level is smaller than the second risk level, then, similar to the case of "-1", since the risk levels do not match, an inquiry message is output as audio from speaker 13 in Figure 1 to notify the user of information (such as an inquiry) to prompt them to take action in response to the situation around them by inquiring about the reason for the discrepancy.

[0122] In this case, since it was determined that the degree of discrepancy was large, the audio will be output at a relatively high volume (higher than the volume when the calculation result is "-1" as described above).

[0123] ===SC3=== In SC3 shown in Figure 8, the control unit 15 records various types of information.

[0124] Specifically, although optional, for example, the surrounding situation information acquired by SA1 in Figure 2 and the behavioral state information acquired by SB1 in Figure 6 may be recorded and held in the first area of ​​the recording unit 14 (an area for recording information for a certain period of time or a certain amount of data), and if a discrepancy is determined in SC1, the corresponding surrounding situation information and behavioral state information may be recorded in the second area of ​​the recording unit 14 (an area for recording and accumulating information until a deletion command is issued).

[0125] The information recorded in the second area is arbitrary. For example, if SC1 determines that there is a discrepancy and that the first risk level is greater than the second risk level, the system may be configured to record the corresponding surrounding situation information and behavioral state information. As a variation, the system may be configured to record the information if, within a certain period of time, no change in behavioral state information occurs that would reduce (i.e., eliminate) this discrepancy. This concludes the explanation of the determination process.

[0126] (Effects of the embodiment) Thus, according to this embodiment, by performing a discrepancy-related determination, which is a determination regarding the discrepancy between the standard behavior and the subject's behavior, it becomes possible to appropriately determine, for example, the discrepancy between the behavior the subject should take and the subject's actual behavior.

[0127] Furthermore, by identifying one of the multiple candidate standard behaviors as the standard behavior, and one of the multiple candidate target behaviors as the target behavior, for example, the standard behavior and the target behavior can be appropriately identified, making it possible to appropriately determine the discrepancy between the behavior the target should take and the behavior of the target.

[0128] Furthermore, by performing a determination of whether or not there is a discrepancy between the first risk level (degree of relevance to achieving the first objective) and the second risk level (degree of relevance to achieving the second objective) as a discrepancy-related determination, it becomes possible to make an appropriate determination from the perspective of whether or not there is a discrepancy.

[0129] Furthermore, by determining the degree of discrepancy between the first risk level (degree of relevance to achieving the first objective) and the second risk level (degree of relevance to achieving the second objective) as a discrepancy-related determination, it becomes possible to make appropriate determinations, for example, from the perspective of the degree of discrepancy.

[0130] Furthermore, by determining the relative magnitude of the first risk level (degree of relevance to achieving the first objective) and the second risk level (degree of relevance to achieving the second objective) as a divergence relationship determination, it becomes possible to make appropriate determinations, for example, from the perspective of the relative magnitude.

[0131] Furthermore, by providing notifications regarding the subject based on the determination results of the determination means, it becomes possible to provide useful information, for example, regarding the discrepancy between the actions the subject should take and the actions the subject actually takes.

[0132] [Modifications of the embodiment] While embodiments of the present invention have been described above, the specific configurations and means of the present invention can be arbitrarily modified and improved within the scope of the technical idea of ​​the present invention as described in the claims. Such modifications will be described below.

[0133] (Regarding the problems to be solved and the effects of the invention) First, the problems that the invention aims to solve and the effects of the invention are not limited to those described above, and may vary depending on the implementation environment and details of the invention's configuration. In some cases, only a portion of the problems described above may be solved, or only a portion of the effects described above may be achieved.

[0134] (Regarding decentralization and integration) Furthermore, the electrical components described above are functional concepts and do not necessarily need to be physically configured as shown in the illustrations. In other words, the specific forms of distribution and integration of each part are not limited to those shown in the illustrations, and all or part of them can be functionally or physically distributed or integrated in any unit according to various loads and usage conditions. Also, the term "apparatus" in this application is not limited to those composed of a single apparatus, but includes those composed of multiple apparatuses. Similarly, the term "system" in this application is not limited to those composed of multiple apparatuses, but includes those composed of a single apparatus. In other words, "system" may be interpreted as including "apparatus," and "apparatus" may be interpreted as including "system." In a narrower sense, the concept of "apparatus" may be interpreted as being included in the concept of "system."

