Information processing apparatus, information processing method, and computer-readable medium

Abstract

Provided is an information processing apparatus capable of appropriately ascertaining the vehicle model of a self-driving vehicle that may exhibit abnormal movement. The information processing apparatus acquires a captured image of at least one point through which a vehicle can travel. The information processing apparatus identifies an attribute that pertains to the autonomous driving of the vehicle in the image. The information processing apparatus acquires, for each of the attributes of at least two vehicles, travel data that indicates vehicle behavior at the point. The information processing apparatus uses a prescribed determination criterion to determine whether the degree of deviation, from the total travel data, of the travel data pertaining to each of the attributes that correspond to the vehicle models of the self-driving vehicles is high. The information processing apparatus identifies, as a first vehicle model, the vehicle model of a self-driving vehicle having an attribute for which the degree of deviation is determined to be high. The information processing apparatus performs a process for presenting information that pertains to the identified first vehicle model.

Description

Information Processing Apparatus, Information Processing Method, and Computer-Readable Medium

[0001] The present disclosure relates to an information processing apparatus, an information processing method, and a computer-readable medium.

[0002] Patent Document 1 discloses a road management system. The system according to Patent Document 1 captures an image of a vehicle on a road to generate image data. The system according to Patent Document 1 recognizes whether the vehicle is an autonomous vehicle, and when the recognition result indicates an autonomous vehicle, generates feature quantity information of the autonomous vehicle based on the image data. The system according to Patent Document 1 provides the feature quantity information and information regarding the position of the autonomous vehicle.

[0003] Japanese Patent Application Laid-Open No. 2017-084260

[0004] In the technology according to Patent Document 1, only the appearance features of the autonomous vehicle are used to manage the autonomous vehicle. Therefore, with the technology according to Patent Document 1, there is a possibility that the vehicle type of an autonomous vehicle that may perform an abnormal operation cannot be specified. Therefore, it is desirable to appropriately grasp the vehicle type of an autonomous vehicle that may perform an abnormal operation.

[0005] An object of the present disclosure is to solve such problems, and to provide an information processing apparatus, an information processing method, and a program capable of appropriately grasping the vehicle type of an autonomous vehicle that may perform an abnormal operation.

[0006] The information processing device according to this disclosure includes: an image acquisition means for acquiring images obtained by photographing one or more locations where a vehicle can travel; an attribute identification means for identifying attributes related to the autonomous driving of a vehicle shown in the image; a driving data acquisition means for acquiring driving data indicating the behavior of the vehicle at the location for each of at least two vehicle attributes; a deviation degree determination means for determining whether there is a large deviation between overall driving data, which is driving data for vehicles of at least two attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle, using predetermined determination criteria; a vehicle type identification means for identifying the type of autonomous vehicle for the attribute determined to have a large deviation as a first vehicle type; and a presentation means for performing processing for presenting information about the identified first vehicle type.

[0007] The information processing method according to this disclosure acquires images obtained by photographing one or more locations where a vehicle can travel, identifies attributes related to autonomous driving of the vehicles shown in the images, acquires driving data indicating the behavior of the vehicles at the locations for each of at least two vehicle attributes, determines whether there is a large discrepancy between the overall driving data, which is the driving data for at least two vehicle attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle, using predetermined criteria, identifies the type of autonomous vehicle for the attribute determined to have a large discrepancy as the first type of vehicle, and performs processing to present information regarding the identified first type of vehicle.

[0008] The program according to this disclosure causes a computer to perform the following steps: acquire images obtained by photographing one or more locations where a vehicle can travel; identify attributes related to autonomous driving of the vehicle shown in the images; acquire driving data indicating the behavior of the vehicle at the location for each of at least two vehicle attributes; determine, using predetermined criteria, whether there is a large discrepancy between the overall driving data, which is the driving data for at least two vehicle attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle; identify the type of autonomous vehicle for the attribute determined to have a large discrepancy as a first type of vehicle; and perform processing to present information regarding the identified first type of vehicle.

[0009] According to this disclosure, it is possible to provide an information processing device, an information processing method, and a program that can appropriately identify the types of autonomous vehicles that may be exhibiting abnormal behavior.

[0010] This is a diagram showing the configuration of the information processing device related to this disclosure. This is a flowchart showing the information processing method performed by the information processing device related to this disclosure. This is a diagram illustrating the configuration of the information presentation system related to this disclosure. This is a diagram illustrating the hardware configuration of the information processing device related to this disclosure. This is a diagram illustrating the functional configuration of the information processing device related to this disclosure. This is a diagram for explaining the process of acquiring driving data by the information processing device related to this disclosure. This is a diagram illustrating behavioral data acquired by the driving data acquisition unit related to this disclosure. This is a diagram illustrating driving data acquired by the driving data acquisition process of the information processing device related to this disclosure. This is a diagram illustrating driving data acquired by the driving data acquisition process of the information processing device related to this disclosure. This is a diagram for explaining the process of the deviation degree determination unit related to this disclosure. This is a flowchart showing an example of a process performed by the information processing device related to this disclosure. This is a diagram illustrating the functional configuration of the information processing device related to this disclosure. This is a flowchart showing an example of a process performed by the information processing device related to this disclosure. This is a diagram illustrating the functional configuration of the information processing device related to this disclosure. This is a diagram for explaining the process of the deviation degree determination unit related to this disclosure. This is a flowchart showing an example of a process performed by the information processing device related to this disclosure.

[0011] The embodiments will be described below with reference to the drawings. For clarity of explanation, the following descriptions and drawings have been omitted and simplified as appropriate. In addition, the same elements are denoted by the same reference numerals in each drawing, and redundant explanations have been omitted where necessary. Furthermore, although the following explanation will use drawings, the drawings used in the description of a particular embodiment are not limited to that embodiment. Each drawing may be applicable to all embodiments.

[0012] (Embodiment 1) Figure 1 is a diagram showing the configuration of an information processing device 1 according to the present disclosure. The information processing device 1 includes an image acquisition unit 2, an attribute identification unit 4, a driving data acquisition unit 6, a deviation degree determination unit 8, a vehicle type identification unit 10, and a presentation unit 12. The image acquisition unit 2 functions as an image acquisition means. The attribute identification unit 4 functions as an attribute identification means. The driving data acquisition unit 6 functions as a driving data acquisition means. The deviation degree determination unit 8 functions as a deviation degree determination means. The vehicle type identification unit 10 functions as a vehicle type identification means. The presentation unit 12 functions as a presentation means. The information processing device 1 uses these components to present vehicle types of autonomous vehicles that may exhibit abnormal behavior.

[0013] The information processing device 1 can be implemented by a computer. The information processing device 1 can be implemented, for example, by cloud computing. Alternatively, the information processing device 1 can be implemented, for example, by multiple computers connected to each other in a communicative manner. Furthermore, each component of the information processing device 1 may be implemented by being distributed across multiple computers. In other words, the computers implementing each of the above-mentioned components may be physically separate. Also, the functions of each of the multiple components may be implemented by multiple computers.

[0014] Figure 2 is a flowchart illustrating an information processing method performed by the information processing device 1 according to this disclosure. It can also be said that Figure 2 shows an information presentation method performed by the information processing device 1. The image acquisition unit 2 acquires images obtained by photographing one or more points on which a vehicle can travel (step S12). Here, "points on which a vehicle can travel" are predetermined locations on a road that can be photographed by the imaging device. These locations are, for example, intersections, but are not limited to these. These locations may be parking areas or any points on a road.

[0015] Furthermore, an "image" is obtained by at least one imaging device capturing a location. The image may be a moving image (video) or a still image. The image may also be a frame image constituting a video. In the following, the term "image" also refers to "image data representing an image" as the object of processing in information processing.

[0016] The attribute identification unit 4 identifies the attributes related to the autonomous driving of the vehicle shown in the image (step S14). Here, the "attributes related to the autonomous driving of the vehicle" are defined, for example, whether the vehicle is an autonomous vehicle or not, and if the vehicle is an autonomous vehicle, its make and model. For example, the "attributes" may be "manual vehicle," "autonomous vehicle model A," and "autonomous vehicle model B." An autonomous vehicle corresponds to a vehicle that can perform driving operations automatically without driver intervention. On the other hand, a manual vehicle is a vehicle that operates through manual driving. In other words, a manual vehicle corresponds to a vehicle that performs driving operations through driver intervention. To put it another way, a manual vehicle corresponds to a vehicle that is not an autonomous vehicle as described above. The level of autonomous driving in an "autonomous vehicle" may be any level except for a level in which no autonomous driving technology is installed. Also, the "autonomous vehicle model" (the make and model of the autonomous vehicle) corresponds to the specifications related to autonomous driving. That is, if vehicles have the same specifications related to autonomous driving, their autonomous vehicle models may be the same. Autonomous vehicle types may be determined, for example, according to the specifications of the autonomous driving algorithm and sensors. Alternatively, autonomous vehicle types may be determined, for example, by the vehicle manufacturer and the name assigned by the vehicle manufacturer. Furthermore, "attributes" may, but are not limited to, those that can be identified using the image acquired in processing S12. Specific examples of methods for identifying attributes will be described later.

[0017] The driving data acquisition unit 6 acquires driving data (step S16). Specifically, the driving data acquisition unit 6 acquires driving data indicating the behavior of the vehicle at a given location for each of at least two vehicle attributes. The driving data acquisition unit 6 may also acquire driving data for each type of behavior. Here, "vehicle behavior" corresponds to the operations performed when the vehicle is driving. The types of behavior may include, but are not limited to, at least one of the following behaviors: behavior related to stopping, behavior related to turning, behavior related to driving straight, behavior related to changing lanes, and behavior related to starting.

[0018] Furthermore, the "driving data" may show the frequency distribution of each vehicle behavior. Specifically, the driving data may show the frequency distribution of values ​​for driving indicators that represent each vehicle behavior of each attribute at each location. The driving indicators correspond to indicators for evaluating the driving content for each behavior. The driving indicators correspond to the vehicle's actions performed by the driving operations when the vehicle performs each behavior. The driving indicators may include, for example, the driving speed. Also, for example, if the type of behavior is "stopping," the driving indicators may include the stopping position. The stopping position may include, for example, the distance from the stopping position of the vehicle to the stop line. Details of the driving data will be described later. The frequency distribution may be shown, for example, as a histogram.

[0019] The deviation degree determination unit 8 determines the degree of deviation between the overall driving data and the driving data for each attribute corresponding to the type of autonomous vehicle (step S18). Specifically, the deviation degree determination unit 8 determines whether the degree of deviation between the overall driving data and the driving data for each attribute corresponding to the type of autonomous vehicle is large or small, using predetermined criteria. Here, "overall driving data" refers to driving data for the attributes of at least two vehicles, but may also refer to driving data for the attributes of all vehicles. "Degree of deviation" corresponds to the difference between "overall driving data" and "driving data for each attribute corresponding to the type of autonomous vehicle." The larger the "degree of deviation," the greater the discrepancy between "driving data for each attribute corresponding to the type of autonomous vehicle" and "overall driving data." "Criteria" corresponds to an index that can evaluate the degree of deviation between "overall driving data" and "driving data for each attribute corresponding to the type of autonomous vehicle."

