Psychological state estimation device, psychological state estimation method, and psychological state estimation program

Abstract

This provides a user interface for estimating an individual's psychological state using multiple types of measurement data and different classification models. [Solution] The psychological state estimation device includes: a first estimation unit that estimates the psychological state of a subject based on a first classification model using first data measured from the subject; a second estimation unit that estimates the psychological state of a subject based on a second classification model using second data measured from the subject; a conversion unit that converts the first estimation result estimated by the first estimation unit into the psychological state of the subject based on the second classification model; and an output unit that outputs information based on the first estimation result converted by the conversion unit and the second estimation result estimated by the second estimation unit, using a classification method corresponding to the second classification model.

Description

Technical Field

[0001] The present disclosure relates to a mental state estimation device, a mental state estimation method, and a mental state estimation program.

Background Art

[0002] Measurement data used for estimating the mental state of a subject can be broadly classified into measurement data obtained by measuring the physical changes of the subject, such as facial expressions, postures, voices, etc., and measurement data obtained by measuring the physiological changes of the subject, such as electroencephalograms, electrocardiograms, pulse waves, etc.

[0003] The measurement data obtained by measuring the physical changes of the subject includes facial image data, skeletal data, voice data, etc. Such measurement data is suitable for estimating the mental state based on an arbitrary emotion classification model (hereinafter sometimes referred to as a "classification model"), for example, a basic emotion model, but has characteristics that it is not suitable for estimating the mental state based on the Russell circumplex model.

[0004] On the other hand, the measurement data obtained by measuring the physiological changes of the subject includes electroencephalogram data, electrocardiogram data, pulse wave data, etc. Such measurement data is suitable for estimating the mental state based on an arbitrary classification model different from the basic emotion model, for example, the Russell circumplex model, but has characteristics that it is not suitable for estimating the mental state based on the basic emotion model.

[0005] Patent Document 1 below discloses an emotion control device having a control unit that determines the emotional state of a user as a composite state (emotional state 1 + emotional state 2), where the emotional state determined based on the first emotion classification means (temporarily referred to as "emotional state 1") and the emotional state based on the second emotion classification means different from the first emotion classification means (temporarily referred to as "emotional state 2").

[0006] On the other hand, a high rate of agreement between the estimated psychological state and the actual psychological state of the individual subject (hereinafter sometimes referred to as "estimation accuracy") is desirable. As a method to improve the estimation accuracy of psychological state, it is conceivable to use a psychological state estimation program to compare and analyze the psychological states estimated from multiple types of measurement data. An example of such a psychological state estimation program is a multimodal AI equipped with a trained model that has learned the relationship between the measurement data and psychological state of multiple individuals. In particular, when the amount of measurement data becomes enormous in order to improve estimation accuracy, the above-mentioned multimodal AI is suitable from the standpoint of processing power. [Prior art documents] [Patent Documents]

[0007] [Patent Document 1] Japanese Patent Publication No. 2017-176580 [Patent Document 2] Japanese Patent Publication No. 2018-102519 [Patent Document 3] Japanese Patent Publication No. 2019-058625 [Patent Document 4] Japanese Patent Publication No. 2020-185138 [Patent Document 5] Japanese Patent Publication No. 2024-071824 [Overview of the Initiative] [Problems that the invention aims to solve]

[0008] Here, the basic emotion model and the Russell circumplex model differ in their methods of classifying psychological states. When the classification models for psychological states differ in the estimation results, the data formats of the estimation results are different, making it difficult to compare and analyze the estimation results to improve the accuracy of the estimation of psychological states. Therefore, when estimating an individual's psychological state using multiple types of measurement data and multiple types of classification models as described above, it is desirable to provide a user interface that integrates each estimation result.

[0009] In one aspect, this disclosure aims to provide a user interface for estimating an individual's psychological state using different classification models based on multiple types of measurement data. [Means for solving the problem]

[0010] According to one embodiment, the psychological state estimation device is A first estimation unit that estimates the psychological state of the subject based on a first classification model using first data obtained from measuring the subject, A second estimation unit estimates the psychological state of the subject based on a second classification model using second data obtained from the subject, A conversion unit converts the first estimation result estimated by the first estimation unit into the psychological state of the subject based on a second classification model, The system includes an output unit that outputs information based on the first estimation result converted by the conversion unit and the second estimation result estimated by the second estimation unit, using a classification method corresponding to the second classification model. [Effects of the Invention]

