INFORMATION CAPTURE DEVICE, LEARNING DEVICE AND INFORMATION CAPTURE PROCEDURES

Abstract

An information acquisition device (100) comprises: an image acquisition information acquisition unit (102) for acquiring image acquisition information of a test subject (P1) placed on a mounting surface (B1a); and a wakefulness information acquisition unit (104) for acquiring information regarding a level of wakefulness of the test subject placed on the mounting surface, wherein the image acquisition information acquisition unit acquires the image acquisition information, comprising a face of the test subject placed on the mounting surface, moving between a first position in which an angle between the mounting surface and a horizontal plane is a first angle, and a second position in which the angle between the mounting surface and the horizontal plane is a second angle.

Description

TECHNICAL AREA

[0001] The present disclosure relates to an information acquisition device, a learning device and an information acquisition method. STATE OF THE ART

[0002] Conventionally, a device has been disclosed that determines the level of alertness of a user using a trained model that was trained with biological information and signal data of the biological information as training data (see patent literature 1). QUOTE LIST PATENT LITERATURE

[0003] Patent Literature 1: JP 2021-033748 A SUMMARY OF THE INVENTIONAL PROBLEM

[0004] In general, it is desirable that the trained model used in the device according to patent literature 1 be trained using a sufficient amount of training data. Therefore, there is a need for a device capable of efficiently acquiring information about the level of alertness of a test subject.

[0005] The present disclosure solves the above-mentioned problem and aims to provide an information acquisition device, a learning device and an information acquisition procedure that are capable of acquiring information about the level of alertness of a test subject. SOLUTION TO THE PROBLEM

[0006] An information acquisition device according to the present disclosure comprises: an image acquisition information acquisition unit for acquiring image acquisition information of a test subject placed on a mounting surface; and a wakefulness information acquisition unit for acquiring information regarding a level of wakefulness of the test subject placed on the mounting surface, wherein the image acquisition information acquisition unit acquires the image acquisition information comprising a face of the test subject placed on the mounting surface, which moves between a first position in which an angle between the mounting surface and a horizontal plane is a first angle, and a second position in which the angle between the mounting surface and the horizontal plane is a second angle. ADVANTAGEOUS EFFECTS OF THE INVENTION

[0007] According to the present disclosure, information regarding the level of alertness of a test subject can be collected. BRIEF DESCRIPTION OF THE DRAWINGS Fig. Figure 1 is a side view illustrating an information collection system according to a first embodiment. Fig. Figure 2 is a side view illustrating the information collection system according to the first embodiment. Fig. Figure 3 is a block diagram illustrating the information collection system according to the first embodiment. Fig. Figure 4 is a block diagram illustrating an example of a hardware configuration of an information gathering device according to the first embodiment. Fig. Figure 5 is a block diagram showing an example of a hardware configuration of the information gathering device according to the first embodiment. Fig. Figure 6 is a flowchart showing a procedure for collecting information using the information collection system according to the first embodiment. Fig. Figure 7 is a block diagram showing an example of a hardware configuration for estimating the physical state of the information gathering device according to the second embodiment. Fig. Figure 8 is a schematic diagram illustrating a neural network used in a learning unit according to the second embodiment. Fig. Figure 9 is a flowchart illustrating the processing associated with the generation of a trained model, which is performed by the body state estimation device according to the second embodiment. Fig. Figure 10 is a flowchart illustrating the processing associated with the estimation of a level of alertness performed by the body state estimation device according to the second embodiment. Fig. Figure 11 is a block diagram illustrating a body condition estimation system according to a third embodiment. Fig. Figure 12 is a flowchart illustrating the processing associated with setting a threshold value, which is performed by the body condition estimation device according to the third embodiment. Fig. Figure 13 is a flowchart illustrating the processing associated with the estimation of a level of alertness performed by the body state estimation device according to the third embodiment. DESCRIPTION OF THE EXECUTION FORMS

[0008] The embodiments according to the present disclosure are described in detail below with reference to the drawings. First embodiment

[0009] First, an information collection system 1A according to the first embodiment is described with reference to Fig. 1 described. Fig. Figure 1 is a side view illustrating the information collection system 1A according to the first embodiment. The information collection system 1A is a system for collecting information, comprising biological information, from a test person (or test subject).

[0010] As in Fig. As illustrated in Figure 1, the information collection system 1A according to the first embodiment comprises an assembly table B1 (or assembly platform B1) on which (or which) a test subject P1 is placed, an image acquisition unit C1 that images the test subject P1, a sensor S1 for capturing biological information from the test subject P1, a drive unit M1 that drives the assembly table B1, an input unit N1 and an information collection device.

[0011] The assembly table B1 has an assembly surface B1a on which the test subject P1 is placed. As for example in Fig. As illustrated in Figure 1, the assembly table B1 is configured as a bed and designed such that the test subject P1 can be mounted in a state where the test subject P1 lies on the assembly surface B1a in a position where the flat assembly surface B1a and a horizontal plane are substantially parallel. Furthermore, the assembly table B1 is mounted so that it can rotate about a pivot axis (not shown) arranged along the assembly surface B1a. It should be noted that in the first embodiment, a position of the assembly surface B1a in which the assembly surface B1a and the horizontal plane are substantially parallel is also referred to as the first position. Furthermore, in the first embodiment, a position of the assembly surface B1a in which the assembly surface B1a intersects the horizontal plane is also referred to as the second position.

[0012] The image acquisition unit C1 captures an image of the test subject P1 placed on the mounting surface B1a. For example, the image acquisition unit C1 comprises a camera with an image sensor, such as a charge-coupled device (CCD) or a complementary metal-oxide semiconductor (CMOS). The image acquisition unit C1 is held on the mounting table B1 in such a way that its position relative to the mounting surface B1a is fixed. Thus, the image acquisition unit C1 can image a specific area of ​​the mounting surface B1a regardless of the rotational position of the mounting table B1. In other words, the image acquisition unit C1 can image a specific area such that the relative position between the mounting surface B1a and a center line L0 of the capture area remains constant regardless of the rotational position of the mounting table B1.For example, the image acquisition unit C1 is positioned such that the face of test subject P1, placed on the assembly table B1, is encompassed within a recording area L1, regardless of the rotational position of the assembly table B1. In other words, the image acquisition unit C1 captures image information of test subject P1, placed on the assembly surface B1a, such that image information captured in a state where the assembly surface B1a is in the second position and image information captured in a state where the assembly surface B1a is in the first position are image information in the same area with respect to the assembly surface B1a.It should be noted that the image acquisition unit C1 can be configured to capture the image information of the test subject P1 placed on the mounting surface B1a such that the image information captured in a state where the mounting surface B1a is in the second position, or in a state where the mounting surface is moving from the first position to the second position, and the image information captured in a state where the mounting surface B1a is in the first position, or in a state where the mounting surface is moving from the second position to the first position, is image information in the same area with respect to the mounting surface B1a.