[0020] Furthermore, the driving data may show the frequency distribution of each behavior of the vehicle. In this case, the deviation degree determination unit 8 determines whether the degree of deviation is large for each behavior between the frequency distribution of that behavior shown in the overall driving data and the frequency distribution of that behavior shown in the driving data for each attribute of the autonomous vehicle. In this case, the determination criterion may be, for example, the difference between the modes of the two distributions, or the difference between the mean values ​​of the two distributions. In this case, the deviation degree determination unit 8 may determine that the degree of deviation is large if the difference between the modes is greater than or equal to a predetermined threshold. Alternatively, the deviation degree determination unit 8 may determine that the degree of deviation is large if the difference between the mean values ​​is greater than or equal to a predetermined threshold.

[0021] The vehicle type identification unit 10 identifies a first vehicle type (step S20). Specifically, the vehicle type identification unit 10 identifies the vehicle type of the autonomous vehicle whose attribute is determined to have a large degree of deviation as the first vehicle type. Here, an autonomous vehicle whose driving data has a large degree of deviation from the overall driving data is a vehicle that behaves differently compared to vehicles with multiple attributes related to the overall driving data. Therefore, an autonomous vehicle whose attribute is determined to have a large degree of deviation is a vehicle that may behave abnormally. Thus, the "first vehicle type" can be defined as the vehicle type of an autonomous vehicle that may behave abnormally, in other words, an abnormal vehicle type.

[0022] The presentation unit 12 performs processing to present information about the identified first vehicle type (step S22). For example, the presentation unit 12 may transmit information about the first vehicle type to a terminal of a destination facility or organization so that the first vehicle type is displayed on that terminal. The information about the first vehicle type may include, for example, identification information of the first vehicle type and behavior regarding driving data that has been determined to have a large degree of deviation for the first vehicle type. The destination may be, for example, a facility or organization that can correct the specifications related to the autonomous driving of the first vehicle type's autonomous vehicle.

[0023] As described above, the information processing device 1 in this disclosure presents information about a first vehicle model that may be exhibiting abnormal behavior. Because the information processing device 1 in this disclosure is configured in this way, the user to whom the information is presented can appropriately identify the vehicle model of an autonomous vehicle that may be exhibiting abnormal behavior. Therefore, it becomes possible to improve autonomous driving technology by appropriately correcting the autonomous driving specifications of the autonomous vehicle model that may be exhibiting abnormal behavior.

[0024] Furthermore, the information processing method executed by the information processing device 1 also makes it possible to appropriately identify the types of autonomous vehicles that may be exhibiting abnormal behavior. In addition, the program that executes the information processing method also makes it possible to appropriately identify the types of autonomous vehicles that may be exhibiting abnormal behavior.

[0025] (Embodiment 2) Next, Embodiment 2 will be described with reference to the drawings. For clarity of explanation, the following descriptions and drawings have been omitted and simplified as appropriate. In addition, the same elements are denoted by the same reference numerals in each drawing, and redundant explanations have been omitted as necessary. The contents of Embodiment 2 correspond to the details of the contents of Embodiment 1.

[0026] Figure 3 is a diagram illustrating the configuration of the information presentation system 20 according to this disclosure. The information presentation system 20 includes one or more roadside imaging devices 30, one or more terminal devices 40, and an information processing device 100. The roadside imaging devices 30, terminal devices 40, and information processing device 100 have functions such as computers.

[0027] The information processing device 100 corresponds to the information processing device 1 shown in Figure 1. The information processing device 100 may be, for example, a server. Alternatively, the information processing device 100 may be, for example, an information terminal. Furthermore, the information processing device 100 may be implemented by, for example, cloud computing. Also, the information processing device 100 may be implemented by, for example, multiple computers that are connected to each other in a communicative manner. In addition, each component of the information processing device 100, which will be described later, may be implemented by multiple computers. As will be described later, the information processing device 100 identifies a vehicle with abnormal operation (a first vehicle) that may be performing abnormal operations and presents information about this vehicle with abnormal operation, which is information about the vehicle with abnormal operation.

[0028] The roadside imaging device 30 and the information processing device 100 are connected to each other via a wired or wireless network for communication. The information processing device 100 and the terminal device 40 are also connected to each other via a wired or wireless network for communication. The wireless network may be, for example, a network using a communication line standard such as LTE (Long Term Evolution), or a network used in a specific area such as Wi-Fi (registered trademark) or local 5G. The wired network may be, for example, a LAN (Local Area Network) or optical fiber.

[0029] The roadside imaging device 30 is, for example, a camera, but is not limited to that. The roadside imaging device 30 is installed so as to be able to photograph the aforementioned locations, that is, locations where vehicles can travel. For example, the roadside imaging device 30 may be installed at an intersection. In this case, the roadside imaging device 30 is positioned so that the intersection is included in the shooting range. Note that roadside imaging devices 30 may be installed at each of multiple locations. Also, multiple roadside imaging devices 30 may be installed at a single location. Furthermore, if the roadside imaging device 30 is mounted on an aircraft such as a drone or artificial satellite, one roadside imaging device 30 may photograph multiple locations. The roadside imaging device 30 transmits the road image, which is an image obtained by photographing the installed location, to the information processing device 100.

[0030] Terminal device 40 is a terminal of the organization to which the information processing device 100 presents abnormal vehicle information. Terminal device 40 may be installed in a designated facility, which is a facility of the designated organization. Terminal device 40 receives abnormal vehicle information from the information processing device 100. Terminal device 40 outputs the received abnormal vehicle information to the user. The designated facility is, for example, a vehicle manufacturer that produces autonomous vehicles related to abnormal vehicle types, but is not limited to this. The designated organization may be a police organization or an organization that has jurisdiction over vehicle safety. At the designated facility, corrections may be made to the specifications (algorithms or sensors, etc.) of autonomous vehicles related to abnormal vehicle types.

[0031] Figure 4 is a diagram illustrating the hardware configuration of the information processing device 100 according to this disclosure. As shown in Figure 4, the information processing device 100 has as its main hardware components a control unit 102, a storage unit 104, a communication unit 106, and an interface unit 108 (IF). The control unit 102, storage unit 104, communication unit 106, and interface unit 108 are interconnected via a data bus or the like. The roadside imaging device 30 and terminal device 40 described above may also have the hardware configuration of the information processing device 100 shown in Figure 4.

[0032] The control unit 102 is a processor, such as a CPU (Central Processing Unit). The control unit 102 has the function of an arithmetic unit that performs control processing and arithmetic processing. The control unit 102 may have multiple processors. The storage unit 104 is a storage device, such as a memory or a hard disk. The storage unit 104 is, for example, a ROM (Read Only Memory) or RAM (Random Access Memory). The storage unit 104 has the function of storing control programs and arithmetic programs executed by the control unit 102. In other words, the storage unit 104, which is a memory, stores one or more instructions. The storage unit 104 also has the function of temporarily storing processing data. The storage unit 104 may include a database. The storage unit 104 may also have multiple memories.

[0033] The communication unit 106 performs the processing necessary for the information processing device 100 to communicate with other devices via a network. The communication unit 106 may include a communication port, router, firewall, etc. The interface unit 108 (IF; Interface) is, for example, a user interface (UI). The interface unit 108 has an input device such as a keyboard, touch panel, or mouse, and an output device such as a display or speaker. The interface unit 108 may be configured such that the input device and the output device are integrated, for example, a touchscreen or touch panel. The interface unit 108 accepts data input operations from a user such as an operator or worker and outputs information to the user. The interface unit 108 may output abnormal vehicle information.

[0034] Figure 5 is a diagram illustrating the functional configuration of the information processing device 100 according to this disclosure. The information processing device 100 includes, as its components, an image acquisition unit 120, a vehicle attribute identification unit 122, a trained model storage unit 124, a driving data acquisition unit 130, and a driving data storage unit 132. The information processing device 100 also includes, as its components, a deviation degree determination unit 140, an abnormal vehicle type identification unit 150, and a vehicle type information presentation unit 170. Using these components, the information processing device 100 identifies an abnormal vehicle type (a first vehicle type) and presents abnormal vehicle type information relating to this abnormal vehicle type to the destination facility.

[0035] As mentioned above, the information processing device 100 does not need to be composed of a single physical device. In this case, each of the above-mentioned components may be realized by multiple physically separate devices. This is also true in other embodiments described later.

[0036] The image acquisition unit 120 functions as an image acquisition means. The vehicle attribute identification unit 122 functions as a vehicle attribute identification means (attribute identification means). The trained model storage unit 124 functions as a trained model storage means. The driving data acquisition unit 130 functions as a driving data acquisition means. The driving data storage unit 132 functions as a driving data storage means. The deviation degree determination unit 140 functions as a deviation degree determination means. The abnormal vehicle type identification unit 150 functions as an abnormal vehicle type identification means (vehicle type identification means). The vehicle type information presentation unit 170 functions as a vehicle information presentation means (presentation means).

[0037] Each of the above-described components can be realized, for example, by executing a program under the control of the control unit 102. More specifically, each component can be realized by the control unit 102 executing a program (instruction) stored in the memory unit 104. Alternatively, each component can be realized by recording the necessary program on any non-volatile recording medium and installing it as needed. Furthermore, each component is not limited to being realized by software programs, but may also be realized by any combination of hardware, firmware, and software. In addition, each component may be realized using a user-programmable integrated circuit, such as an FPGA (field-programmable gate array) or a microcontroller. In this case, the program composed of the above-described components may be realized using this integrated circuit. These points are also true in other embodiments described later.

[0038] The image acquisition unit 120 corresponds to the image acquisition unit 2 shown in Figure 1. The image acquisition unit 120 acquires road images obtained by photographing one or more points where a vehicle can travel. Specifically, the image acquisition unit 120 acquires road images obtained by the roadside imaging device 30 from the roadside imaging device 30. The image acquisition unit 120 may also acquire road images by having the communication unit 106 receive road images from the roadside imaging device 30.

[0039] The image acquisition unit 120 can acquire road images from each of the multiple roadside imaging devices 30 installed at multiple locations. In this case, the road image is accompanied by location information indicating which location was captured. For example, the road image may be accompanied by identification information of the roadside imaging device 30 that captured the road image as location information. Alternatively, the road image may be accompanied by identification information or location information of the location where the roadside imaging device 30 that captured the road image is installed as location information.