[0135] (Regarding shape, numerical values, structure, and time series) With regard to the components illustrated in the embodiments and drawings, their shapes, numerical values, or the interrelationships of the structure or time series of multiple components can be arbitrarily modified and improved within the scope of the technical concept of the present invention.

[0136] (Identification of standard behaviors and behaviors on the part of the target group (Part 1)) In Figure 2 (SA2) and Figure 6 (SB2), we explained the case where one of the three options in Figures 5 and 7—"walking," "walking slowly," and "standing still"—is identified as the baseline behavior and the subject's behavior, but this is not the only case.

[0137] For example, "running" (risk level = "0") may be added, and one of the four may be designated as the standard behavior and the subject's behavior. Alternatively, for example, one of five or more, or two, may be designated as the standard behavior and the subject's behavior. Alternatively, for example, the number of standard behaviors and subject's behaviors to be designated is not limited to one, but two or more may be designated as the standard behavior and the subject's behavior.

[0138] Furthermore, if two or more standard behaviors and two or more subject-side behaviors are identified, the deviation relationship determination in SC1 of Figure 8 may be configured to treat the sum of the risk levels corresponding to the two or more standard behaviors identified in SA2 of Figure 2 as the first risk level, and the sum of the risk levels corresponding to the two or more subject-side behaviors identified in SB2 of Figure 6 as the second risk level.

[0139] Furthermore, the system may be configured to identify specific baseline behaviors and subject behaviors, such as "walking" with a 40% probability and "walking slowly" with an 80% probability, along with the probability of each being likely. In this case, SC1 in Figure 8 may be configured to perform a calculation that multiplies each probability by the corresponding risk level value, and then use the result of this calculation to make a determination.

[0140] Furthermore, for example, SA2 in Figure 2 and SB2 in Figure 6 may be configured to identify the baseline behavior and the subject's behavior regardless of the candidates such as "walking," "walking slowly," and "standing still" in Figures 5 and 7 (i.e., without providing any candidates).

[0141] While the specific implementation method is arbitrary, it may be configured to use a pre-trained large-scale language model (so-called generative AI) implemented in any location, such as on the cloud (i.e., on an external computer), for identification. In other words, it may be configured to send ambient information and predetermined instructions (prompts) to the model as input, receive information indicating a standard action output from the model, and identify the standard action corresponding to the received information. The same applies to the actions of the target person.

[0142] Furthermore, for example, one of the standard behaviors or the subject's behaviors may be identified based on candidates, as described in the embodiment, while the other may be identified regardless of candidates, as described above.

[0143] (Identification of standard behaviors and behaviors on the part of the target group (Part 2)) SA2 in Figure 2 and SB2 in Figure 6 may be omitted.

[0144] In this case, for example, when ambient situation information and behavioral state information are input, a pre-trained model for judgment is generated using a predetermined learning method (such as machine learning with supervised data) and recorded in the recording unit 14. This model outputs a judgment result for a discrepancy relationship judgment, which is a judgment regarding the discrepancy between a standard behavior corresponding to the ambient situation information and the subject's behavior corresponding to the behavioral state information. The judgment result output by this pre-trained model for judgment is arbitrary, but for example, it may be configured to output information showing the information exemplified in the judgment column of Figure 9 as the judgment result, or it may be configured to output information exemplified in the "Calculation Result" column of Figure 9 as the judgment result.

[0145] Furthermore, the control unit 15 may be configured to perform a discrepancy relationship determination in SC1 of Figure 8 by inputting the ambient situation information acquired in SA1 of Figure 2 and the behavioral state information acquired in SB1 of Figure 6 to the trained model for determination recorded in the recording unit 14, and by obtaining the determination result output from the trained model for determination. After this, SC2 and SC3 of Figure 8 may be executed in the same manner as described in the embodiment.

[0146] (Identification of standard behaviors and behaviors of the target individuals (Part 3)) Furthermore, as mentioned above, instead of pre-determining actions and selecting from them, the system may determine suitable actions or states as candidates for standard actions, output actions or states not as words like "stopped" but as numerical values ​​such as a level index representing the degree of danger, output as a combination of the action state and the level index, and configure the system to present multiple patterns of output. Note that the terms "determination" and "output" used herein may be interpreted as referring to "specification."