[0013] In particular, the image acquisition unit C1 is positioned where the test subject P1 can be viewed from below (in Fig. (Illustrated in Figure 1 on the left) the test subject P1, placed on the assembly table B1, can be imaged in such a way that the face of the test subject P1, regardless of the rotational position of the assembly table B1, encompasses the recording area L1. The image acquisition unit C1 outputs the captured image information. It should be noted that the image acquisition unit C1 can be a visible light camera, which detects visible light, or an infrared camera, which detects infrared rays.

[0014] Sensor S1 outputs a signal corresponding to the biological activity of test subject P1. For example, sensor S1 is positioned so that it is in contact with test subject P1, who is placed on the assembly table B1. It detects the biological activity of test subject P1 and outputs a signal corresponding to this biological activity as biological information about test subject P1. Specifically, sensor S1 detects the heartbeat of test subject P1 and outputs a signal corresponding to this biological information.It should be noted that the biological information output by sensor S1 to test subject P1 is not limited to the heartbeat, but could, for example, include an electrocardiogram waveform, blood pressure, brain waves, blood oxygen saturation (SpO2), body temperature, skin luminosity, skin moisture content, or similar data. It could be any one of these pieces of information or a combination of many of them. Furthermore, sensor S1 could be an electrode for detecting an electric current flowing through the human body, a sensor for detecting infrared radiation, a sensor for detecting radio waves, or a sensor that utilizes another physical phenomenon.

[0015] The drive unit M1 drives the assembly table B1 so that the assembly table B1 rotates about the axis of rotation. For example, the drive unit M1 comprises a motor and a delay motor and rotates the assembly table B1 by supplying power and a control signal from the information gathering device 100. In other words, the drive unit M1 operates the assembly table B1 so that the assembly surface B1a moves between the first position, in which the assembly surface B1a and the horizontal plane are essentially parallel, and the second position, in which the assembly surface B1a and the horizontal plane intersect.In other words, the drive unit M1 operates the assembly table B1 such that the assembly surface B1a moves between the first position, in which the angle between the assembly surface B1a and the horizontal plane forms a first angle, and the second position, in which the angle between the assembly surface B1a and the horizontal plane forms a second angle.

[0016] Fig. Figure 2 is a side view illustrating the information gathering system 1A in a state where the assembly table B1 is rotated and the assembly surface B1a is in the second position. In this state, the test subject P1 is placed on the assembly surface B1a in an upright position with their head positioned higher than their feet. As described above, the information gathering system 1A is configured to perform a head-up tilt test by inducing reflex syncope in the test subject P1. This is achieved by moving the test subject P1, placed on the assembly table B1, from a supine to an upright position and observing the presence or absence of reflex syncope.

[0017] The input unit N1 receives an input command from an operator of the information collection system 1A. For example, the input unit N1 receives an input command to enter information regarding the level of consciousness of test subject P1 placed on the assembly surface B1a. In particular, the input unit N1 receives an input command indicating the presence or absence of unconsciousness of test subject P1 placed on the assembly surface B1a. For example, the input unit N1 includes an input device such as a switch, a touch panel, a keyboard, and a mouse that receives an input command. The input unit N1 outputs a signal corresponding to the input command. For example, if it is determined that test subject P1 has unconscious, the operator performs an input command on the input unit N1 indicating that test subject P1 has unconscious.

[0018] Fig. Figure 3 is a block diagram representing the information collection system 1A according to the first embodiment. As in Fig. As illustrated in Figure 3, the information gathering device 100, as an information acquisition device, comprises a control unit 101, an image acquisition information acquisition unit 102, a biological information acquisition unit 103, and a wakefulness information acquisition unit 104, and is electrically connected to the drive unit M1, the imaging unit C1, the sensor S1, and the input unit N1. The control unit 101 controls the drive of the drive unit M1 by outputting a control signal to the drive unit M1. The image acquisition information acquisition unit 102 acquires image acquisition information from the image acquisition unit C1. In particular, the image acquisition unit C1 acquires image acquisition information of the test subject P1 placed on the mounting surface B1a. The biological information acquisition unit 103 acquires biological information of the test subject P1 from the sensor S1.In other words, the Biology Information Acquisition Unit 103 acquires the biological information of test subject P1 independently of the image acquisition information acquired by the Image Acquisition Information Acquisition Unit 102. The Wakefulness Information Acquisition Unit 104 acquires a signal from the Input Unit N1. For example, the Wakefulness Information Acquisition Unit 104 acquires information corresponding to an input operation in which the operator enters information about the wakefulness level of test subject P1. In other words, the Wakefulness Information Acquisition Unit 104 acquires information about the wakefulness level of test subject P1.

[0019] Next, a hardware configuration of the information collection device 100 according to the first embodiment is described with reference to the Fig. 4 and Fig. 5 described. Fig. Figure 4 is a block diagram illustrating an example of a hardware configuration of the information gathering device 100 according to the first embodiment, and Fig. 5 is a block diagram showing an example of a hardware configuration that differs from the one in Fig. The hardware configuration of the information collection device 100 shown in Figure 4 differs according to the first embodiment. For example, as shown in Figure 4. Fig. As illustrated in Figure 4, the information gathering device 100 comprises a processor 100a, a memory 100b, and an I / O port 100c, and is configured such that the processor 100a reads and executes a program stored in memory 100b. Memory 100b can be, for example, non-volatile or volatile semiconductor memory such as RAM, ROM, flash memory, EPROM, or EEPROM. Additionally, memory 100b can be a magnetic disk, a flexible disk, an optical disk, a compact disc, a mini-disc, a DVD, or the like. Furthermore, memory 100b can be an HDD or an SSD.

[0020] Furthermore, the information collection device 100 includes, for example, as shown in Fig. Figure 5 illustrates a processing circuit 100d and the I / O port 100c, which are dedicated hardware. The processing circuit 100d comprises, for example, a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, a low-segregated integration (LSI) system, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or a combination thereof. Each function of the information gathering device 100 is implemented by the processor 100a or the processing circuit 100d, which are dedicated hardware that executes a program that is software, firmware, or a combination of software and firmware.