[0040] The vehicle attribute identification unit 122 corresponds to the attribute identification unit 4 shown in Figure 1. The vehicle attribute identification unit 122 identifies attributes related to autonomous driving of a vehicle that appears in the road image. Specifically, the vehicle attribute identification unit 122 may identify attributes related to autonomous driving of a vehicle from the shape of the vehicle that appears in the road image. The vehicle attribute identification unit 122 may identify the vehicle attribute as "autonomous vehicle" if the shape of the vehicle that appears in the road image is a shape characteristic of an autonomous vehicle. For example, the vehicle attribute identification unit 122 may identify the vehicle attribute as "autonomous vehicle" if a sensor such as a LiDAR (Light Detection and Ranging) is mounted on the upper part of the vehicle that appears in the road image.

[0041] Furthermore, the vehicle attribute identification unit 122 may identify the vehicle attribute as "autonomous vehicle" if there is no driver in the vehicle shown in the road image. The vehicle attribute identification unit 122 may also identify the vehicle attribute as the attribute corresponding to a mark attached to the vehicle shown in the road image. The mark attached to the vehicle may be, for example, an emblem or an inscription indicating the name of the vehicle, which can identify the vehicle's manufacturer or model. For example, if the mark attached to the vehicle corresponds to autonomous vehicle model A, the vehicle attribute identification unit 122 may identify the vehicle attribute as "autonomous vehicle model A". Also, for example, if the mark attached to the vehicle corresponds to autonomous vehicle model B, the vehicle attribute identification unit 122 may identify the vehicle attribute as "autonomous vehicle model B". Furthermore, the vehicle attribute identification unit 122 may identify the vehicle attribute as "manual vehicle" if the mark attached to the vehicle corresponds to a manual vehicle model.

[0042] Furthermore, the vehicle attribute identification unit 122 may identify the attributes of the vehicle related to autonomous driving, as described above, using a trained model obtained by taking an image of a vehicle as input and learning to output the attributes of the vehicle related to autonomous driving. The trained model is generated by a machine learning algorithm such as a neural network. For example, the trained model may be generated by taking an image of a vehicle as input and performing a learning process to output the attributes of the vehicle related to autonomous driving. For example, the trained model may be generated by taking an image of an autonomous vehicle of a certain model A as input and performing a learning process to output the attributes of the vehicle related to autonomous driving as "Model A". Alternatively, for example, the trained model may be generated by taking an image of an autonomous vehicle of a certain model B as input and performing a learning process to output the attributes of the vehicle related to autonomous driving as "Model B". Alternatively, for example, the trained model may be generated by taking an image of a manually driven vehicle as input and performing a learning process to output the attributes of the vehicle related to autonomous driving as "Manually driven vehicle". In this case, the trained model may estimate the attributes of the vehicle related to autonomous driving from the appearance of the vehicle.

[0043] Furthermore, the vehicle attribute identification unit 122 may identify the attributes of the vehicle related to autonomous driving by communicating with the vehicle. Specifically, the vehicle attribute identification unit 122 may identify the attributes of the vehicle by receiving a signal from the vehicle indicating the attributes of the vehicle related to autonomous driving. Alternatively, the vehicle attribute identification unit 122 may identify the type of autonomous vehicle by receiving a signal from the autonomous vehicle indicating the type of autonomous vehicle. On the other hand, if the vehicle attribute identification unit 122 is unable to receive a signal from the vehicle indicating the type of autonomous vehicle, it may identify the vehicle as a manually driven vehicle. Note that a communication device installed at the location where the roadside imaging device 30 is installed may communicate with the vehicle and receive a signal indicating the attributes of the vehicle. The vehicle attribute identification unit 122 may then identify the attributes of the vehicle by acquiring that signal from the communication device.

[0044] The learned model storage unit 124 stores the learned models used for identifying the attributes by the vehicle attribute identification unit 122. Further, the learned model storage unit 124 may perform learning processing on the learned model as described above.

[0045] The driving data acquisition unit 130 corresponds to the driving data acquisition unit 6 shown in FIG. 1. The driving data acquisition unit 130 acquires driving data indicating the behavior of the vehicle at a location for each of the attributes of at least two vehicles. The driving data acquisition unit 130 may acquire driving data indicating the behavior of the vehicle at a location for each of at least two attributes for each type of behavior. The driving data acquisition unit 130 may acquire driving data indicating the behavior of the vehicle at a location for each of all the attributes of the vehicle for each type of behavior. The driving data acquisition unit 130 may generate driving data indicating the behavior of the vehicle at each location using the road image acquired by the image acquisition unit 120.

[0046] For the sake of explanation, for example, it is assumed that the vehicle attribute identification unit 122 identifies, as vehicle attributes, attribute A (for example, autonomous driving vehicle type A), attribute B (for example, autonomous driving vehicle type B), attribute C (for example, autonomous driving vehicle type C), and attribute X (for example, manual driving vehicle). In this case, the driving data acquisition unit 130 acquires, for each type of behavior, the driving data indicating the behavior of the vehicle of attribute A at location A. Further, the driving data acquisition unit 130 acquires, for each type of behavior, the driving data indicating the behavior of the vehicle of attribute B at location A. Further, the driving data acquisition unit 130 acquires, for each type of behavior, the driving data indicating the behavior of the vehicle of attribute C at location A. Further, the driving data acquisition unit 130 acquires, for each type of behavior, the driving data indicating the behavior of the vehicle of attribute X at location A.

[0047] For example, the driving data acquisition unit 130 acquires driving data indicating the "behavior related to stopping" of a vehicle with attribute A at point A. Similarly, the driving data acquisition unit 130 acquires driving data indicating each of the "behavior related to turning", "behavior related to straight driving", "behavior related to lane change", and "behavior related to starting" of the vehicle with attribute A at point A. Similarly, the driving data acquisition unit 130 acquires driving data indicating the behaviors of vehicles with other attributes B, C, and X at point A for each type of behavior. Further, the driving data acquisition unit 130 acquires driving data in the same manner for a plurality of other points.

[0048] The driving data storage unit 132 stores the driving data acquired by the driving data acquisition unit 130. The driving data storage unit 132 stores the driving data for each attribute of the vehicle specified by the vehicle attribute specifying unit 122. Also, the driving data storage unit 132 stores the driving data for each type of behavior. Further, the driving data storage unit 132 may store the overall driving data for each type of behavior.

[0049] FIG. 6 is a diagram for explaining the driving data acquisition process by the information processing apparatus 100 according to the present disclosure. FIG. 6 shows an example in which two roadside imaging devices 30A and 30B are installed at a point Pt1 which is an intersection. Hereinafter, the processing for point Pt1 will be described, but the same applies to other points. In FIG. 6, the broken line indicates the boundary of the imaging range of the roadside imaging device 30.

[0050] The image acquisition unit 120 acquires road images from each of the roadside imaging devices 30A and 30B. It is assumed that the road images are moving images. The vehicle attribute specifying unit 122 specifies the attributes of the vehicles VcA, VcB, VcC, and VcD reflected in the road images as described above using the road images acquired from each of the roadside imaging devices 30A and 30B.

[0051] The driving data acquisition unit 130 synthesizes road images acquired from roadside imaging devices 30A and 30B to generate a road image showing the entire area of ​​point Pt1. For each of the vehicles VcA, VcB, VcC, and VcD visible in the road image, the driving data acquisition unit 130 acquires behavior data indicating the driving indicators of each behavior at point Pt1, in association with the attributes of each vehicle. Examples of behavior data will be described later. The processing for vehicle VcA will be described below, but the same applies to vehicles VcB, VcC, and VcD.

[0052] The driving data acquisition unit 130 tracks vehicle VcA by detecting its position in each frame of the road image and acquires the trajectory of vehicle VcA at point Pt1. Furthermore, by detecting the position of vehicle VcA in each frame of the road image, it acquires the speed of vehicle VcA at point Pt1 from the distance traveled by vehicle VcA between frames and the frame rate.

[0053] Furthermore, the driving data acquisition unit 130 acquires behavioral data indicating driving indicators for each behavior from the trajectory and speed of the vehicle VcA at point Pt1. The driving data acquisition unit 130 may also acquire behavioral data indicating driving indicators for stopping behavior from the trajectory and speed of the vehicle VcA at point Pt1. For example, the driving data acquisition unit 130 may acquire the stopping position of the vehicle VcA as a driving indicator for the stopping behavior of the vehicle VcA. For example, the driving data acquisition unit 130 may acquire the relative position of the stopping position of the vehicle VcA with respect to the stop line as the stopping position of the vehicle VcA. Alternatively, for example, the driving data acquisition unit 130 may acquire the distance from the stop line to the stopping position of the vehicle VcA as the stopping position of the vehicle VcA. Alternatively, for example, the driving data acquisition unit 130 may acquire the deceleration of the vehicle VcA as a driving indicator for the stopping behavior of the vehicle VcA. For example, the driving data acquisition unit 130 may acquire the average speed from a position a predetermined distance before the stopping position of vehicle VcA to the stopping position as the degree of deceleration of vehicle VcA. Alternatively, for example, the driving data acquisition unit 130 may acquire the deceleration (negative acceleration) until vehicle VcA comes to a stop as the degree of deceleration of vehicle VcA.

[0054] Furthermore, the driving data acquisition unit 130 may acquire behavioral data indicating driving indicators of turning behavior from the trajectory and speed of vehicle VcA at point Pt1. For example, the driving data acquisition unit 130 may acquire a value indicating the degree of curvature of the curve of the trajectory when vehicle VcA makes a right or left turn, as a driving indicator of turning behavior for vehicle VcA. For example, the driving data acquisition unit 130 may acquire the radius of curvature of the curve as a value indicating the degree of curvature of the curve. Alternatively, for example, the driving data acquisition unit 130 may acquire the shortest distance between a predetermined position at the intersection (e.g., the center of the intersection) and the curve as a value indicating the degree of curvature of the curve. Alternatively, for example, the driving data acquisition unit 130 may acquire the speed when vehicle VcA is making a right or left turn, as a driving indicator of turning behavior for vehicle VcA.

[0055] Furthermore, the driving data acquisition unit 130 may acquire behavioral data indicating driving indicators for straight-line driving behavior from the trajectory and speed of vehicle VcA at point Pt1. For example, the driving data acquisition unit 130 may acquire the relative position of vehicle VcA's driving position in the lane as a driving indicator for straight-line driving behavior of vehicle VcA. In this case, for example, the driving data acquisition unit 130 may acquire the distance between the lane boundary line and the driving position of vehicle VcA. Alternatively, for example, the driving data acquisition unit 130 may acquire the speed of vehicle VcA when it is driving in a straight line as a driving indicator for straight-line driving behavior of vehicle VcA.