[0011] According to this disclosure, it is possible to provide a user interface for estimating an individual's psychological state using multiple types of measurement data and different classification models. [Brief explanation of the drawing]

[0012] [Figure 1] This figure shows an example of the system configuration of a psychological state estimation system equipped with a psychological state estimation device according to the first embodiment. [Figure 2] This figure shows an example of the hardware configuration of a psychological state estimation device. [Figure 3] This figure shows the details of the functional configuration of the psychological state estimation unit realized in the psychological state estimation device according to the first embodiment. [Figure 4] This figure shows a specific example of processing performed by the basic emotion estimation unit. [Figure 5] The first figure shows a specific example of processing by the conversion unit. [Figure 6] It is a diagram showing a specific example of the processing by the Russell ring estimation unit. [Figure 7A] It is a first diagram showing a specific example of the processing by the integration unit. [Figure 7B] It is a second diagram showing a specific example of the processing by the integration unit. [Figure 8A] It is a first diagram showing a specific example of the processing by the display unit. [Figure 8B] It is a second diagram showing a specific example of the processing by the display unit. [Figure 9] It is a flowchart showing the flow of the mental state estimation process by the mental state estimation device according to the first embodiment. [Figure 10] It is a diagram showing the detailed functional configuration of the mental state estimation unit realized in the mental state estimation device according to the second embodiment. [Figure 11] It is a second diagram showing a specific example of the processing by the conversion unit. [Figure 12A] It is a third diagram showing a specific example of the processing by the integration unit. [Figure 12B] It is a fourth diagram showing a specific example of the processing by the integration unit. [Figure 13A] It is a third diagram showing a specific example of the processing by the display unit. [Figure 13B] It is a fourth diagram showing a specific example of the processing by the display unit. [Figure 14] It is a flowchart showing the flow of the mental state estimation process by the mental state estimation device according to the second embodiment.

Mode for Carrying Out the Invention

[0013] Hereinafter, each embodiment will be described with reference to the accompanying drawings. In the present specification and the drawings, components having substantially the same functional configuration are denoted by the same reference numerals, and redundant description is omitted.

[0014] [First Embodiment] <System Configuration of Mental State Estimation System> First, the system configuration of a psychological state estimation system equipped with a psychological state estimation device according to the first embodiment will be described. Figure 1 is a diagram showing an example of the system configuration of a psychological state estimation system equipped with a psychological state estimation device according to the first embodiment.

[0015] As shown in Figure 1, the psychological state estimation system 100 comprises a measuring device 111, a measuring device 112, and a psychological state estimation device 120.

[0016] The measuring device 111 is a measuring device (in the first embodiment, an example of the first measuring device) that measures physical changes in the subject 130. Measuring physical changes in the subject 130 means, for example, measuring the subject 130's facial expressions, posture, voice, etc. Therefore, the measuring device 111 includes an imaging device, a voice acquisition device, etc. The measurement data measured by the measuring device 111 (in the first embodiment, an example of the first data; here, referred to as measurement data A) is transmitted to the psychological state estimation device 120.

[0017] The measuring device 112 is a measuring device (in the first embodiment, an example of the second measuring device) that measures the physiological changes of the subject 130. Measuring the physiological changes of the subject 130 means, for example, measuring the subject 130's electroencephalogram, electrocardiogram, pulse wave, etc. Therefore, the measuring device 112 includes an electroencephalograph, electrocardiograph, pulse wave meter, etc. The measurement data measured by the measuring device 112 (in the first embodiment, an example of the second data; here, referred to as measurement data B) is transmitted to the psychological state estimation device 120.

[0018] The psychological state estimation device 120 has a psychological state estimation program installed, and when this program is executed, the psychological state estimation device 120 realizes the psychological state estimation unit 121.

[0019] The psychological state estimation unit 121 estimates the psychological state based on the basic emotion model based on measurement data A transmitted from the measurement device 111, and estimates the psychological state based on the Russell circumplex model based on measurement data B transmitted from the measurement device 111.

[0020] The psychological state estimation unit 121 converts the estimation result of the psychological state estimated based on the basic emotion model into a psychological state based on the Russell annular model, and integrates the converted estimation result with the estimation result of the psychological state based on the Russell annular model estimated based on the measurement data B.

[0021] The psychological state estimation unit 121 displays the integrated estimation results using a classification method corresponding to the Russell annular model.