[0021] Next, with reference to Fig. 6 a method for collecting information using the information collection system 1A according to the first embodiment is described. Fig. Figure 6 is a flowchart illustrating a procedure for collecting information using the information collection system 1A according to the first embodiment. As shown in Fig. As illustrated in Figure 6, when collecting information, the operator first checks whether the assembly surface B1a of the assembly table B1 is parallel to the horizontal plane and places the test subject P1 on the assembly surface B1a in a prone position (step ST1). In other words, the operator places the test subject P1 on the assembly surface B1a, which is in the first position. In this step, the operator positions the test subject P1 on the assembly surface B1a in such a way that the test subject P1 is in a posture from which information can be captured by the information collection device 100. For example, in this step, the operator secures the test subject P1 to the assembly surface B1a with a strap (not shown).

[0022] When step ST1 is executed, the operator attaches sensor S1 to test subject P1 and begins the acquisition of biological information from test subject P1 by the information collection device 100 (step ST2). In this step, the information collection device 100 begins acquiring biological information from test subject P1 via sensor S1 while it is positioned on the mounting surface B1a in the first position. For example, the information collection device 100, using sensor S1, acquires biological information from test subject P1 that changes over time.

[0023] When the information collection device 100 performs the processing of step ST2, the operator begins the acquisition of image information of test subject P1 by the information collection device 100 (step ST3). In this step, the information collection device 100 begins acquiring image information of test subject P1, which is being imaged by the image acquisition unit C1. For example, in this step, the operator activates the image acquisition unit C1 and the information collection device 100, and the image acquisition unit C1 begins acquiring image information of test subject P1. Furthermore, the information collection device 100 acquires, for example, image information of test subject P1 as motion image information changing over time by the image acquisition unit C1. When the processing of step ST3 is performed, the information collection device 100 also begins timing by the control unit 101.

[0024] The information collection device 100 determines whether a predetermined time 1 has elapsed after the execution of step ST3 (step ST4). During this processing, the control unit 101 compares the predefined predetermined time 1 with the clocked time and determines whether the time since the execution of step ST3 has exceeded the predetermined time 1. The predetermined time 1 is, for example, a preset time that is greater than or equal to 5 minutes and less than or equal to 10 minutes. If the predetermined time 1 has not elapsed after the processing in step ST3 (NO in step ST4), the information collection device 100 waits before proceeding with the subsequent processing.

[0025] When the predetermined time 1 has elapsed since the execution of step ST3 by the information gathering device 100 (YES in step ST4), the information gathering device 100 changes the inclination of the assembly table B1 to the second angle (step ST5). During this processing, the control unit 101 controls the drive unit M1 so that the assembly table B1 rotates such that the angle between the assembly surface B1a and the horizontal plane forms the second angle, and the test subject P1 is in an upright position. For example, the second angle is a preset value that is greater than or equal to 60 degrees and less than or equal to 80 degrees.

[0026] When the information-gathering device performs step ST5, the operator observes test subject P1 to determine whether or not test subject P1 has lost consciousness (step ST6). In this step, the operator determines whether test subject P1 has lost consciousness based on observations such as test subject P1's appearance, their response to an external stimulus, and biological information obtained from sensor S1. For example, the operator determines whether test subject P1 has lost consciousness based on test subject P1's weakness, the disappearance of their response to an external stimulus, a decrease in test subject P1's vital signs, and similar factors.

[0027] If, in step ST6, it is determined that test subject P1 has not lost consciousness (NO in step ST6), i.e., if, in step ST6, the operator does not perform any input operation that indicates test subject P1 has lost consciousness on the input unit N1, the information gathering device 100 determines whether a predetermined time 2 has elapsed after the processing of step ST3 (step ST7). During this processing, the control unit 101 compares the predetermined time 2 with the clocked time and determines whether the time after the processing of step ST3 has exceeded the predetermined time 2. For example, the predetermined time 2 is a preset time that is equal to or greater than 50 minutes and equal to or less than 60 minutes.If the predetermined time 2 has not elapsed after processing in step ST3 (NO in step ST7), the information gathering device 100 returns to processing in step ST6.

[0028] If, during the processing of step ST6, it is determined that test subject P1 has lost consciousness (YES in step ST6), and if the clocked time in the processing of step ST7 has exceeded the predetermined time 2 (YES in step ST7), the information collection device 100 changes the inclination of the assembly table B1 to the first angle (step ST8). In this step, the operator performs an input operation indicating that test subject P1 has lost consciousness on the input unit N1. Thus, the consciousness information acquisition unit 104 receives information about the presence or absence of unconsciousness of test subject P1, who is positioned on the assembly surface B1a. Furthermore, in this step, the control unit 101 controls the drive unit M1 so that the assembly table B1 rotates such that the angle between the assembly surface B1a and the horizontal plane forms the first angle, i.e.,The test subject P1 is in a supine position. In other words, the control unit 101 drives the drive unit M1 to rotate the setup table B1 so that the angle between the support surface B1a and the horizontal plane forms the first angle, i.e., the test subject P1 is in a supine position. Furthermore, when the processing of step ST8 is carried out, the information gathering device 100 begins timing via the control unit 101.

[0029] When the information gathering device 100 performs step ST8, it determines whether a predetermined time 3 has elapsed since the execution of step ST8. During this processing, the control unit 101 compares the predetermined time 3 with the clocked time and determines whether the time since the execution of step ST8 exceeds the predetermined time 3. For example, the predetermined time 3 is a preset time that is equal to or greater than 5 minutes and equal to or less than 10 minutes. If the predetermined time 3 has not elapsed after the processing in step ST8 (NO in step ST9), the information gathering device 100 waits before proceeding with the subsequent processing.

[0030] When the predetermined time 3 has elapsed after the processing of step ST8 by the information gathering device 100 (YES in step ST9), the information gathering device 100 terminates the acquisition of image acquisition information by the image acquisition unit C1 (step ST10) and terminates the acquisition of biological information by the sensor S1 (step ST11). When the information gathering device 100 completes steps ST10 and ST11, the information acquisition of test subject P1 using the information gathering system 1A is complete. By performing steps ST1 to ST11 for a large number of test subjects using the information gathering system 1A, the operator acquires information that is available as training data for generating a trained model to determine the level of alertness of the person from image acquisition information of the person for whom the level of alertness is to be determined.

[0031] For example, if characteristic information is contained in the image acquisition information of test subject P1, and if test subject P1 is determined to be unconscious in the processing of step ST6, it is possible to generate a trained model to predict unconsciousness of the subject from the image acquisition information of the subject by generating a trained model using training data that includes data to which a label of unconsciousness, encompassing the characteristic information, is assigned, and data to which a label of non-unconsciousness, not encompassing the characteristic information, is assigned.

[0032] Furthermore, if, for example, the image acquisition information of test subject P1 contains a characteristic after a fainting spell, after it was determined in step ST6 that test subject P1 had fainted, it is possible to generate a trained model to determine whether the subject has fainted or not from the subject's image acquisition information by creating a trained model using training data that includes data to which a label "fainted" with the characteristic information has been assigned, and data to which a label "not fainted" without the characteristic information has been assigned.