[0056] Furthermore, the driving data acquisition unit 130 may acquire behavioral data indicating driving indicators of lane change behavior from the trajectory and speed of vehicle VcA at point Pt1. For example, the driving data acquisition unit 130 may acquire a value indicating the timing of a lane change as a driving indicator of the lane change behavior of vehicle VcA. In this case, for example, when vehicle VcA changes lanes to a right-turn lane in order to make a right turn, the driving data acquisition unit 130 may acquire the distance between the point where vehicle VcA made the lane change and the stop line as a value indicating the timing of the lane change. Alternatively, for example, the driving data acquisition unit 130 may acquire the distance from the position where vehicle VcA started the lane change to the position where vehicle VcA finished the lane change as a value indicating the timing of the lane change. Alternatively, for example, the driving data acquisition unit 130 may acquire the time taken from when vehicle VcA started the lane change until it finished the lane change as a value indicating the timing of the lane change. Furthermore, for example, the driving data acquisition unit 130 may acquire the speed of vehicle VcA when it is changing lanes, as a driving index of the vehicle VcA's behavior regarding lane changes.

[0057] Furthermore, the driving data acquisition unit 130 may acquire behavioral data indicating driving indicators of starting behavior from the trajectory and speed of the vehicle VcA at point Pt1. For example, the driving data acquisition unit 130 may acquire a value indicating the timing of starting as a driving indicator of the starting behavior of the vehicle VcA. In this case, for example, the driving data acquisition unit 130 may acquire the time from when the traffic signal changes to an indication of "permission to proceed" (for example, the traffic signal turning "green") until the vehicle VcA starts moving as a value indicating the timing of starting. Alternatively, for example, the driving data acquisition unit 130 may acquire the acceleration when the vehicle VcA starts moving as a driving indicator of the starting behavior of the vehicle VcA.

[0058] Figure 7 is a diagram illustrating behavioral data acquired by the driving data acquisition unit 130 according to this disclosure. In the example of Figure 7, behavioral data for vehicles VcA, VcB, VcC, and VcD at point Pt1 is illustrated. In the example of Figure 7, the attribute of vehicle VcA is specified as attribute A. The attribute of vehicle VcB is specified as attribute B. The attribute of vehicle VcC is specified as attribute C. The attribute of vehicle VcD is specified as attribute X.

[0059] Furthermore, Figure 7 shows the values ​​of driving index A1 and driving index A2 for each of the vehicles VcA, VcB, VcC, and VcD in behavior A, as behavior data for behavior A. Also, Figure 7 shows the values ​​of driving index B1 and driving index B2 for each of the vehicles VcA, VcB, VcC, and VcD in behavior B, as behavior data for behavior B. Behavior A is, for example, "behavior related to stopping". Driving index A1 is, for example, "stopping position", and driving index A2 is, for example, "deceleration rate". Behavior B is, for example, "behavior related to turning". Driving index B1 is, for example, "degree of turning", and driving index B2 is, for example, "speed".

[0060] Furthermore, the driving data acquisition unit 130 aggregates the behavioral data for each attribute at point Pt1 to acquire driving data that shows the frequency distribution of each behavior for each attribute. In other words, the driving data can be said to be statistical data obtained by aggregating the behavioral data for each attribute. For example, the driving data acquisition unit 130 uses multiple behavioral data X1 of the driving index X1 of behavior X by a vehicle of attribute A to perform the following processing. Note that behavior X is, for example, "behavior related to stopping," and driving index X1 is, for example, "stopping position."

[0061] The driving data acquisition unit 130 calculates (counts) the frequency of values ​​that fit into each interval (class) of the driving index X1 for attribute A. This allows the driving data acquisition unit 130 to acquire driving data X1 that shows the frequency distribution of the driving index X1 for behavior X for attribute A. The driving data acquisition unit 130 also performs a similar process for other driving indexes of behavior X for attribute A, thereby acquiring driving data for each driving index of behavior X for attribute A. Furthermore, the driving data acquisition unit 130 performs a similar process for other behaviors related to attribute A, thereby acquiring driving data for each behavior related to attribute A. Finally, the driving data acquisition unit 130 performs a similar process for each of the other attributes B, C, and X, thereby acquiring driving data for each of the behaviors related to attribute B, C, and X.

[0062] Furthermore, the driving data acquisition unit 130 acquires overall driving data, which is driving data relating to vehicles with at least two attributes for each type of behavior. For example, the driving data acquisition unit 130 may acquire overall driving data X1 relating to vehicles with attributes A, B, and C corresponding to autonomous driving vehicles, for the driving index X1 of behavior X. Specifically, the driving data acquisition unit 130 may acquire overall driving data X1 by performing the above-described processing using multiple behavior data X1 of the driving index X1 of behavior X by vehicles with attributes A, B, and C. Alternatively, the driving data acquisition unit 130 may acquire overall driving data X1 by combining the driving data X1 for attribute A, the driving data X1 for attribute B, and the driving data X1 for attribute C.

[0063] The driving data acquisition unit 130 may acquire overall driving data, which is driving data for all vehicles with all attributes, including manually driven vehicles. In this case, the driving data acquisition unit 130 acquires overall driving data X1 for vehicles with attributes A, B, C, and X, based on the driving index X1 of behavior X. Specifically, the driving data acquisition unit 130 may acquire overall driving data X1 by performing the above-described processing using multiple behavior data X1 of the driving index X1 of behavior X by vehicles with attributes A, B, C, and X. Alternatively, the driving data acquisition unit 130 may acquire overall driving data X1 by combining the driving data X1 for attribute A, the driving data X1 for attribute B, the driving data X1 for attribute C, and the driving data X1 for attribute X.

[0064] Figures 8 and 9 illustrate driving data acquired by the driving data acquisition process of the information processing device 100 according to this disclosure. Figures 8 and 9 may show, for example, overall driving data relating to attribute A, attribute B, and attribute C, or overall driving data relating to attribute A, attribute B, attribute C, and attribute X. Alternatively, Figures 8 and 9 may show driving data relating to attribute A, attribute B, and attribute C, respectively, corresponding to an autonomous driving vehicle (autonomous driving vehicle driving data).

[0065] Figure 8 illustrates driving data D11 and D12 related to "stopping behavior" at a certain point Pt1. Driving data D11 is a graph showing the frequency distribution of stopping positions. In the graph of driving data D11, the horizontal axis shows the stopping position, and the vertical axis shows the frequency for each section of the stopping position. A stopping position with a value greater than the position of the stop line indicates that the vehicle stopped beyond the stop line. Driving data D12 is a graph showing the frequency distribution of deceleration. In the graph of driving data D12, the horizontal axis shows the deceleration, and the vertical axis shows the frequency for each section of the deceleration. Mark M11 indicates the value of the stopping position of vehicle VcA, and mark M12 indicates the value of the deceleration of vehicle VcA.

[0066] Figure 9 illustrates driving data D21 and D22 related to the "behavior related to turning" when turning right at a certain point Pt1. Driving data D21 is a graph showing the frequency distribution of turning degree. In the graph of driving data D21, the horizontal axis shows the turning degree, and the vertical axis shows the frequency for each section of the turning degree. A larger turning degree value indicates a larger radius of curvature. Driving data D22 is a graph showing the frequency distribution of speed during the right turn. In the graph of driving data D22, the horizontal axis shows the speed, and the vertical axis shows the frequency for each section of the speed. Mark M21 indicates the turning degree value of vehicle VcA when turning right. Mark M22 indicates the speed of vehicle VcA during the right turn.

[0067] The deviation degree determination unit 140 corresponds to the deviation degree determination unit 8 shown in Figure 1. The deviation degree determination unit 140 compares the overall driving data with the driving data (autonomous vehicle driving data) for each attribute corresponding to the autonomous vehicle. The deviation degree determination unit 140 then determines the degree of deviation between the overall driving data and the driving data for each attribute corresponding to the autonomous vehicle. Specifically, for each type of behavior, the deviation degree determination unit 140 determines whether the degree of deviation between the overall driving data and the driving data for the vehicle with the attribute corresponding to the autonomous vehicle is large or small, using predetermined criteria. For example, for each type of behavior, the deviation degree determination unit 140 determines whether the degree of deviation between the frequency distribution of the behavior shown in the overall driving data and the frequency distribution of the behavior shown in the driving data for each attribute corresponding to the autonomous vehicle is large or small.

[0068] The deviation degree determination unit 140 then determines that the deviation degree for a certain behavior is large for that behavior if the deviation degree for a certain behavior with respect to a certain attribute corresponding to an autonomous driving vehicle meets the determination criteria. For example, the deviation degree determination unit 140 determines that the deviation degree for that behavior is large for that attribute if the deviation degree for a certain behavior with respect to a certain attribute is greater than or equal to a predetermined threshold.

[0069] Figure 10 is a diagram illustrating the processing of the deviation degree determination unit 140 according to this disclosure. Figure 10 shows an example of driving data for the driving index "degree of turning" in "behavior related to turning" at a certain point. The dashed line shows the overall driving data D31. The overall driving data D31 is a graph showing the frequency distribution of the degree of turning for vehicles of at least two attributes. The solid line shows the driving data D32 for vehicles of attribute A, which correspond to autonomous driving vehicles. The driving data D32 is a graph showing the frequency distribution of the degree of turning for vehicles of attribute A. Note that although Figure 10 shows "behavior related to turning" as an example of a type of behavior, the same applies to the degree of deviation for other behaviors. Also, the following explanation describes an example of comparing driving data for vehicles of attribute A with overall driving data, but note that the same applies to other attributes (autonomous driving vehicles). The deviation degree determination unit 140 determines the degree of deviation between the frequency distribution of the degree of turning shown in the overall driving data D31 and the frequency distribution of the degree of turning shown in the driving data D32 related to attribute A.

[0070] In the example shown in Figure 10, the dispersion (variance) of the driving data D32 is considerably smaller compared to the dispersion (variance) of the overall driving data D31. In other words, as mentioned above, if the autonomous vehicle models are the same, the specifications for autonomous driving in multiple autonomous vehicles are the same. Furthermore, the operation of an autonomous vehicle is less dependent on the individual driving operations of the driver. Therefore, if multiple autonomous vehicles with the same attributes perform a certain action at a certain point, they are likely to perform similar actions. Consequently, the dispersion (spread of frequency distribution) of the driving data D32 related to the attributes of autonomous vehicles can be extremely small.

[0071] In the example shown in Figure 10, for example, the deviation degree determination unit 140 may determine the deviation degree as the difference between the mode of the frequency distribution of the overall driving data D31 and the mode of the frequency distribution of the driving data D32 relating to attribute A. The deviation degree determination unit 140 may then determine that the deviation degree for "behavior related to turning" is large if the difference between the mode of the frequency distribution of the overall driving data D31 and the mode of the frequency distribution of the driving data D32 is greater than or equal to a predetermined threshold. In particular, in this case, the deviation degree determination unit 140 may determine that the deviation degree for the degree of turning in "behavior related to turning" is large. The same applies when using the mean or median, as described later.