[0147] Furthermore, the surrounding situation information and behavioral status information may also be configured to include location information (information including information showing routes on a map, etc.). In this case, for example, future actions based on current actions and locations, such as "If I take the rightward route, if I continue on this route..." or "If I reach this point at a predetermined time (for example, in 1 second, etc.)..." may be specified as the reference action and the subject's action.

[0148] (Identification of standard behaviors and behaviors on the part of the target group (Part 4)) Furthermore, as explained in "(Identification of Reference Behaviors and Target Behaviors (Part 1))", if the system is configured to identify two or more reference behaviors and / or target behaviors, it is also possible to configure the system to identify two or more reference behaviors and / or target behaviors along with their recommendation rates. For example, it may be configured to identify "running" as 60%, "walking" as 80%, "walking slowly" as 70%, "standing still" as 50%, etc. Note that "recommendation rate" is a concept that indicates the degree of recommendation to include an behavior in the deviation association assessment; for example, a higher number indicates a greater degree of recommendation.

[0149] In this configuration, for example, one or more standard behaviors and target behaviors may be selected considering the recommendation rate, and each process may be performed based on the selected standard behavior and target behavior. Alternatively, for example, the system may be configured to use the recommendation rate in the deviation association determination without selecting standard behaviors and target behaviors.

[0150] The specific process for determining the recommendation rate may be configured to be executed using, for example, a pre-trained model generated by a predetermined machine learning method, or it may be configured to be executed using other methods (the same applies to other processes).

[0151] (Regarding candidate behavior criteria and candidate behavior) Candidate criterion behavior and candidate subject behavior can be interpreted as concepts that assume any time axis within a time axis that includes the past, present, and future. For example, if a history from the past is assumed, "gradually slowing down, so it looks like it's about to stop," which represents a specific behavioral trajectory leading to the present, may be adopted as the candidate criterion behavior and candidate subject behavior. The implementation method of such concepts is arbitrary, but for example, "gradually slowing down, so it looks like it's about to stop" may be used as candidate criterion behavior information in Figure 5 and candidate subject behavior information in Figure 7.

[0152] (Regarding the arrangement of functions) Furthermore, the various functions described in the above embodiment may be arranged as desired. That is, all or part of each component of the determination system may be configured to be implemented in various devices or systems other than the terminal device 100.

[0153] For example, the determination process shown in Figure 8 in the terminal device 100 may be configured to be performed by a cloud computer that can communicate with the terminal device 100. For example, the cloud computer may communicate with the terminal device 100, receive and acquire the identification results of SA2 and SB2, and perform the determination process shown in Figure 8 based on these identification results. In this case, SC2 may be configured so that the cloud computer notifies the user via the terminal device 100, or it may be configured to send the notification to a terminal device (smartphone, tablet, personal computer, etc.) owned by a person related to the user (e.g., a guardian, etc.). Furthermore, SC3 may be configured so that the cloud computer records in its own record unit, or it may be configured to record in the record unit of an external device (terminal device 100 or another device (various databases, etc.)).

[0154] Furthermore, the configuration of the terminal device 100 in Figure 1 is illustrative, and for example, if a function for inputting sound information is required, it may be configured to include a microphone. In other words, the terminal device 100 in Figure 1 can be modified to include any corresponding components depending on the required function.

[0155] (Regarding the notification (Part 1)) Furthermore, although the above embodiment describes a case in which notifications regarding the target person are made using the speaker 13 in Figure 1, it is not limited to this. For example, the terminal device 100 may be equipped with a display unit (e.g., an indicator light consisting of an LED, various displays, etc.) and configured to make notifications regarding the target person via the display unit. Alternatively, it may be configured to communicate with an external device (e.g., a terminal device owned by a person related to the user (e.g., a guardian, etc.)) and make notifications via the external device.

[0156] (Regarding the notification (Part 2)) Furthermore, although the above embodiment describes the case of adjusting the volume in SC2 of Figure 8, it is not limited to this.

[0157] For example, the system could be configured to output a mild warning message in light wording when the calculation result is "+1" (i.e., to keep the warning light), and to output a serious warning message in important wording when the calculation result is "+2" (i.e., to make the warning sound like an important warning). In this case, the volume of the audio output may be the same for both cases, or the volume for the "+2" case may be louder than for the "+1" case.