[0033] Furthermore, for example, if characteristic information is comprehensively contained in a change in the image acquisition information of test subject P1 before and after fainting, and if test subject P1 is determined to be unconscious in the processing of step ST6, it is possible to generate a trained model to predict whether the person has fainted or not from the person's image acquisition information by creating a trained model using training data that includes data to which a label of fainting including the characteristic information is assigned, and data to which a label of not fainting without the characteristic information is assigned.It should be noted that the operator can perform steps ST1 to ST11 multiple times for a test subject using the information collection system 1A to gather information that is available as training data for generating a trained model to determine the level of alertness of the test subject from the image acquisition information of the test subject for whom the level of alertness is to be determined.

[0034] As described above, the information collection system 1A according to the first embodiment comprises the image acquisition information acquisition unit 102 for acquiring image acquisition information of the test subject P1 placed on the mounting surface B1a and the alertness information acquisition unit 104 for acquiring information on the level of alertness of the test subject placed on the mounting surface B1a, wherein the image acquisition information acquisition unit 102 acquires image acquisition information that includes the face of the test subject P1 placed on the mounting surface B1a, which moves between the first position, where the angle between the mounting surface B1a and the horizontal plane is the first angle, and the second position, where the angle between the mounting surface B1a and the horizontal plane is the second angle.As described above, the information collection system 1A can induce a change in the level of alertness of test subject P1 by changing the angle between the mounting surface B1a, on which test subject P1 is placed, and the horizontal plane, and obtain the image acquisition information of test subject P1 when the level of alertness changes, and thus it is possible to collect information that can be used as training data for generating a trained model that determines the level of alertness of the person based on the image acquisition information.

[0035] For example, in a vehicle like a car that includes an image capture unit depicting an occupant, if the occupant's health status is to be estimated using image capture information from the unit, it's conceivable to perform the estimation using a trained model generated from image capture information of the occupant as training data should the health status change. However, it is rare for the occupant's physical condition to change, for example, fainting, when the occupant actually enters the vehicle, and it is difficult to collect a sufficient amount of image capture information.The information collection system 1A according to the first embodiment can induce a change in the level of alertness of the test subject in an environment simulating a state when entering a vehicle, and capture image capture information of the test subject when the level of alertness changes, without capturing image capture information of an occupant when actually entering the vehicle.

[0036] For example, if a trained model for estimating a change in a driver's alertness level is generated based on image information from a vehicle camera depicting the driver, it is possible to acquire information in a state where the position and recording area of ​​the image acquisition unit C1 of the information gathering system 1A are pre-adjusted to the position and recording area of ​​the vehicle's in-vehicle camera using the trained model. The reliability of the trained model can be improved by generating the trained model using the information acquired in this way.In particular, if the vehicle camera of the vehicle, which the trained model uses, is arranged so that it images the driver from diagonally front left and diagonally downwards from the driver, it is possible to improve the reliability of the generated trained model by holding the image acquisition unit C1 of the information gathering system 1A on the mounting table B1 in such a way that it images the test subject P1 from diagonally front left and diagonally downwards from the test subject P1, regardless of the angle between the mounting surface B1a and the horizontal plane.Furthermore, if, for example, the vehicle camera of the vehicle using the trained model is positioned so that it images the driver diagonally from the front right and diagonally above the driver, it is possible to improve the reliability of the generated trained model by holding the image acquisition unit C1 of the information gathering system 1A on the mounting table B1 in such a way that it captures the test subject P1 diagonally from the front left and diagonally downwards, regardless of the angle between the mounting surface B1a and the horizontal plane.

[0037] Furthermore, the information collection system 1A according to the first embodiment comprises the biological information acquisition unit 103, which receives the biological information of the test subject P1 via the sensor S1. For example, if the set of features relating to the biological information of the test subject P1 is extracted from the image acquisition information, and the trained model, which determines the level of alertness of the test subject from the image acquisition information, is generated using the set of features extracted from the image acquisition information as training data, it is possible to improve the accuracy of extracting the set of features, the reliability of extracting the set of features from the image acquisition information, and the efficiency of generating the training data.In particular, if the feature set relating to the heartbeat of test subject P1 is extracted from the image acquisition information, and the trained model, which determines the level of alertness of the test subject from the image acquisition information, is generated using the feature set extracted from the image acquisition information as training data, it is possible to improve the extraction accuracy when extracting the feature set relating to the heartbeat of test subject P1 from the image acquisition information by the sensor S1 together with the image acquisition information, and to improve the efficiency in generating the training data as well as the reliability of the generated trained model.

[0038] It should be noted that the biological information of test subject P1, the set of features of which is extracted from the image acquisition information, and the biological information acquired by sensor S1 preferably comprise, but are not limited to, information that is identical or corresponding to each other. Both the biological information of test subject P1, the set of features of which is extracted from the image acquisition information, and the biological information acquired by the sensor need only be information that influences the level of alertness of test subject P1. For example, the biological information of test subject P1, the set of features of which is extracted from the image acquisition information, could be a change in the expression of test subject P1, and the biological information acquired by sensor S1 could be information relating to the heartbeat of test subject P1.Furthermore, identical or corresponding information is closely related information, and examples include a heartbeat and an RR interval, an average blood pressure and a diastolic or systolic blood pressure, as well as skin color and body temperature.

[0039] The image acquisition unit 102 captures image information, including the face of test subject P1, which is mounted on the mounting surface B1a, located at least in the second position. Since the change in the level of alertness, including the loss of consciousness of test subject P1 in the head-up tilt test, frequently occurs when test subject P1 is in an upright position and is often accompanied by a change in appearance in the facial area, the image acquisition unit 102 captures the image information that includes the face of test subject P1 in an upright position. This allows for the easy capture of image information containing characteristic information regarding the change in the level of alertness and improves the efficiency of training data generation as well as the reliability of the resulting trained model.

[0040] In the first embodiment, the information collection device 100 is configured such that the wakefulness information acquisition unit 104 acquires information indicating that the test subject P1 has become unconscious through an input operation at the input unit N1, but is not limited to this.The alertness information acquisition unit only needs to be configured to acquire information about the level of alertness of the test subject P1 placed on the mounting surface B1a, and, for example, the alertness information acquisition unit can be configured to acquire information indicating a state in which consciousness has not been completely lost, but the level of alertness has decreased due to an input operation on the input unit, or it can be configured to acquire information indicating a state in which the level of alertness has decreased without unconsciousness, or a state in which there are signs of a decrease in the level of alertness.