[0072] Furthermore, the "predetermined threshold" may be set from parameters obtained from the overall driving data D31. For example, the "predetermined threshold" may be set appropriately using the standard deviation of the overall driving data D31. In this case, the "predetermined threshold" may be set to, for example, twice the standard deviation of the overall driving data D31. Alternatively, the "predetermined threshold" may be, for example, the difference between the mode of the overall driving data D31 and the value corresponding to the upper and lower x% (e.g., 5%) of the overall driving data D31 in terms of turning (driving index). The same applies when using the mean or median, as described later.

[0073] Alternatively, for example, the deviation degree determination unit 140 may determine the deviation degree by the difference between the average value of the frequency distribution of the overall driving data D31 and the average value of the frequency distribution of the driving data D32 related to attribute A. Alternatively, for example, the deviation degree determination unit 140 may determine the deviation degree by the difference between the median value of the frequency distribution of the overall driving data D31 and the median value of the frequency distribution of the driving data D32 related to attribute A.

[0074] Furthermore, for example, the deviation degree determination unit 140 may determine the degree of deviation by determining that the smaller the area of ​​overlap between the frequency distribution of the overall driving data D31 and the frequency distribution of the driving data D32 related to attribute A, the greater the degree of deviation. The deviation degree determination unit 140 may also determine that the degree of deviation regarding "behavior related to turning" is large if the area of ​​overlap between the frequency distribution of the overall driving data D31 and the frequency distribution of the driving data D32 is below a predetermined threshold. In particular, in this case, the deviation degree determination unit 140 may determine that the degree of deviation regarding the degree of turning in "behavior related to turning" is large.

[0075] Furthermore, for example, the deviation degree determination unit 140 may use the overall driving data D31 to set a target value for "behavior related to turning." The deviation degree determination unit 140 may then use the difference between the set target value and the driving data D32 related to attribute A as the deviation degree.

[0076] The abnormal vehicle identification unit 150 corresponds to the vehicle identification unit 10 shown in Figure 1. The abnormal vehicle identification unit 150 identifies abnormal vehicle types. The abnormal vehicle identification unit 150 identifies the vehicle type of an autonomous vehicle whose attribute is determined to have a large degree of deviation by the deviation degree determination unit 140 as an abnormal vehicle type (first vehicle type). The abnormal vehicle identification unit 150 also identifies the behavior determined to have a large degree of deviation for the identified abnormal vehicle type as a behavior in which the autonomous vehicle of that abnormal vehicle type is highly likely to perform an abnormal operation. The abnormal vehicle identification unit 150 then identifies the behavior in which the autonomous vehicle of the abnormal vehicle type is highly likely to perform an abnormal operation as an abnormal behavior (first behavior) related to that abnormal vehicle type. In this specification, "abnormal operation" is used to mean an operation that differs from the overall operation trend of the vehicle with the attribute corresponding to the overall driving data.

[0077] Specifically, for example, the abnormal vehicle identification unit 150 may identify an autonomous vehicle with an attribute that has been determined to have a large degree of deviation for a predetermined number of behavior types Th1 or more as an abnormal vehicle (first vehicle). For example, suppose that when the predetermined number Th1 is 2, it is determined that there is a large degree of deviation between the driving data of an autonomous vehicle with attribute A and the overall driving data for behavior A (e.g., "behavior related to turning") and behavior B (e.g., "behavior related to stopping"). In this case, the abnormal vehicle identification unit 150 identifies the autonomous vehicle with attribute A as an abnormal vehicle. In this case, the abnormal vehicle identification unit 150 also identifies that there is a high probability that the autonomous vehicle with attribute A, which is an abnormal vehicle, will perform abnormal actions for behavior A (e.g., "behavior related to turning") and behavior B (e.g., "behavior related to stopping"). Therefore, the abnormal vehicle identification unit 150 identifies behavior A (for example, "behavior related to turning") and behavior B (for example, "behavior related to stopping") as abnormal behaviors (first behaviors) related to attribute A.

[0078] Furthermore, the abnormal vehicle identification unit 150 may identify an abnormal vehicle (first vehicle) that has been determined to have a large degree of deviation for each type of behavior. For example, suppose that for behavior A (e.g., "behavior related to turning"), it is determined that there is a large degree of deviation between the driving data for each of the autonomous vehicles with attributes A and B and the overall driving data. In this case, the abnormal vehicle identification unit 150 may identify the autonomous vehicles with attributes A and B as abnormal vehicles for behavior A. In this case, the abnormal vehicle identification unit 150 may also identify behavior A (e.g., "behavior related to turning") as abnormal behavior related to the abnormal vehicles with attributes A and B. In particular, in the example of Figure 10, the abnormal vehicle identification unit 150 may identify the degree of turning in "behavior related to turning" as abnormal behavior related to the autonomous vehicles with attributes A and B that are abnormal vehicles.

[0079] Furthermore, abnormal behavior corresponds to driving data related to the behavior of the abnormal vehicle that shows a large degree of deviation from the overall driving data. Therefore, abnormal behavior is likely to be behavior in which the autonomous vehicle of the abnormal vehicle performs unique actions compared to vehicles with attributes corresponding to the overall driving data. For example, if "behavior related to turning" is abnormal behavior, the autonomous vehicle of the abnormal vehicle may perform unique actions in "behavior related to turning," that is, in turning operations such as turning right or left or driving around curves. In particular, in the example in Figure 10, the degree of turning in "behavior related to turning" may be abnormal in the autonomous vehicle of the abnormal vehicle.

[0080] Furthermore, the deviation degree determination unit 140 may determine the degree of deviation for each attribute using driving data obtained from multiple locations. That is, the deviation degree determination unit 140 may determine whether the degree of deviation between the overall driving data and the driving data for each attribute corresponding to the autonomous vehicle type is large at each of the multiple locations. The abnormal vehicle type identification unit 150 may then identify attribute A as an abnormal vehicle type if it is determined that the degree of deviation between the overall driving data and the driving data for attribute A is large at a predetermined number of Th2 or more locations. That is, the abnormal vehicle type identification unit 150 may identify the autonomous vehicle type of an attribute that has been determined to have a large degree of deviation at a predetermined number of Th2 or more locations as an abnormal vehicle type (first vehicle type).

[0081] Furthermore, the deviation degree determination unit 140 may determine the degree of deviation for each attribute for each behavior using driving data obtained at multiple locations. That is, the deviation degree determination unit 140 may determine whether the degree of deviation between the overall driving data and the driving data for each attribute of the autonomous vehicle is large at each of the multiple locations for each behavior. The abnormal vehicle identification unit 150 may identify attribute A as an abnormal vehicle if it is determined that the degree of deviation between the overall driving data and the driving data for attribute A is large at a predetermined number of Th2 or more locations for a certain behavior X. In this case, the abnormal vehicle identification unit 150 may also identify behavior X as an abnormal behavior (first behavior) related to attribute A. Furthermore, the abnormal vehicle identification unit 150 may identify attribute A as an abnormal vehicle if it is determined that the degree of deviation between the overall driving data and the driving data for attribute A is large at a predetermined number of Th1 or more locations for a predetermined number of behaviors at a predetermined number of Th2 or more locations.

[0082] The vehicle information display unit 170 corresponds to the display unit 12 shown in Figure 1. The vehicle information display unit 170 performs processing to display abnormal vehicle information relating to the abnormal vehicle identified by the abnormal vehicle identification unit 150 to the destination facility. Specifically, the vehicle information display unit 170 performs processing to transmit the abnormal vehicle information to the terminal device 40. The vehicle information display unit 170 may also control the communication unit 106 to transmit the abnormal vehicle information to the terminal device 40. Furthermore, the vehicle information display unit 170 may perform processing to transmit an instruction to the terminal device 40 so that the abnormal vehicle information is displayed on the display of the terminal device 40. The abnormal vehicle information indicates, for example, the abnormal vehicle and the abnormal behavior relating to the abnormal vehicle. Therefore, the vehicle information display unit 170 performs processing to display information indicating the identified abnormal vehicle (first vehicle) and the abnormal behavior (first behavior) that has been determined to have a large degree of deviation for that abnormal vehicle. The vehicle information display unit 170 may also display a message on the terminal device 40 prompting the user to correct the specifications related to the autonomous driving of the abnormal vehicle (for example, the autonomous driving algorithm).

[0083] Then, the vehicle information display unit 170 transmits the abnormal vehicle information to the terminal device 40, which displays the abnormal vehicle information. In this way, the abnormal vehicle information is presented to the receiving facility. By presenting the abnormal vehicle information to the receiving facility in this manner, the administrator of the receiving facility can appropriately identify the vehicle type of autonomous vehicle that may be performing abnormal operations. In other words, the administrator of the receiving facility can easily identify the abnormal vehicle type and the abnormal behavior associated with that vehicle type. The receiving facility is, for example, a vehicle manufacturer that produces the abnormal autonomous vehicle. In this case, the vehicle manufacturer can appropriately correct the specifications of the autonomous driving related to the abnormal vehicle. This allows the vehicle manufacturer to improve the autonomous driving capabilities of the autonomous vehicle.

[0084] Figure 11 is a flowchart illustrating an example of processing performed by the information processing device 100 according to this disclosure. Figure 11 also shows an information processing method performed by the information processing device 100 according to this disclosure. Furthermore, Figure 11 can also be said to show an information presentation method performed by the information processing device 100 according to this disclosure.

[0085] Furthermore, if each component shown in Figure 5 is implemented by cloud computing or multiple computers, the process illustrated in Figure 11 will be executed by the information processing device 100 implemented by cloud computing or multiple computers. Please note that the process described in Figure 11 is merely an example. These points also apply to other flowcharts.

[0086] As described above, the image acquisition unit 120 acquires road images obtained by photographing one or more locations where a vehicle can travel (step S100). As described above, the vehicle attribute identification unit 122 identifies the attributes related to autonomous driving of the vehicle shown in the road image (step S102). As described above, the driving data acquisition unit 130 acquires behavior data showing driving indicators for each behavior of each vehicle at each location (step S104). As described above, the driving data acquisition unit 130 uses the behavior data to acquire driving data (autonomous vehicle driving data) for each behavior related to the attributes of the autonomous vehicle type for each behavior (step S106). As described above, the driving data acquisition unit 130 uses the behavior data to acquire overall driving data for at least two attributes of the vehicle for each behavior (step S108).

[0087] As described above, the deviation degree determination unit 140 determines the degree of deviation between the overall driving data and the autonomous driving vehicle driving data (step S110). As described above, the abnormal vehicle identification unit 150 identifies an abnormal vehicle that is highly likely to perform abnormal operations (step S112). As described above, the vehicle information presentation unit 170 performs processing to present abnormal vehicle information, which indicates the abnormal vehicle and the abnormal behavior related to the abnormal vehicle, to the vehicle manufacturer that manufactures the autonomous driving vehicle of the abnormal vehicle (step S114).