[0022] Thus, the psychological state estimation device 120 according to the first embodiment integrates the estimation results of psychological states based on different types of models. As a result, the psychological state estimation device 120 according to the first embodiment can provide a user interface for estimating the psychological state of a subject 130 using measurement data A and measurement data B with different classification methods.

[0023] <Hardware configuration of the psychological state estimation device> Next, the hardware configuration of the psychological state estimation device 120 will be described. Figure 2 shows an example of the hardware configuration of the psychological state estimation device. As shown in Figure 2, the psychological state estimation device 120 includes a processor 201, memory 202, auxiliary storage device 203, connection device 204, communication device 205, and drive device 206. Each piece of hardware in the psychological state estimation device 120 is interconnected via a bus 207.

[0024] The processor 201 has various computing devices such as a CPU (Central Processing Unit) and a GPU (Graphics Processing Unit). The processor 201 reads various programs (for example, a psychological state estimation program) into memory 202 and executes them.

[0025] Memory 202 has main memory devices such as ROM (Read Only Memory) and RAM (Random Access Memory). The processor 201 and memory 202 form a so-called computer, and the computer realizes various functions by having the processor 201 execute various programs read from memory 202.

[0026] The auxiliary storage device 203 stores various programs and various data used when those programs are executed by the processor 201.

[0027] The connection device 204 connects the operating device 211 and the display device 212 to the psychological state estimation device 120.

[0028] The communication device 205 communicates with the measuring device 111 and receives measurement data A transmitted from the measuring device 111. The communication device 205 also communicates with the measuring device 112 and receives measurement data B transmitted from the measuring device 112.

[0029] The drive device 206 is a device for setting the recording medium 213. The recording medium 213 here includes media that record information optically, electrically, or magnetically, such as CD-ROMs, flexible disks, and magneto-optical disks. The recording medium 213 may also include semiconductor memory that records information electrically, such as ROMs and flash memory.

[0030] The various programs to be installed on the auxiliary storage device 203 are installed, for example, when the distributed recording medium 213 is set in the drive device 206 and the various programs recorded on the recording medium 213 are read by the drive device 206. Alternatively, the various programs to be installed on the auxiliary storage device 203 may be installed by downloading them from a network (not shown) via the communication device 205.

[0031] <Functional Configuration of the Psychological State Estimation Device> Next, the functional configuration of the psychological state estimation device 120 will be described. Figure 3 is a diagram showing the details of the functional configuration of the psychological state estimation unit realized in the psychological state estimation device according to the first embodiment. As described above, the psychological state estimation device 120 has a psychological state estimation program installed, and when this program is executed, the psychological state estimation device 120 realizes the psychological state estimation unit 121.

[0032] As shown in Figure 3, the psychological state estimation unit 121 includes a data acquisition unit 311, a basic emotion estimation unit 312, a conversion unit 313, a data acquisition unit 321, a Russell circle estimation unit 322, an integration unit 323, and a display unit 324.

[0033] The data acquisition unit 311 acquires the measurement data A transmitted from the measuring device 111 and notifies the basic emotion estimation unit 312.

[0034] The basic emotion estimation unit 312, in the first embodiment, is an example of the first estimation unit, and estimates the psychological state based on the basic emotion model (in the first embodiment, an example of the first classification model) based on the measurement data A. The basic emotion estimation unit 312 notifies the conversion unit 313 of the estimation result (in the first embodiment, an example of the first estimation result).

[0035] The conversion unit 313 converts the estimated psychological state based on the basic emotion model into a psychological state based on the Russell annular model, and notifies the integration unit 323 of the converted estimated result.

[0036] The data acquisition unit 311 acquires the measurement data B transmitted from the measuring device 112 and notifies the Russell ring estimation unit 322.

[0037] The Russell ring estimation unit 322, in the first embodiment, is an example of the second estimation unit, and estimates the psychological state based on the Russell ring model (an example of the second classification model in the first embodiment) based on the measurement data B. The Russell ring estimation unit 322 notifies the integration unit 323 of the estimation result (an example of the second estimation result in the first embodiment).