[0158] Alternatively, for example, if the calculation result is "-1", a question is output audibly (i.e., the process is kept to a light question), and if the calculation result is "-2", a question is output audibly asking if there are any problems (i.e., the process is carried out to ask if there are any problems). In this case, the volume of the audio output may be the same for both cases, or the volume may be louder when the calculation result is "-2" than when it is "-1".

[0159] (Regarding the degree of relevance to achieving the objective) Furthermore, in the above embodiment, the case where the "degree of relevance to achieving the objective" is the degree of danger related to the user's (elementary school student's) achievement of the objective of avoiding danger (for example, avoiding dangers related to traffic), was explained, but it is not limited to this. For example, the "degree of relevance to achieving the objective" may be the degree of ease of progress (i.e., the degree of ease of progress) related to the user's (person's) achievement of the objective of smooth progress, or it may be any other degree. As the degree of ease of progress, five levels may be set: "no obstacles," "sparse pedestrians," "crowded but moving with the flow," "crowded and going against the flow," and "obstacles present, unable to move," and the system may be configured to specify the standard action and the action taken by the target person corresponding to each level.

[0160] (Regarding the determination of discrepancies) Furthermore, the deviation relationship determination of SC1 in Figure 8 may be configured to perform processing other than the processing described in the embodiment. For example, it may be configured to perform determination by scoring using a function that uses the results of the two specific means (i.e., the specific results of SA2 in Figure 2 and SB2 in Figure 6), or it may be configured to perform determination by comparison results that also take into account changes in the results of the two specific means (i.e., the specific results of SA2 in Figure 2 and SB2 in Figure 6). Alternatively, it may be configured to perform only one or more of the determinations exemplified as deviation relationship determinations. In other words, for example, "determination by scoring using a function that uses the results of the two specific means (specific results of the first specific means and the second specific means)" or "determination by comparison results that also take into account changes in the two specific means (comparison results that also take into account changes in the specific results of the first specific means and the second specific means)" may be configured as deviation relationship determinations.

[0161] (Regarding various processes) Furthermore, known pre-processing or post-processing may be applied to the various processes described in the above embodiments. For example, the process of SA2 in Figure 2 may be configured to perform known pre-processing (for example, processing to convert a fisheye image to a cylindrical image, or optical flow processing on an image) on the ambient information acquired by SA1, and then process using the information generated by the pre-processing. In this case, the first trained model may be configured to correspond to the said pre-processing.

[0162] (Regarding combinations) Furthermore, the features of the above embodiment and the features of the modified example may be combined in any way.

[0163] (Note) The determination system described in Appendix 1 comprises: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is the action the subject should take in response to the surrounding circumstances of the subject, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information to identify the state of the subject's actions; a second identification means for identifying a subject-side action, which is the action of the subject, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination regarding the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means.

[0164] The determination system in Appendix 2, in the determination system described in Appendix 1, wherein the first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of relevance to achieving the objective of the target person as the standard action, the second identification means identifies one of a plurality of candidate target person actions predetermined based on the degree of relevance to achieving the objective as the target person action, and the determination means performs the deviation relevance determination based on the degree of relevance to achieving the objective.

[0165] The determination system in Appendix 3, in the determination system described in Appendix 2, includes a determination by the determination means of whether there is a discrepancy between the first degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the standard behavior identified by the first identification means, and the second degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the target person's behavior identified by the second identification means, as the discrepancy relevance determination.

[0166] The determination system in Appendix 4, in the determination system described in Appendix 2, includes a determination by the determination means of determining the degree of discrepancy between the first degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the standard behavior identified by the first identification means, and the second degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the target person's behavior identified by the second identification means, as the discrepancy relevance determination.

[0167] The determination system in Appendix 5, in the determination system described in Appendix 2, determines the relative magnitude of the degrees of relevance to achieving the objective between the first degree of relevance to achieving the objective and the second degree of relevance to achieving the objective when the first degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the standard action identified by the first identification means, and the second degree of relevance to achieving the objective, which is the degree of relevance to achieving the objective corresponding to the target person action identified by the second identification means, as the deviance relationship determination.