[0088] In the example described above, the overall driving data was assumed to be driving data for vehicles with at least two attributes. On the other hand, to identify abnormal vehicle types, it is preferable to have a larger number of attributes in the overall driving data. Therefore, the overall driving data may be driving data for vehicles with all attributes.

[0089] (Embodiment 3) Next, Embodiment 3 will be described with reference to the drawings. For clarity of explanation, the following descriptions and drawings have been omitted and simplified as appropriate. In addition, the same elements are denoted by the same reference numerals in each drawing, and redundant explanations have been omitted as necessary. The information presentation system 20 according to Embodiment 3 is substantially the same as that shown in Figure 3, so its description will be omitted. The hardware configuration of the information processing device according to Embodiment 3 is substantially the same as that shown in Figure 4, so its description will be omitted.

[0090] Figure 12 is a diagram illustrating the functional configuration of the information processing device 100 according to the present disclosure. Similar to Embodiment 2, the information processing device 100 has as its components an image acquisition unit 120, a vehicle attribute identification unit 122, a trained model storage unit 124, a driving data acquisition unit 130, and a driving data storage unit 132. The information processing device 100 also has as its components a deviation degree determination unit 140, an abnormal vehicle type identification unit 150, and a vehicle type information presentation unit 170. Furthermore, in Embodiment 3, the information processing device 100 has an operation identification unit 160. The operation identification unit 160 functions as an operation identification means.

[0091] The operation identification unit 160 identifies an abnormal operation (first operation) that is specific to the autonomous vehicle of the abnormal vehicle type (first vehicle type) in the abnormal behavior (first behavior). In other words, the operation identification unit 160 identifies an abnormal operation (first operation) that is specific to the autonomous vehicle of the abnormal vehicle type (first vehicle type) in the abnormal behavior (first behavior) that is determined to have a large degree of deviation for the abnormal vehicle type (first vehicle type). Here, "operation in behavior X" corresponds to the vehicle's operation performed by the operation when the vehicle performs behavior X. Also, "operation in behavior X" corresponds to the specific way the vehicle moves, represented by the driving index of behavior X, when behavior X is performed. And "abnormal operation X" corresponds to the specific characteristics (tendencies) of the autonomous vehicle of the abnormal vehicle type in abnormal behavior X. Specifically, "abnormal operation X" corresponds to the characteristic movement of the autonomous vehicle of the abnormal vehicle type in abnormal behavior X, compared with the movement of the vehicle with attributes corresponding to the overall driving data. In other words, "abnormal behavior X" corresponds to the unique movement of the abnormal vehicle type autonomous vehicle, compared to the movement of the vehicle with attributes corresponding to the overall driving data in abnormal behavior X. More specifically, "abnormal behavior X" corresponds to the trend of the driving indicators in the driving data of the abnormal vehicle type autonomous vehicle, compared to the trend of the driving indicators in the overall driving data for abnormal behavior X.

[0092] For example, if behavior X is "behavior related to stopping," then "action in behavior X" corresponds to how much the vehicle decelerates when stopping, and at what position relative to the stop line the vehicle stops. And, if behavior X is "behavior related to stopping," then "abnormal action X" corresponds to, for example, an abnormal self-driving car decelerating with a large deceleration rate when stopping, and an abnormal self-driving car stopping beyond the stop line.

[0093] Furthermore, for example, if behavior X is "behavior related to turning," then "action in behavior X" corresponds to how much the vehicle turns and at what speed the vehicle turns. And if abnormal behavior X is "behavior related to turning," then "abnormal action X" corresponds to, for example, an abnormal self-driving car making a wide turn and an abnormal self-driving car turning at high speed.

[0094] Furthermore, for example, if behavior X is "behavior related to driving in a straight line," then "actions in behavior X" correspond to the position of the vehicle within the lane and the speed at which the vehicle is driving in a straight line. And if abnormal behavior X is "behavior related to driving in a straight line," then "abnormal action X" corresponds to, for example, an abnormal self-driving car driving across the lane boundary and an abnormal self-driving car driving in a straight line at high speed.

[0095] Furthermore, for example, if behavior X is "behavior related to lane changes," then "actions in behavior X" correspond to when the vehicle changes lanes and how fast the vehicle travels while changing lanes. And if abnormal behavior X is "behavior related to lane changes," then "abnormal action X" corresponds to, for example, an abnormal autonomous vehicle changing lanes at a late timing and an abnormal autonomous vehicle changing lanes at a high speed.

[0096] Furthermore, for example, if behavior X is "behavior related to starting," then "action in behavior X" corresponds to when the vehicle starts moving and with what acceleration the vehicle starts moving. And if abnormal behavior X is "behavior related to starting," then "abnormal action X" corresponds to, for example, an abnormal self-driving car starting at a late timing and an abnormal self-driving car starting with high acceleration.

[0097] The operation identification unit 160 may identify an operation as abnormal operation X if the frequency of the operation performed by the abnormal vehicle type autonomous vehicle in the abnormal operation X is higher than the frequency of the operation performed by the vehicle with the attribute corresponding to the overall driving data. For example, in the example of Figure 10, suppose the deviation degree determination unit 140 determines that there is a large deviation between the frequency distribution of turning degrees shown in the overall driving data D31 and the frequency distribution of turning degrees shown in the driving data D32 related to attribute A of the autonomous vehicle. In this case, the abnormal vehicle type identification unit 150 identifies attribute A as an abnormal vehicle type and identifies the "turning behavior" as abnormal behavior. In the example of Figure 10, regarding the driving index "turning degree" in the "turning behavior," the frequency of sections with large turning degrees is higher in the driving data D32 related to attribute A of the abnormal vehicle type compared to the overall driving data D31.

[0098] Therefore, the operation identification unit 160 identifies abnormal operation X as an autonomous vehicle with attribute A making a large turn in the "behavior related to turning." In other words, the operation identification unit 160 identifies abnormal operation X as an autonomous vehicle with attribute A making a wide turn while traveling around a curve in the "behavior related to turning."

[0099] As described above, by identifying the abnormal operation X, the operation identification unit 160 can identify operations specific to the abnormal autonomous vehicle, compared with vehicles with attributes corresponding to the overall driving data, that is, vehicles containing attributes other than those of the abnormal autonomous vehicle. For example, if the abnormal vehicle has attribute A, the operation identification unit 160 can identify operations specific to the autonomous vehicle with attribute A, compared with autonomous vehicles of other attributes or manually driven vehicles.

[0100] The vehicle information display unit 170 performs processing to display the abnormal vehicle and abnormal behavior related to the abnormal vehicle, identified by the abnormal vehicle identification unit 150, and the abnormal operation identified by the operation identification unit 160, to the destination facility. Specifically, the vehicle information display unit 170 performs processing to transmit abnormal vehicle information, which indicates the abnormal vehicle, the abnormal behavior related to the abnormal vehicle, and the abnormal operation in the abnormal behavior, to the terminal device 40. The vehicle information display unit 170 may also control the communication unit 106 to transmit the abnormal vehicle information to the terminal device 40. Furthermore, the vehicle information display unit 170 may perform processing to transmit an instruction to the terminal device 40 so that the abnormal vehicle information is displayed on the display of the terminal device 40. The vehicle information display unit 170 may also cause the terminal device 40 to display a message prompting the correction of the specifications related to the automatic driving of the abnormal vehicle (e.g., the automatic driving algorithm).

[0101] Then, the vehicle information display unit 170 transmits the abnormal vehicle information to the terminal device 40, which displays the abnormal vehicle information. In this way, the abnormal vehicle information is presented to the receiving facility. By presenting the abnormal vehicle information to the receiving facility in this manner, the administrator can appropriately identify the vehicle type of autonomous vehicle that may be performing abnormal operations. In other words, the administrator of the receiving facility can easily identify the abnormal vehicle, the abnormal behavior related to that vehicle, and the abnormal operation within that abnormal behavior. The receiving facility is, for example, a vehicle manufacturer that produces the abnormal autonomous vehicle. In this case, the vehicle manufacturer can use the information on the abnormal behavior and operation to appropriately correct the specifications of the autonomous driving related to the abnormal vehicle.

[0102] For example, suppose the abnormal vehicle has attribute A, the abnormal behavior is related to "turning behavior," and the abnormal operation is "an autonomous vehicle with attribute A making a wide turn around a curve." In this case, the vehicle manufacturer can modify the specifications for autonomous driving of the autonomous vehicle with attribute A so that it suppresses "an autonomous vehicle with attribute A making a wide turn around a curve" in terms of "turning behavior." This allows the vehicle manufacturer to improve the autonomous driving capabilities of the autonomous vehicle.

[0103] Figure 13 is a flowchart illustrating an example of processing performed by the information processing device 100 according to this disclosure. Figure 13 also illustrates an information processing method performed by the information processing device 100 according to this disclosure. Furthermore, Figure 13 can be said to illustrate an information presentation method performed by the information processing device 100 according to this disclosure.

[0104] The information processing device 100 performs substantially the same processing as S100 to S112 described above using Figure 11 (step S202). The operation identification unit 160 identifies the abnormal operation in the abnormal behavior of the abnormal vehicle by the abnormal vehicle type autonomous vehicle, as described above (step S210). The vehicle type information presentation unit 170 performs processing to present abnormal vehicle type information, which indicates the abnormal vehicle type, the abnormal behavior related to the abnormal vehicle type, and the abnormal operation in the abnormal behavior, to the vehicle manufacturer that manufactures the abnormal vehicle type autonomous vehicle (step S212).

[0105] (Embodiment 4) Next, Embodiment 4 will be described with reference to the drawings. For clarity of explanation, the following descriptions and drawings have been omitted and simplified as appropriate. In addition, the same elements are denoted by the same reference numerals in each drawing, and redundant explanations have been omitted as necessary. The information presentation system 20 according to Embodiment 4 is substantially the same as that shown in Figure 3, so its description will be omitted. The hardware configuration of the information processing device according to Embodiment 4 is substantially the same as that shown in Figure 4, so its description will be omitted.

[0106] Figure 14 is a diagram illustrating the functional configuration of the information processing device 100 according to the present disclosure. Similar to Embodiment 2, the information processing device 100 has as its components an image acquisition unit 120, a vehicle attribute identification unit 122, a trained model storage unit 124, a driving data acquisition unit 130, and a driving data storage unit 132. The information processing device 100 also has as its components a deviation degree determination unit 140, an abnormal vehicle type identification unit 150, and a vehicle type information presentation unit 170. Furthermore, in Embodiment 4, the information processing device 100 has a location identification unit 166 and a location presentation unit 176. The location identification unit 166 has the function of a location identification means. The location presentation unit 176 has the function of a location presentation means (presentation means).