[0041] Furthermore, in the first embodiment, the information acquisition device 100 is configured such that the image acquisition information acquisition unit 102 acquires the image acquisition information of the test subject P1 as time-changing motion image information from the imaging unit C1, but is not limited to this. The image acquisition information acquisition unit need only be configured to acquire the image acquisition information of the test subject P1 placed on the mounting surface B1a, and, for example, the image acquisition information acquisition unit can be configured to acquire image acquisition information that is still image information from the imaging unit C1 at a specific single time (time point), or it can be configured to acquire image acquisition information that is still image information at multiple time points when information is acquired from a test subject.

[0042] Furthermore, in the first embodiment, the information acquisition device 100 is configured such that the image acquisition information acquisition unit 102 acquires the image acquisition information of the test subject P1 from a supine position through an upright position and back to a supine position by means of the imaging unit C1, but is not limited to this. The image acquisition information acquisition unit only needs to be configured such that it acquires image acquisition information, comprising the face of the test subject P1, which is placed on the mounting surface B1a, at least at every point in time after the test subject P1 has changed from a supine to an upright position.In other words, the image acquisition information acquisition unit only needs to be configured to acquire the image acquisition information, comprising the face of test subject P1, which is positioned on mounting surface B1a, at all times, at least after the position of mounting surface B1a has changed from the first position to the second position. For example, the image acquisition information acquisition unit can begin acquiring the image acquisition information from image acquisition unit C1 when the position of mounting surface B1a changes from the first position to the second mounting surface, or it can be configured to stop acquiring the image acquisition information when the position of mounting surface B1a changes from the second position to the first position.

[0043] Furthermore, in the first embodiment, the information collection device 100 is configured such that the control unit 101 controls the angle between the mounting surface B1a and the horizontal plane such that the mounting surface B1a moves between the first position, in which the angle between the mounting surface B1a and the horizontal plane is substantially parallel; the first position, in which the angle between the mounting surface B1a and the horizontal plane is substantially parallel, so that the test subject P1 is in a prone position; and the second position, in which the mounting surface B1a intersects the horizontal plane such that the test subject P1 is in an upright position. However, the possibilities are not limited to these.It is sufficient if the control unit controls the rotational position of the mounting table B1 such that a change in the level of alertness of the test subject P1 is caused by a change in the body posture of the test subject P1 placed on the mounting surface B1a. For example, the mounting surface on which the test subject P1 is placed need not be a flat surface, but can be formed from a multitude of overlapping surfaces, from a curved surface, or from a soft material that can be easily deformed by an external force. In particular, the control unit can be configured to control the position of the mounting surface between the first position, in which the test subject P1 is placed in a supine position, and the second position, in which the test subject P1 is placed in a seated position.Additionally, the mounting surface can include a first surface that comes into contact with the back of the test subject P1 in a supine position to support the test subject P1, and a second surface that comes into contact with the sole of the foot of the test subject P1 in an upright position to support the test subject P1.

[0044] Furthermore, in the first embodiment, the information collection device 100 is configured such that the biological information acquisition unit 103 acquires the biological information of test subject P1 via an input signal from sensor S1, but is not limited to this. For example, the biological information acquisition unit can be configured to acquire the biological information of test subject P1 via the input signal from input unit N1, or it can be configured to acquire the biological information of test subject P1 via sensor S1 and the input signal from input unit N1.In particular, the biological information acquisition unit can be configured to acquire the biological information of test subject P1 by the operator performing an input operation at the input unit relating to the biological information of test subject P1 acquired by a sensor or by the operator's observation of test subject P1 (not illustrated). Examples of biological information of test subject P1 obtained by the operator's observation of test subject P1 comprehensively include features of test subject P1's appearance, such as information about expression, degree of eye opening, complexion, pupil diameter, degree of mouth opening, and the presence or absence of fatigue.In addition, the information entered by the input unit into the information collection device 100 is not limited to the above-mentioned information, but may also include other information about the test subject, for example, the age, medical history, height, weight, gender, and the like of the test subject.

[0045] Furthermore, in the first embodiment of the information acquisition system 1A, the image acquisition unit C1 is held on the assembly table B1 such that the relative position between the image acquisition unit C1 and the assembly surface B1a remains constant, but this is not limited to the first embodiment. It is sufficient if the information acquisition system 1A is configured to acquire image acquisition information in a specific area of ​​the assembly surface B1a regardless of the position of the assembly surface B1a. For example, the information acquisition system can include an image position control unit (not shown) that controls the position of the image acquisition unit so that the image acquisition unit moves with the rotation of the assembly table and can be configured to acquire image acquisition information in a specific area of ​​the assembly surface B1a regardless of the position of the assembly surface B1a.

[0046] Second embodiment Next, a body state estimation system 2A according to the second embodiment is described with reference to the Fig. 7, Fig. 8, Fig. 9 to Fig. 10 described. Fig. Figure 7 is a block diagram illustrating the body state estimation system 2A according to the second embodiment. As shown in Fig. As illustrated in Figure 7, the body state estimation system 2A according to the second embodiment comprises the information collection device 100 according to the first embodiment and a body state estimation device 200. The description of the information collection device 100, which overlaps with that of the first embodiment, is omitted.

[0047] The body state estimation device 200, as a device for estimating the level of alertness and as a learning device, is a device that, using the information acquired by the information gathering device 100, generates a trained model and, using the trained model, estimates a person's level of alertness from image acquisition information of the person. As in Fig. As illustrated in Figure 7, the body state estimation device 200 comprises a feature sampling unit, a learning unit 202, a model storage unit 203 and a state estimation unit 204.

[0048] The feature extraction unit 201 extracts a feature set from image information and biological information of a test subject acquired by the information collection device 100. For example, the feature extraction unit 201 can be configured to extract one of the following features: an RR interval, a heartbeat, a diastolic blood pressure value, a systolic blood pressure value, an average blood pressure value, an eye opening degree, a mouth opening degree, movement of expressive muscles, a pupillary change rate, a luminance change rate, and the like from the image information and biological information of the test subject acquired by the information collection device 100, or it can be configured to extract several of these feature sets.

[0049] Learning Unit 202, acting as a trained model generation unit, generates a trained model using the feature set extracted by Feature Extraction Unit 201 and information about the test subject's level of alertness. This training data is linked to the image acquisition information and the biological information from which the feature set was extracted. A well-known supervised learning algorithm can be used as the training algorithm employed by Learning Unit 202 to generate the trained model. The following describes a case in which Learning Unit 202 uses a neural network as an example of the training algorithm.

[0050] Fig. Figure 8 is a schematic diagram illustrating a neural network used in learning unit 202 according to the second embodiment. For example, learning unit 202 learns the relationship between the input image acquisition information and the test subject's level of alertness with respect to the image acquisition information through so-called supervised learning according to the neural network model. Here, supervised learning refers to a process in which training data, consisting of a set of input and output data (labels), is provided to learning unit 202 as a learning device to learn features in the training data and derive an output from the input.