[0168] The determination program described in Appendix 6 causes the computer to function as: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is the action the subject should take in response to the surrounding circumstances of the subject, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information to identify the state of the subject's actions; a second identification means for identifying the subject's actions, which is the actions of the subject, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a discrepancy-related determination, which is a determination regarding the discrepancy between the standard action identified by the first identification means and the subject's actions identified by the second identification means.

[0169] The determination device described in Appendix 7 comprises: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a first identification means for identifying a standard action, which is an action that the subject should take in response to the surrounding circumstances of the subject, based on the ambient information acquired by the first acquisition means; a second acquisition means for acquiring behavioral state information to identify the state of the subject's actions; a second identification means for identifying a subject-side action, which is an action taken by the subject, based on the behavioral state information acquired by the second acquisition means; and a determination means for performing a deviation-related determination, which is a determination regarding the deviation between the standard action identified by the first identification means and the subject-side action identified by the second identification means.

[0170] The determination method of Appendix 8 includes: a first acquisition step of acquiring ambient information indicating the surrounding circumstances of a subject; a first identification step of identifying a standard action, which is the action the subject should take in response to the surrounding circumstances of the subject, based on the ambient information acquired in the first acquisition step; a second acquisition step of acquiring behavioral state information to identify the state of the subject's actions; a second identification step of identifying a subject-side action, which is the action of the subject, based on the behavioral state information acquired in the second acquisition step; and a determination step of performing a discrepancy-related determination, which is a determination of the discrepancy between the standard action identified in the first identification step and the subject-side action identified in the second identification step.

[0171] The determination system described in Appendix 9 comprises: a first acquisition means for acquiring ambient information indicating the surrounding circumstances of a subject; a second acquisition means for acquiring behavioral state information indicating the state of the subject's actions; and a determination means for performing a discrepancy-related determination, which is a determination regarding the discrepancy between a standard action, which is the action the subject should take in response to the surrounding circumstances of the subject, and a subject-side action, which is the action the subject takes, based on the ambient information acquired by the first acquisition means and the behavioral state information acquired by the second acquisition means.

[0172] The determination system of Appendix 10 further comprises a notification means that provides notification regarding the subject based on the determination result of the determination means, in the determination system described in any one of Appendix 1 to 5 or 9.

[0173] (Effect of the note) According to the judgment system described in Appendix 1, the judgment program described in Appendix 6, the judgment device described in Appendix 7, the judgment method described in Appendix 8, and the judgment system described in Appendix 9, by performing a discrepancy-related judgment, which is a judgment regarding the discrepancy between the standard behavior and the subject's behavior, it becomes possible to appropriately determine, for example, the discrepancy between the behavior the subject should take and the subject's behavior.

[0174] According to the judgment system described in Appendix 2, by identifying one of the candidate standard behaviors as the standard behavior and one of the candidate target behaviors as the target behavior, it is possible to appropriately identify the standard behavior and the target behavior, for example, and thus make it possible to appropriately judge the discrepancy between the behavior the target should take and the behavior of the target.

[0175] According to the judgment system described in Appendix 3, by performing a judgment on whether or not there is a discrepancy between the degree of relevance to achieving the first objective and the degree of relevance to achieving the second objective as a discrepancy-related judgment, it becomes possible to make an appropriate judgment, for example, from the perspective of whether or not there is a discrepancy.

[0176] According to the determination system described in Appendix 4, by determining the degree of discrepancy between the degree of relevance to achieving the first objective and the degree of relevance to achieving the second objective as a discrepancy-related determination, it becomes possible to make an appropriate determination from the perspective of the degree of discrepancy, for example.

[0177] According to the determination system described in Appendix 5, the relative magnitudes of the degree of relevance to achieving the first objective and the degree of relevance to achieving the second objective are determined as a divergence relationship determination, making it possible to make appropriate determinations, for example, from the perspective of the relative magnitudes.

[0178] According to the determination system described in Appendix 10, by providing notification regarding the subject based on the determination result of the determination means, it becomes possible to provide useful information, for example, regarding the discrepancy between the action the subject should take and the action the subject is actually taking. [Explanation of symbols]

[0179] 11 Cameras 12 sensors 13 speakers 14 Records Section 15 Control Unit 100 terminal devices

Claims

1. A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The system comprises a determination means that performs a deviation-related determination, which is a determination of the deviation between the standard behavior identified by the first identification means and the target person's behavior identified by the second identification means, The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of whether or not there is a discrepancy between the first degree of risk, which is the degree of risk corresponding to the standard action identified by the first identification means, and the second degree of risk, which is the pre-risk degree corresponding to the target person's action identified by the second identification means, as the discrepancy-related determination. Judgment system.