[0107] In Embodiment 4, the attribute types include manually driven vehicles and each type of autonomous vehicle. The overall driving data is driving data for vehicles with at least two attributes, including the attributes of manually driven vehicles. The driving data used for comparison with the overall driving data includes driving data for attributes corresponding to at least two types of autonomous vehicles. The destination facilities are facilities of organizations (such as local governments or traffic control organizations) that carry out infrastructure development at each location.

[0108] Furthermore, the deviation degree determination unit 140 determines, for each location, whether the deviation degree between the overall driving data and the driving data for vehicles with attributes corresponding to at least two types of autonomous vehicles is large, using predetermined determination criteria. In other words, in Embodiment 4, the deviation degree determination unit 140 compares, for each location, the overall driving data for vehicles with multiple attributes, including the attributes of manually driven vehicles, with the driving data for autonomous vehicles with multiple attributes. Here, "driving data for autonomous vehicles with multiple attributes" indicates the overall trend of the behavior of autonomous vehicles with multiple attributes.

[0109] Figure 15 is a diagram illustrating the processing of the deviation degree determination unit 140 according to this disclosure. Figure 15 illustrates driving data for the driving index "degree of turning" in "behavior related to turning" at a certain point. The dashed line shows the overall driving data D41. The overall driving data D41 is a graph showing the frequency distribution of the degree of turning for vehicles with at least two attributes, including the attribute of a manually driven vehicle. The solid line shows the driving data D42 for autonomous vehicles with at least two attributes. The driving data D42 is a graph showing the frequency distribution of the degree of turning for autonomous vehicles with attributes A, B, and C, for example. Note that although Figure 15 shows "behavior related to turning" as an example of a type of behavior, the same applies to the degree of deviation for other behaviors.

[0110] The deviation degree determination unit 140 determines the degree of deviation for each location between the frequency distribution of turning shown in the overall driving data D41 and the frequency distribution of turning shown in the driving data D42 relating to autonomous vehicles of attribute A, attribute B, and attribute C. The deviation degree determination unit 140 may determine the degree of deviation as the difference between the mode of the frequency distribution of the overall driving data D41 and the mode of the frequency distribution of the driving data D42 relating to autonomous vehicles of attribute A, attribute B, and attribute C, similar to the method described in Embodiment 2. The deviation degree determination unit 140 may then determine that the degree of deviation regarding "behavior related to turning" is large if the difference between the mode of the frequency distribution of the overall driving data D41 and the mode of the frequency distribution of the driving data D42 is greater than or equal to a predetermined threshold. The "predetermined threshold" may be different from the one described in Embodiment 2.

[0111] Furthermore, the deviation degree determination unit 140 may determine the deviation degree by the difference between the average value of the frequency distribution of the overall driving data D41 and the average value of the frequency distribution of the driving data D42 relating to autonomous vehicles of attributes A, B, and C, similar to the method described in Embodiment 2. Alternatively, the deviation degree determination unit 140 may determine the deviation degree by the difference between the median of the frequency distribution of the overall driving data D41 and the median of the frequency distribution of the driving data D42 relating to autonomous vehicles of attributes A, B, and C. Alternatively, the deviation degree determination unit 140 may determine the deviation degree by considering that the smaller the overlap area between the frequency distribution of the overall driving data D41 and the frequency distribution of the driving data D42 relating to autonomous vehicles of attributes A, B, and C, the greater the deviation degree. Furthermore, the deviation degree determination unit 140 may set a target value for "behavior related to turning" using the overall driving data D41. The deviation degree determination unit 140 may then use the difference between the set target value and the driving data D42 related to the autonomous vehicle with attributes A, B, and C as the deviation degree.

[0112] The location identification unit 166 identifies locations where a large degree of discrepancy is determined to be present as locations requiring improvement (first locations). The location identification unit 166 identifies locations where a large degree of discrepancy is determined between the overall driving data and the driving data relating to vehicles with attributes corresponding to at least two types of autonomous vehicles as locations requiring improvement. For example, suppose that for location Pt1, a large degree of discrepancy is determined between the overall driving data and the driving data relating to vehicles with attributes corresponding to at least two types of autonomous vehicles. In this case, the location identification unit 166 identifies location Pt1 as a location requiring improvement. A "location requiring improvement" is a location where it is desirable to improve the surrounding conditions so that autonomous driving by the autonomous vehicle can be performed appropriately.

[0113] Furthermore, the location identification unit 166 may identify locations where it has determined that the degree of deviation is large for a predetermined number of behavior types Th3 or more as locations requiring improvement. For example, suppose that the predetermined number Th3 is 2, and the deviation degree determination unit 140 determines that at location Pt1, the degree of deviation is large for behavior A and behavior B between the driving data of the autonomous vehicle with attributes A, B, and C and the overall driving data. In this case, the location identification unit 166 identifies location Pt1 as a location requiring improvement (first location).

[0114] The location display unit 176 performs processing to display information regarding locations requiring improvement (first locations). Specifically, the location display unit 176 performs processing to transmit the location requiring improvement information to a terminal device 40 in an organization that performs infrastructure development. The location display unit 176 may also control the communication unit 106 to transmit the location requiring improvement information to the terminal device 40. Furthermore, the location display unit 176 may perform processing to transmit an instruction to the terminal device 40 so that the location requiring improvement information is displayed on the display of the terminal device 40. The location requiring improvement information may, for example, indicate the location requiring improvement and the behavior of the driving data at the location requiring improvement where the degree of deviation from the overall driving data is determined to be large. In addition, the location display unit 176 may cause the terminal device 40 to display a message urging the improvement of the infrastructure at the location requiring improvement so that the autonomous vehicle can operate appropriately.

[0115] Then, the location display unit 176 transmits the location information requiring improvement to the terminal device 40, which displays the location information requiring improvement. In this way, the location information requiring improvement is presented to the destination facility. By presenting the location information requiring improvement to the destination facility in this manner, the administrator can appropriately identify locations where autonomous vehicles with multiple attributes may behave differently from vehicles with attributes related to the overall driving data. In other words, the administrator of the destination facility can easily identify locations requiring improvement and the behaviors at those locations where autonomous vehicles may behave abnormally.

[0116] Here, if multiple autonomous vehicles with different attributes exhibit abnormal behavior at a particular location, it is more likely that the abnormal behavior is due to a problem at that location rather than a specification issue with the autonomous vehicles themselves. Therefore, the organization responsible for infrastructure development can improve the infrastructure at locations requiring improvement so that autonomous vehicles behave appropriately. For example, if an autonomous vehicle may behave abnormally because it detects a picture on a sign around an intersection, the sign can be corrected. Also, if an autonomous vehicle may behave abnormally because it has difficulty detecting traffic signals, the location of the traffic signals can be improved so that autonomous vehicles can detect them more easily. Thus, autonomous vehicles can be made to behave appropriately.

[0117] Figure 16 is a flowchart illustrating an example of processing performed by the information processing device 100 according to this disclosure. Figure 16 also illustrates an information processing method performed by the information processing device 100 according to this disclosure. Furthermore, Figure 16 can be said to illustrate an information presentation method performed by the information processing device 100 according to this disclosure.

[0118] The image acquisition unit 120 acquires road images obtained by photographing one or more points where a vehicle can travel, similar to the process in S100 (step S300). The vehicle attribute identification unit 122 identifies attributes related to the autonomous driving of the vehicle shown in the road image, similar to the process in S102 (step S302). In this embodiment 4, the vehicle attribute identification unit 122 identifies a plurality of attributes, including the attributes of a manually driven vehicle. The information processing device 100 performs substantially the same processing as described in S104 to S114 using Figure 11 (step S304).

[0119] The driving data acquisition unit 130 uses the behavior data to acquire driving data (autonomous vehicle driving data) for vehicles with at least two attributes related to autonomous vehicle types for each behavior (step S310). The driving data acquisition unit 130 uses the behavior data to acquire overall driving data for vehicles with at least two attributes, including a manually driven vehicle, for each behavior (step S312).

[0120] As described above, the deviation degree determination unit 140 determines the degree of deviation between the overall driving data and the autonomous vehicle driving data for each point (step S314). As described above, the location identification unit 166 identifies the locations that need improvement (step S316). As described above, the location presentation unit 176 performs processing to present the location information that needs improvement to the organization that will carry out infrastructure development (step S318).

[0121] In the example described above, the overall driving data was assumed to be driving data for vehicles with at least two attributes, including manually driven vehicles. Similarly, the autonomous vehicle driving data was assumed to be driving data for vehicles with at least two attributes related to autonomous vehicles. On the other hand, in order to identify areas for improvement, it is preferable to have a large number of attributes for both the overall driving data and the autonomous vehicle driving data. Therefore, the overall driving data may be driving data for vehicles with all attributes. Similarly, the autonomous vehicle driving data may be driving data for vehicles with all attributes of autonomous vehicles.

[0122] (Variations) The present invention is not limited to the embodiments described above, and can be modified as appropriate without departing from the spirit of the invention. For example, the order of each process in the flowchart described above can be changed as appropriate. Also, one or more of the processes in the flowchart described above may be omitted.

[0123] Furthermore, for example, each process in the flowchart described above may be implemented using a trained model learned by machine learning. In other words, the deviation degree determination unit 140 described above may use a trained model learned by machine learning to determine the degree of deviation between the overall driving data and the driving data relating to the attributes of the autonomous vehicle. In this case, the trained model may be trained to take the overall driving data and the driving data relating to the attributes of the autonomous vehicle as input and output the degree of deviation between the two. Also, Embodiment 4 and Embodiment 3 may be applied to each other. That is, in Embodiment 4, abnormal operation relating to abnormal vehicle types may be identified.

[0124] Although the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure can be made as can be understood by those skilled in the art within the scope of the present disclosure. Furthermore, each embodiment can be combined with other embodiments as appropriate.

[0125] Each drawing is merely illustrative to illustrate one or more embodiments. Each drawing may be associated with one or more other embodiments, rather than being associated with only one specific embodiment. As those skilled in the art will understand, various features or steps described with reference to any one drawing can be combined with features or steps shown in one or more other drawings, for example, to create embodiments not explicitly shown or described. Not all features or steps shown in any one drawing to illustrate an exemplary embodiment are necessarily required, and some features or steps may be omitted. The order of steps described in any of the drawings may be changed as appropriate.

[0126] The program described above, when loaded into a computer, includes a set of instructions (or software code) for causing the computer to perform one or more of the functions described in the embodiments. The program may be stored in a non-temporary computer-readable medium or a physical storage medium. Examples, but not limited to, include random-access memory (RAM), read-only memory (ROM), flash memory, solid-state drive (SSD) or other memory technologies, CD-ROM, digital versatile disk (DVD), Blu-ray® disc or other optical disc storage, magnetic cassette, magnetic tape, magnetic disk storage or other magnetic storage devices. The program may be transmitted over a temporary computer-readable medium or a communication medium. Examples, but not limited to, include electrical, optical, acoustic or other forms of propagating signals. The program may also include a program product.