[0051] The neural network comprises an input layer containing a multitude of neurons, an intermediate layer (hidden layer) also containing a multitude of neurons, and an output layer containing a multitude of neurons. The intermediate layer can consist of one layer, two, or more layers. For example, a three-layer neural network, as in Fig. Figure 8 shows that when a large number of inputs are entered into the input layer (X1-X3), the value is multiplied by a weight W1 (w11-w16) and entered into the intermediate layer (Y1-Y2) as input, and the result is further multiplied by a weight W2 (w21-w26) and output by the output layer (Z1-Z3).

[0052] This output varies depending on the values ​​of weights W1 and W2. In the second embodiment, the neural network learns whether the person is in a state of reduced alertness or a state showing signs of reduced alertness through supervised learning based on training data generated from the feature set extracted by feature extraction unit 201. For example, the neural network learns from the training data generated from feature extraction unit 201 whether the test subject is unconscious or showing signs of unconsciousness.As described above, the neural network learns by adjusting the weights W1 and W2 so that the result output by the output layer, after the extracted feature set has been fed into the input layer, approximates the feature set that indicates a decrease in the person's level of alertness or a sign of a decrease in the person's level of alertness. The learning unit 202 generates a trained model by performing the learning described above and stores the generated trained model in the model storage unit 203.

[0053] The state estimation unit 204, acting as an inference unit, performs an inference regarding the person's level of alertness based on the trained model stored in the model storage unit 203 and the image acquisition information of the person for whom the level of alertness is to be estimated. For example, based on the trained model stored in the model storage unit 203 and the image acquisition information of the person for whom the level of alertness is to be estimated, the state estimation unit 204 estimates whether there is a state in which the level of alertness has decreased or a state in which there are signs of a decrease in the person's level of alertness.The feature sets used by the state estimation unit 204 for estimation using the feature set can be image information or biological information acquired by the information gathering device 100, or image information or biological information acquired by an image device or a biological information gathering device (not illustrated). The state estimation unit 204 outputs the estimation result.

[0054] The body state estimator 200 can comprise a processor, memory, and an I / O port and be configured such that the processor reads and executes a program stored in memory, or it can comprise a processing circuit and an I / O port, which are dedicated hardware, and be configured such that the processing circuit executes the program. Since the hardware configuration of the body state estimator 200 is similar to that of the information gathering device 100 according to the first embodiment, its description is omitted here.

[0055] Next, the processing performed by the body condition estimation device 200 according to the second embodiment is described with reference to the Fig. 9 and Fig. 10 will be carried out. Fig. Figure 9 is a flowchart illustrating the processing involved in generating a trained model, which is performed by the body state estimator 200 according to the second embodiment. At the start of the processing, the body state estimator 200 first acquires the biological information acquired by the information gathering device 100 (step ST21) and then acquires the image acquisition information acquired by the information gathering device 100 (step ST22). It should be noted that the body state estimator 200 can be configured to acquire information from the information gathering device 100 while electrically connected to it for information transmission, or it can be configured to acquire information from the information gathering device 100 via a recording medium.

[0056] After processing steps ST21 and ST22, the Body Condition Estimator 200 extracts a feature set from the acquired information (step ST23). It should be noted that during step ST23, the Body Condition Estimator 200 may be configured to extract the feature set relating to the test subject's biological information from the image data.For example, the body state estimator 200 can be configured to extract the feature set of image information corresponding to the biological information acquired by sensor S1, based on the image information and the biological information acquired by sensor S1, when the image information is acquired by the imaging unit C1, or to extract the feature set relating to the biological information of the test subject, based on a characteristic change in the appearance of the human body according to a known biological activity. The body state estimator 200 can use a known algorithm when extracting the feature set from a captured image.

[0057] When processing step ST23 is performed, the body state estimator 200 generates a trained model (step ST24) based on the extracted feature set and the information regarding the test subject's level of alertness, which is linked to the image information and the biological information from which the feature set was extracted. During this processing, the learning unit 202 can generate the trained model using a known machine training algorithm, such as the neural network described above. After processing step ST24, the body state estimator 200 stores the generated trained model in the model storage unit 203 and terminates processing (step ST25).

[0058] Next, the processing related to the level of alertness estimation performed by the body state estimator 200 according to the second embodiment is described, illustrating as an example a processing in which the body state estimator 200 estimates the level of alertness of the person based on image acquisition information and biological information acquired by the imaging device and the biological information acquisition device (not shown). Fig. Figure 10 is a flowchart illustrating the processing associated with the level of alertness estimation performed by the body state estimation device 200 according to the second embodiment. It should be noted that in the second embodiment, the state estimation unit 204 forms an image acquisition information acquisition unit that acquires image information, including an image of the person's face.

[0059] Initially, at the start of processing, the Body State Estimator 200 acquires biological information about the person from the biological information acquisition device (not shown) (step ST31) and acquires image acquisition information, comprising an image of the person's face, from the imaging device (not shown) (step ST32). For example, the Body State Estimator 200 acquires the biological information and the image information in a state where it is electrically connected to the image acquisition device and the biological information acquisition device (not illustrated) to enable information exchange.For example, the imaging device (not shown) is an imaging device that images an occupant of a vehicle, and the information acquisition device (not shown) is a sensor provided in the vehicle that captures biological information of the occupant.

[0060] After processing steps ST31 and ST32, the body state estimator 200 extracts a feature set from the acquired information (step ST33). During this processing, the feature set extraction unit 201 extracts a feature set from the acquired information according to an algorithm when the trained model is generated. For example, the feature set extraction unit 201 extracts a feature set relating to the biological information of the test subject from the acquired image information.

[0061] After processing step ST33, the body state estimator 200 accesses the information stored in the model storage unit 203 and reads the trained model (step ST34). It should be noted that if the learning unit 202 is configured to generate multiple trained models, the body state estimator 200 may be configured to select one of the trained models based on the extracted feature set.

[0062] When processing step ST34 is performed, the body state estimator 200 estimates the person's level of alertness using the state estimator 204 based on the extracted feature set and the trained model stored in the model storage unit 203 (step ST35). After processing step ST35, the body state estimator 200 outputs information about the estimated level of alertness of the person and terminates processing (step ST36).