2. A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The system comprises a determination means that performs a deviation-related determination, which is a determination of the deviation between the standard behavior identified by the first identification means and the target person's behavior identified by the second identification means, The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of the degree of discrepancy between the first degree of risk, which is the degree of risk corresponding to the standard behavior identified by the first identification means, and the second degree of risk, which is the degree of risk corresponding to the target person's behavior identified by the second identification means, as the discrepancy-related determination. Judgment system.

3. A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The system comprises a determination means that performs a deviation-related determination, which is a determination of the deviation between the standard behavior identified by the first identification means and the target person's behavior identified by the second identification means, The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of the relative magnitude of the risk levels between the first risk level and the second risk level when there is a discrepancy between the first risk level, which is the risk level corresponding to the standard behavior identified by the first identification means, and the second risk level, which is the risk level corresponding to the target person's behavior identified by the second identification means, as the discrepancy relationship determination. Judgment system.

4. Computers, A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The first identifying means functions as a determination means that performs a discrepancy-related determination, which is a determination regarding the discrepancy between the standard behavior identified by the first identifying means and the target person's behavior identified by the second identifying means. The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of whether or not there is a discrepancy between the first degree of risk, which is the degree of risk corresponding to the standard behavior identified by the first identification means, and the second degree of risk, which is the degree of risk corresponding to the target person's behavior identified by the second identification means, as the discrepancy-related determination. Judgment program.

5. Computers, A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The first identifying means functions as a determination means that performs a discrepancy-related determination, which is a determination regarding the discrepancy between the standard behavior identified by the first identifying means and the target person's behavior identified by the second identifying means. The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of the degree of discrepancy between the first degree of risk, which is the degree of risk corresponding to the standard behavior identified by the first identification means, and the second degree of risk, which is the degree of risk corresponding to the target person's behavior identified by the second identification means, as the discrepancy-related determination. Judgment program.

6. Computers, A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The first identifying means functions as a determination means that performs a discrepancy-related determination, which is a determination regarding the discrepancy between the standard behavior identified by the first identifying means and the target person's behavior identified by the second identifying means. The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of the relative magnitude of the risk levels between the first risk level and the second risk level when there is a discrepancy between the first risk level, which is the risk level corresponding to the standard behavior identified by the first identification means, and the second risk level, which is the risk level corresponding to the target person's behavior identified by the second identification means, as the discrepancy relationship determination. Judgment program.

7. A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The system comprises a determination means that performs a deviation-related determination, which is a determination of the deviation between the standard behavior identified by the first identification means and the target person's behavior identified by the second identification means, The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of whether or not there is a discrepancy between the first degree of risk, which is the degree of risk corresponding to the standard behavior identified by the first identification means, and the second degree of risk, which is the degree of risk corresponding to the target person's behavior identified by the second identification means, as the discrepancy-related determination. Judgment device.

8. A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The system comprises a determination means that performs a deviation-related determination, which is a determination of the deviation between the standard behavior identified by the first identification means and the target person's behavior identified by the second identification means, The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of the degree of discrepancy between the first degree of risk, which is the degree of risk corresponding to the standard behavior identified by the first identification means, and the second degree of risk, which is the degree of risk corresponding to the target person's behavior identified by the second identification means, as the discrepancy-related determination. Judgment device.

9. A first acquisition means for acquiring surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired by the first acquisition means, a first identification means identifies a standard action that the subject should take in response to the surrounding situation of the subject, which is related to movement. A second acquisition means for acquiring behavioral state information to identify the state of the subject's behavior, A second identification means identifies, based on the behavioral state information acquired by the second acquisition means, the behavior of the subject, which is a subject-side behavior related to movement, The system comprises a determination means that performs a deviation-related determination, which is a determination of the deviation between the standard behavior identified by the first identification means and the target person's behavior identified by the second identification means, The first identification means identifies one of a plurality of candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, as the standard action. The second identification means identifies one of the candidate target behaviors predetermined based on the degree of risk as the target behavior, The determination means performs the deviation-related determination based on the degree of risk, The determination means performs the determination of the relative magnitude of the risk levels between the first risk level and the second risk level when there is a discrepancy between the first risk level, which is the risk level corresponding to the standard behavior identified by the first identification means, and the second risk level, which is the risk level corresponding to the target person's behavior identified by the second identification means, as the discrepancy relationship determination. Judgment device.