[0127] Although the present invention has been described above with reference to embodiments, the present invention is not limited thereto. Various modifications to the structure and details of the present invention can be made that are understandable to those skilled in the art within the scope of the invention.

[0128] Some or all of the above embodiments may also be described as follows, but are not limited to the following: (Note 1) An information processing device comprising: an image acquisition means for acquiring images obtained by photographing one or more locations where a vehicle can travel; an attribute identification means for identifying attributes related to the autonomous driving of a vehicle shown in the image; a driving data acquisition means for acquiring driving data indicating the behavior of the vehicle at the location for each of at least two vehicle attributes; a deviation degree determination means for determining whether there is a large deviation between overall driving data, which is driving data for vehicles of at least two attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle, using predetermined determination criteria; a vehicle type identification means for identifying the type of autonomous vehicle for the attribute determined to have a large deviation as a first vehicle type; and a presentation means for performing processing for presenting information relating to the identified first vehicle type. (Note 2) The information processing device according to Note 1, wherein the presentation means performs processing for presenting information relating to the identified first vehicle type to a vehicle manufacturer that manufactures an autonomous vehicle of the first vehicle type. (Note 3) The information processing device according to Note 1, wherein the deviation degree determination means determines whether the deviation degree is large at each of the plurality of points between the overall driving data and the driving data relating to each attribute corresponding to the type of autonomous vehicle, and the vehicle type identification means identifies the type of autonomous vehicle for the attribute that has been determined to have a large deviation degree at a predetermined number of points as the first vehicle type. (Note 4) The information processing device according to Note 1, wherein the type of behavior at the point includes at least one of the behaviors related to stopping, turning, going straight, changing lanes, and starting. (Note 5) The information processing device according to Note 1, wherein the deviation degree determination means determines whether the deviation degree is large for each type of behavior between the frequency distribution of the behavior shown in the overall driving data and the frequency distribution of the behavior shown in the driving data relating to each attribute corresponding to the type of autonomous vehicle.(Note 6) The information processing device according to Note 4 or 5, wherein the vehicle type identification means identifies the vehicle type of an autonomous vehicle whose attribute is determined to have a large degree of deviation for a predetermined number or more types of behavior as the first vehicle type. (Note 7) The information processing device according to Note 4 or 5, wherein the vehicle type identification means identifies the first vehicle type which is determined to have a large degree of deviation for each type of behavior. (Note 8) The information processing device according to Note 4 or 5, wherein the presentation means performs processing to present information indicating the identified first vehicle type and the first behavior which is determined to have a large degree of deviation for the first vehicle type. (Note 9) The information processing device according to Note 4 or 5, further comprising: an action identification means that identifies a first action which is an action specific to the autonomous vehicle of the first vehicle type in the first behavior which is determined to have a large degree of deviation for the first vehicle type. (Note 10) The information processing device according to Note 9, wherein the operation identification means identifies an operation as the first operation when the frequency of an operation performed by an autonomous vehicle of the first vehicle type in the first behavior is higher than the frequency of such operation performed by a vehicle with attributes corresponding to the overall driving data. (Note 11) The information processing device according to Note 9, wherein the presentation means performs processing to present information indicating the identified first vehicle type, the first behavior for the first vehicle type, and the first operation for the first vehicle type. (Note 12) The attributes include manual-driving vehicles and each type of autonomous vehicle, the deviation degree determination means determines for each of the points whether the deviation degree is large or small between the overall driving data for vehicles with at least two attributes, including the attributes of manual-driving vehicles, and the driving data for attributes corresponding to at least two types of autonomous vehicles, using predetermined determination criteria, and further comprises a point identification means that identifies the points where the deviation degree is determined to be large as first points, and the presentation means performs processing for presenting information about the first points, the information processing device according to Note 1.(Note 13) An information processing method comprising: acquiring images obtained by photographing one or more locations where a vehicle can travel; identifying attributes related to autonomous driving of the vehicle shown in the image; acquiring driving data indicating the behavior of the vehicle at the location for each of at least two vehicle attributes; determining, using predetermined criteria, whether the degree of discrepancy between the overall driving data, which is the driving data for at least two vehicle attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle is large; identifying the type of autonomous vehicle for the attribute determined to have a large degree of discrepancy as the first type of vehicle; and performing processing to present information regarding the identified first type of vehicle. (Note 14) A non-temporary computer-readable medium containing a program that causes a computer to execute the following steps: acquiring images obtained by photographing one or more locations where a vehicle can travel; identifying attributes related to autonomous driving of the vehicle shown in the image; acquiring driving data indicating the behavior of the vehicle at the location for each of at least two vehicle attributes; determining, using predetermined criteria, whether there is a large discrepancy between the overall driving data, which is the driving data for at least two vehicle attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle; identifying the type of autonomous vehicle for the attribute determined to have a large discrepancy as a first type of vehicle; and performing processing to present information about the identified first type of vehicle.

[0129] Some or all of the elements (e.g., configuration and function) described in Appendices 2 to 12 that are dependent on Appendice 1 may also be dependent on Appendices 13 and 14 in the same manner as those described in Appendices 2 to 12. Some or all of the elements described in any appendice may be applied to various hardware, software, recording means, systems, and methods for recording software.

[0130] 1 Information Processing Device 2 Image Acquisition Unit 4 Attribute Identification Unit 6 Driving Data Acquisition Unit 8 Degree of Deviation Determination Unit 10 Vehicle Type Identification Unit 12 Presentation Unit 20 Information Presentation System 30 Roadside Imaging Device 40 Terminal Device 100 Information Processing Device 120 Image Acquisition Unit 122 Vehicle Attribute Identification Unit 124 Trained Model Storage Unit 130 Driving Data Acquisition Unit 132 Driving Data Storage Unit 140 Degree of Deviation Determination Unit 150 Abnormal Vehicle Type Identification Unit 160 Operation Identification Unit 166 Location Identification Unit 170 Vehicle Type Information Presentation Unit 176 Location Presentation Unit

Claims

1. An information processing device comprising: an image acquisition means for acquiring images obtained by photographing one or more locations where a vehicle can travel; an attribute identification means for identifying attributes related to the autonomous driving of a vehicle shown in the image; a driving data acquisition means for acquiring driving data indicating the behavior of the vehicle at the location for each of at least two vehicle attributes; a deviation degree determination means for determining whether there is a large deviation between overall driving data, which is driving data for vehicles of at least two attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle, using predetermined determination criteria; a vehicle type identification means for identifying the type of autonomous vehicle for the attribute determined to have a large deviation as a first vehicle type; and a presentation means for performing processing for presenting information about the identified first vehicle type.

2. The information processing apparatus according to claim 1, wherein the presenting means performs processing to present information relating to the specified first vehicle type to a vehicle manufacturer that manufactures an autonomous vehicle of the first vehicle type.

3. The information processing apparatus according to claim 1, wherein the deviation degree determination means determines whether the deviation degree between the overall driving data and the driving data relating to each attribute corresponding to the type of autonomous vehicle is large at each of the plurality of locations, and the vehicle type identification means identifies the type of autonomous vehicle for the attribute for which the deviation degree has been determined to be large at a predetermined number of locations as the first vehicle type.

4. The information processing device according to claim 1, wherein the type of behavior at the point includes at least one of the following behaviors: behavior related to stopping, behavior related to turning, behavior related to going straight, behavior related to changing lanes, and behavior related to starting.

5. The information processing apparatus according to claim 1, wherein the deviation degree determination means determines, for each type of behavior, whether the deviation degree is large between the frequency distribution of the behavior shown in the overall driving data and the frequency distribution of the behavior shown in the driving data for each attribute corresponding to the type of autonomous vehicle.

6. The information processing apparatus according to claim 4 or 5, wherein the vehicle type identification means identifies the vehicle type of an autonomous vehicle whose attribute is determined to have a large degree of deviation for a predetermined number or more types of behavior as the first vehicle type.

7. The information processing apparatus according to claim 4 or 5, wherein the vehicle type identification means identifies the first vehicle type that is determined to have a large degree of deviation for each of the types of behavior.

8. The information processing apparatus according to claim 4 or 5, wherein the presenting means performs processing to present information indicating the identified first vehicle type and the first behavior for which the degree of deviation is determined to be large.

9. An information processing device according to claim 4 or 5, further comprising: an action identification means for identifying a first action that is specific to an autonomous vehicle of the first model in a first behavior in which the degree of deviation is determined to be large for the first model.

10. The information processing apparatus according to claim 9, wherein the operation identification means identifies an operation as the first operation when the frequency of an operation performed by an autonomous vehicle of the first type in the first behavior is higher than the frequency of such operation performed by a vehicle with attributes corresponding to the overall driving data.

11. The information processing apparatus according to claim 9, wherein the presentation means performs processing to present information indicating a specified first vehicle type, a first behavior relating to the first vehicle type, and a first operation relating to the first vehicle type.

12. The information processing apparatus according to claim 1, wherein the attributes include manual-driving vehicles and each type of autonomous vehicle, the deviation degree determination means determines for each of the points whether the deviation degree is large between the overall driving data relating to at least two attributes of vehicles, including the attributes of manual-driving vehicles, and the driving data relating to the attributes of at least two types of autonomous vehicles, using predetermined determination criteria, and further comprises point identification means for identifying the points where the deviation degree is determined to be large as first points, and the presentation means performs processing for presenting information relating to the first points.

13. An information processing method comprising: acquiring images obtained by photographing one or more locations where a vehicle can travel; identifying attributes related to autonomous driving of the vehicles shown in the images; acquiring driving data indicating the behavior of the vehicles at the locations for each of at least two vehicle attributes; determining, using predetermined criteria, whether there is a large discrepancy between the overall driving data, which is the driving data for at least two vehicle attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle; identifying the type of autonomous vehicle for the attribute determined to have a large discrepancy as the first type of vehicle; and performing processing to present information regarding the identified first type of vehicle.

14. A non-temporary computer-readable medium containing a program that causes a computer to execute the following steps: acquiring images obtained by photographing one or more locations where a vehicle can travel; identifying attributes related to autonomous driving of the vehicles shown in the images; acquiring driving data indicating the behavior of the vehicles at the locations for each of at least two vehicle attributes; determining, using predetermined criteria, whether there is a large discrepancy between overall driving data, which is driving data for vehicles with at least two attributes, and the driving data for each attribute corresponding to the type of autonomous vehicle; identifying the type of autonomous vehicle for the attribute determined to have a large discrepancy as a first type of vehicle; and performing processing to present information regarding the identified first type of vehicle.

Images