[0063] As described above, the body state estimation system 2A according to the second embodiment comprises the model storage unit 203, which stores the trained model generated using the training data, including the feature set extracted from the image acquisition information, including the image of the test subject's face and information about the test subject's level of alertness, and the state estimation unit 204, which captures the image acquisition information, including the image of the test subject's face, wherein the state estimation unit 204 draws a conclusion regarding the test subject's level of alertness based on the trained model stored in the model storage unit 203 and the captured image acquisition information.Thus, it is possible to estimate the person's level of alertness based on the person's image information, and it is possible to manage the person's physical condition more easily than before.

[0064] Furthermore, according to the second embodiment, the body state estimator 200 comprises the feature extraction unit 201, which extracts a feature set from the image information acquired by the information collection device 100, and a training unit 202, which generates a trained model based on the information about the test subject's level of alertness acquired by the information collection device 100 and the feature set extracted by the feature extraction unit 201. Thus, the body state estimator 200 can generate the trained model based on a sufficient amount of information obtained from the information collection device 100, and the reliability of the trained model can be improved.

[0065] It should be noted that in the second embodiment, the body state estimation system 2A is configured to generate the trained model based on, but not limited to, the biological information of the test subject acquired by the sensor or the like and the extracted set of features, the set of features extracted from the image information of the test subject, and information about the level of alertness of the test subject.The system for estimating physical state only needs to be configured to generate the trained model based on at least the set of features extracted from the image information and information about the test subject's level of alertness. For example, the physical state estimation system 2A can be configured to not capture any biological information that does not depend on the image information, or it can be configured to estimate the test subject's level of alertness without capturing any biological information that does not depend on the image information. Third embodiment

[0066] Next, a body state estimation system 3A according to the third embodiment is described with reference to the Fig. 11, Fig. 12 to Fig. 13 described. Fig. Figure 11 is a block diagram illustrating the body state estimation system 3A according to the third embodiment. As shown in Fig. As illustrated in Figure 11, the body state estimation system 3A according to the third embodiment comprises the information gathering device 100 and the body state estimation device 300 according to the first embodiment. Although the body state estimation device 300 according to the third embodiment differs from the body state estimation device 200 according to the second embodiment in that a result of the body state estimation can be output without using a trained model, some configurations are similar to those of the body state estimation device 200 according to the second embodiment, and the configurations that are similar to those of the second embodiment are designated with the same reference numerals, and their description is omitted.

[0067] The body state estimation device 300, as a device for estimating the level of alertness and setting a threshold, is a device that estimates a person's level of alertness based on the image acquisition information of the person, using the information collected by the information gathering device 100. As in Fig. As illustrated in Figure 11, the device 300 for estimating physical condition comprises a feature quantity sampling unit 201, a threshold adjustment unit 302 and a determination unit 303.

[0068] The threshold adjustment unit 302, acting as a threshold setting unit, establishes a threshold for estimating a person's level of alertness using the feature sample taken from the feature sample unit 201. For example, if the feature sample taken from feature sample unit 201 is one indicating a heartbeat, and it is determined that a person's level of alertness has decreased when a change in heartbeat of 20 beats per minute or more is detected for a predetermined time (e.g., 10 seconds) based on the feature sample, the threshold adjustment unit 302 sets the threshold to 20 beats per minute. Furthermore, the threshold adjustment unit 302 adjusts the previously set threshold to a suitable value based on the feature sample taken from feature sample unit 201.For example, if the threshold for estimating the level of alertness is set to a value where estimating the level of alertness based on the quantity of feature taken from the feature sampling unit 201 and the information about the subject's level of alertness related to the information from which the feature quantity was taken is insufficient, the threshold adjustment unit 302 adjusts the threshold to improve the estimation of the level of alertness.

[0069] The unit of determination 303, acting as the estimation unit, determines the person's level of alertness using the threshold set by the threshold adjustment unit 302. For example, if the quantity of features extracted by the feature sampling unit 201 from the image information or the biological information of the test subject exceeds the threshold set by the threshold adjustment unit 302, the unit of determination 303 estimates that the test subject is in a state of decreased alertness or in a state showing signs of decreased alertness.The set of features used by the identification unit 303 for estimation can be image information or biological information acquired by the information gathering device 100, or image information or biological information acquired by the imaging device or the biological information acquisition unit (not shown). The identification unit 303 outputs the estimation result.

[0070] The body state estimator 300 can comprise a processor, memory, and an I / O port and be configured such that the processor reads and executes a program stored in memory, or it can comprise a processing circuit and an I / O port, which are dedicated hardware, and be configured such that the processing circuit executes the program. Since the hardware configuration of the body state estimator 300 is similar to that of the information gathering device 100 according to the first embodiment, its description is omitted here.

[0071] Next, the processing related to the threshold setting, which is carried out by the body condition estimation device 300 according to the third embodiment, is described, with reference to the Fig. 12 and Fig. 13. Reference is made to this. Fig. Figure 12 is a flowchart illustrating the processing associated with threshold setting performed by the body condition estimator 300 according to the third embodiment. Initially, at the start of processing, the body condition estimator 300 acquires the biological information collected by the information gathering device 100 (step ST21) and acquires the image acquisition information collected by the information gathering device 100 (step ST22). It should be noted that the body condition estimator 300 can be configured to acquire information from the information gathering device 100 while electrically connected to it for information transmission, or it can be configured to acquire information from the information gathering device 100 via a recording medium.

[0072] After processing steps ST21 and ST22, the body condition estimator 300 extracts a feature set from the acquired information (step ST23). After processing step ST23, the body condition estimator 300 sets a threshold for assessing the person's level of alertness (step ST44). During this processing, if the threshold was not predefined, the threshold adjustment unit 302 sets a new threshold; if the threshold was predefined, the threshold adjustment unit sets a new threshold by adjusting the threshold to a more appropriate value. After processing step ST44, the body condition estimator 300 stores a new threshold in the memory unit (not shown) (step ST45).

[0073] Next, the processing related to the estimation of the level of arousal performed by the body state estimator 300 according to the third embodiment is described by illustrating a processing in which the body state estimator 300 estimates the level of arousal of the person based on image acquisition information and biological information acquired by the imaging device and the biological information acquisition device (not shown). Fig.Figure 13 is a flowchart illustrating the processing associated with the level of awakening estimation performed by the body state estimation device 300 according to the third embodiment. It should be noted that in the third embodiment, the determination unit 303 is an image acquisition information acquisition unit that acquires image information, including an image of the person's face.

[0074] Initially, at the start of processing, the physical condition estimator 300 acquires biological information about the person from the biological information acquisition device (not shown) (step ST31) and acquires image acquisition information, including an image of the person's face, from the imaging device (not shown) (step ST32). For example, the physical condition estimator 300 acquires the biological information and the image information in a state where it is electrically connected to the image acquisition device and the biological information acquisition device (not shown) to enable information exchange.For example, the imaging device (not illustrated) is an imaging device that creates an image of a vehicle occupant, and the information acquisition device (not illustrated) is a sensor provided in the vehicle that captures biological information of the occupant.