10. The first acquisition step involves obtaining surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired in the first acquisition step, a first identification step is to identify a standard action that corresponds to the surrounding situation of the subject and is related to movement, which is a standard action that the subject should take. A second acquisition step involves acquiring behavioral status information to identify the behavioral state of the subject, A second identification step involves identifying, based on the behavioral state information acquired in the second acquisition step, the subject's actions, which are subject-side actions related to movement, and The process includes a determination step that performs a discrepancy-related determination, which is a determination of the discrepancy between the standard behavior identified in the first identification step and the subject behavior identified in the second identification step, In the first identification step, one of several candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, is identified as the standard action. In the second identification step, one of the candidate subject behaviors predetermined based on the degree of risk is identified as the subject behavior. In the aforementioned determination step, the aforementioned deviation-related determination is made based on the degree of risk, In the determination step, the determination of whether or not there is a discrepancy between the first degree of risk, which is the degree of risk corresponding to the standard behavior identified in the first identification step, and the second degree of risk, which is the degree of risk corresponding to the subject's behavior identified in the second identification step, is performed as the discrepancy-related determination. Judgment method.

11. The first acquisition step involves obtaining surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired in the first acquisition step, a first identification step is to identify a standard action that corresponds to the surrounding situation of the subject and is related to movement, which is a standard action that the subject should take. A second acquisition step involves acquiring behavioral status information to identify the behavioral state of the subject, A second identification step involves identifying, based on the behavioral state information acquired in the second acquisition step, the subject's actions, which are subject-side actions related to movement, and The process includes a determination step that performs a discrepancy-related determination, which is a determination of the discrepancy between the standard behavior identified in the first identification step and the subject behavior identified in the second identification step, In the first identification step, one of several candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, is identified as the standard action. In the second identification step, one of the candidate subject behaviors predetermined based on the degree of risk is identified as the subject behavior. In the aforementioned determination step, the aforementioned deviation-related determination is made based on the degree of risk, In the determination step, the degree of discrepancy between the first risk level, which is the risk level corresponding to the standard behavior identified in the first identification step, and the second risk level, which is the risk level corresponding to the subject's behavior identified in the second identification step, is determined as the discrepancy-related determination. Judgment method.

12. The first acquisition step involves obtaining surrounding situation information that shows the circumstances around the subject, Based on the surrounding situation information acquired in the first acquisition step, a first identification step is to identify a standard action that corresponds to the surrounding situation of the subject and is related to movement, which is a standard action that the subject should take. A second acquisition step involves acquiring behavioral status information to identify the behavioral state of the subject, A second identification step involves identifying, based on the behavioral state information acquired in the second acquisition step, the subject's actions, which are subject-side actions related to movement, and The process includes a determination step that performs a discrepancy-related determination, which is a determination of the discrepancy between the standard behavior identified in the first identification step and the subject behavior identified in the second identification step, In the first identification step, one of several candidate standard actions predetermined based on the degree of risk related to achieving the objective of avoiding danger for the subject, which indicates the degree of risk that the subject is likely to be involved in a traffic accident, is identified as the standard action. In the second identification step, one of the candidate subject behaviors predetermined based on the degree of risk is identified as the subject behavior. In the aforementioned determination step, the aforementioned deviation-related determination is made based on the degree of risk, In the determination step, when there is a discrepancy between the first risk level, which is the risk level corresponding to the standard behavior identified in the first identification step, and the second risk level, which is the risk level corresponding to the subject's behavior identified in the second identification step, the relative magnitude of the risk levels between the first and second risk levels is determined as the discrepancy relationship determination. Judgment method.

13. The system further comprises a notification means for providing notification regarding the subject based on the determination result of the determination means. A determination system according to any one of claims 1 to 3.

Images