[0075] After processing steps ST31 and ST32, the body condition assessment device 300 extracts a feature set from the acquired information (step ST33). During this processing, the feature set extraction unit 201 extracts a feature set from the acquired information according to a defined threshold. For example, the feature set extraction unit 201 extracts a feature set relating to the test subject's biological information from the acquired image information.

[0076] After processing step ST33, the body condition estimator 300 accesses information stored in a memory unit (not shown) and reads a defined threshold value (step ST54). It should be noted that if the threshold adjustment unit 302 is configured to set multiple threshold values, the body condition estimator 300 may be configured to select one of the threshold values ​​based on the sampled feature set.

[0077] When the processing of step ST54 is performed, the body state estimator 300 estimates the person's level of alertness using the determination unit 303 based on the extracted feature set and the threshold stored in the memory unit (step ST55). After processing step ST55, the body state estimator 300 outputs information about the estimated level of alertness of the person and terminates the processing (step ST56).

[0078] As described above, the body state estimation system 3A according to the third embodiment comprises the threshold adjustment unit 302, which sets a threshold based on the feature set extracted from the image acquisition information, including the image of the test subject's face, and information about the test subject's level of alertness; and the determination unit 303, which acquires the image acquisition information, including the image of the test subject's face. The determination unit 303 estimates the test subject's level of alertness based on the threshold set by the threshold adjustment unit 302 and the acquired image acquisition information. Thus, it is possible to estimate the person's level of alertness based on the image information of the person, and it is possible to manage the person's physical state more easily than before.

[0079] Furthermore, according to the third embodiment, the body state estimation device 300 comprises the feature extraction unit 201, which extracts a feature set from the image information acquired by the information collection device 100, and a threshold adjustment unit 302, which sets a threshold value based on the information about the test subject's level of alertness acquired by the information collection device 100 and the feature set extracted by the feature extraction unit 201. Thus, the body state estimation device 300 can estimate the person's level of alertness based on the threshold set on the basis of a sufficient amount of information acquired by the information collection device 100, thereby improving the reliability of the alertness level assessment.

[0080] It should be noted that in the present disclosure free combinations of the individual embodiments, modifications of any components of the individual embodiments or omissions of any components in the individual embodiments are possible. INDUSTRIAL APPLICABILITY

[0081] An information collection system according to the present disclosure can, for example, be used to manufacture a body condition assessment device for an occupant of a vehicle, such as a car. Furthermore, the body condition assessment device according to the present disclosure can, for example, be used in a vehicle capable of assessing the physical condition of an occupant. REFERENCE MARK LIST

[0082] 1A: Information gathering system, 2A: Body state estimation system, 3A: Body state estimation system, 100: Information gathering device, 101: Control unit, 102: Image information acquisition unit, 103: Biology information acquisition unit, 104: Wakefulness information acquisition unit, 200: Body state estimation device, 201: Feature quantity extraction unit, 202: Learning unit, 203: Model storage unit, 204: State estimation unit, 300: Body state estimation device, 302: Threshold adjustment unit, 303: Determination unit, B1: Assembly table, B1a: Assembly surface, C1: Image acquisition unit, L0: Center line, L1: Acquisition area, M1: Drive unit, N1: Input unit, P1: Test subject, S1: Sensor QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] JP 2021-033748 A

[0003]

Claims

[1] Information acquisition device, comprising: an image acquisition unit for acquiring image information from a test subject placed on a mounting surface; and a wakefulness information acquisition unit for capturing information regarding the level of wakefulness of the test subject placed on the mounting surface, wherein The image acquisition information acquisition unit acquires the image acquisition information, comprising a face of the test subject placed on the mounting surface, which moves between a first position in which an angle between the mounting surface and a horizontal plane is a first angle, and a second position in which the angle between the mounting surface and the horizontal plane is a second angle. [2] Information acquisition device according to claim 1, further comprising: a biological information acquisition unit for capturing biological information from the test subject placed on the mounting surface. [3] Information acquisition device according to claim 1 or 2, wherein the angle to a horizontal plane in the second position is greater than the angle to a horizontal plane in the first position, and The image acquisition information acquisition unit acquires the image acquisition information, which includes at least the face of the test subject placed on the mounting surface, at a specific time after a time at which the position of the mounting surface changes from the first position to the second position. [4] Information acquisition device according to any one of claims 1 to 3, wherein the image acquisition information acquisition unit acquires the image acquisition information of the test subject placed on the mounting surface in such a way, in a state in which the mounting surface is in the second position, in a state in which the mounting surface moves from the first position to the second position, in a state in which the mounting surface moves from the second position to the first position, and in a state in which the mounting surface is in the first position, that the image acquisition information acquired in the state in which the mounting surface is in the second position, or in the state in which the mounting surface moves from the first position to the second position, and that the image acquisition information acquired in the state in which the mounting surface is in the first position,or in the state where the mounting surface moves from the second position to the first position, the image acquisition information is in the same area with respect to the mounting surface. [5] Information acquisition device according to any one of claims 1 to 4, further comprising: a control unit for controlling the angle between the mounting surface and the horizontal plane in such a way that the mounting surface moves between the first position and the second position. [6] Learning device, comprehensive: an image acquisition information acquisition unit for capturing image information from a test subject placed on a mounting surface; a wakefulness information acquisition unit for recording information regarding the level of arousal of the test subject, which is placed on the mounting surface; a feature quantity extraction unit for extracting a feature quantity from the image acquisition information captured by the image acquisition information acquisition unit; and a trained model generation unit for generating a trained model based on information about the level of alertness of the test subject placed on the assembly surface, which was captured by the image acquisition information acquisition unit, and the feature set extracted by the feature set extraction unit, wherein The image acquisition information unit captures the image acquisition information, which includes a face of the test subject placed on the mounting surface, moving between a first position where the angle between the mounting surface and a horizontal plane is a first angle, and a second position where the angle between the mounting surface and the horizontal plane is a second angle. [7] Information acquisition method performed by a device comprising an image acquisition information acquisition unit and a wakefulness information acquisition unit, wherein the information acquisition method comprises the steps: Capture, by the image acquisition information acquisition unit, image acquisition information of a test subject placed on a mounting surface; and The wakefulness information acquisition unit captures information regarding the level of wakefulness of the test subject placed on the assembly surface, whereby The image acquisition information unit captures the image acquisition information, which includes a face of the test subject placed on the mounting surface, moving between a first position where the angle between the mounting surface and a horizontal plane is a first angle, and a second position where the angle between the mounting surface and the horizontal plane is a second angle.

Images