[0038] The integration unit 323 is an example of a determination unit, and it integrates the converted estimation result notified by the conversion unit 313 with the estimation result of the psychological state based on the Russell ring model notified by the Russell ring estimation unit 322. Note that integrating the converted estimation result notified by the conversion unit 313 with the estimation result of the psychological state based on the Russell ring model notified by the Russell ring estimation unit 322 involves... • Determining whether the two estimation results are the same or similar. • When determined to be identical or similar, decide to display the information based on the two estimation results on the Russell annular model graph (i.e., using a classification method appropriate to the Russell annular model) in a manner that indicates a high level of confidence in the estimation result. If it is determined that they do not match and are not similar, The information based on the two estimation results is displayed on a graph of the Russell annular model (i.e., according to a classification method appropriate to the Russell annular model) in a way that indicates that the estimation result has low confidence, or, Display a message indicating that a highly reliable estimation result could not be obtained. To decide, This includes the following: In other words, the integration unit 323 determines the display method when displaying information based on the two estimation results in a single graph, based on the determination result of whether the two estimation results are the same or similar.

[0039] The integration unit 323 notifies the display unit 324 of the converted estimation result notified by the conversion unit 313, the estimation result of the psychological state based on the Russell ring model notified by the Russell ring estimation unit 322, and the determined display mode.

[0040] The display unit 324 is an example of an output unit. • The converted estimation result notified from the conversion unit 313, • The estimation results of the psychological state based on the Russell annular model, as notified by the Russell annular estimation unit 322, The information based on (information based on two estimation results) is displayed on a graph of the Russell annular model (i.e., according to the classification method corresponding to the Russell annular model) in the notified display format (an example of the output format).

[0041] <Specific examples of processing performed by each part of the psychological state estimation unit> Next, we will explain specific examples of processing performed by each part of the psychological state estimation unit 121 (here, the basic emotion estimation unit 312, the conversion unit 313, the Russell circle estimation unit 322, the integration unit 323, and the display unit 324).

[0042] (1) Specific example of processing by the basic emotion estimation unit 312 Figure 4 shows a specific example of processing by the basic emotion estimation unit. As described above, the basic emotion estimation unit 312 estimates the psychological state based on the basic emotion model based on the measurement data A. In Figure 4, the symbol 410 shows an example of the estimation result of the psychological state based on the basic emotion model estimated by the basic emotion estimation unit 312. As shown by the symbol 410, the basic emotion model includes seven basic emotions: "anger," "disgust," "fear," "joy," "sadness," "surprise," and "neutral." As shown by the symbol 410, the estimation result of the psychological state based on the basic emotion model is represented by a graph showing the probability of each of the seven basic emotions.

[0043] (2) Specific example of processing by the conversion unit 313 Figure 5 is the first diagram showing a specific example of processing by the conversion unit. As described above, the conversion unit 313 converts the estimation result of the psychological state based on the basic emotion model into a psychological state based on the Russell ring model. In Figure 5, the symbol 510 shows an example of the converted estimation result, which is obtained by converting the estimation result of the psychological state based on the basic emotion model (symbol 410) into a psychological state based on the Russell ring model. As shown by the symbol 510, the Russell ring model is represented by a graph that includes axes for "joy" and "irritation", axes for "excitement" and "sleepiness", axes for "tension" and "relief", and axes for "emotion" and "boredom".

[0044] As shown in reference numeral 510, the estimation result of the psychological state based on the basic emotion model (reference numeral 410) is converted to a psychological state based on the Russell annular model, and the converted estimation result is reflected in the region on the graph of the Russell annular model (reference numeral 511).

[0045] Note that the symbol 510 in Figure 5 is an example of the estimated result after conversion. Generally, when the estimated results of psychological states based on the basic emotion model are converted to psychological states based on the Russell circumplex model, the following tendencies are observed. If the estimated psychological state based on the basic emotion model is "anger," "disgust," "fear" (or "contempt"), the transformed estimated result is reflected in the second quadrant of the Russell cyclic model graph. If the estimated psychological state based on the basic emotion model is "joy," the transformed estimated result will be reflected in the first or fourth quadrant on the Russell annular model graph. • If the estimated psychological state based on the basic emotion model is "sadness," the transformed estimated result is reflected in the third quadrant of the Russell annular model graph. However, there are also cases where the transformed estimated result is reflected in the first or second quadrant of the Russell annular model graph. If the estimated psychological state based on the basic emotion model is "surprise," the transformed estimate will be reflected in the first or second quadrant on the Russell cyclic model graph. • If the estimated psychological state based on the basic emotion model is "neutral," the transformed estimated result is reflected in the central region of the Russell annular model graph.

[0046] Furthermore, the example in Figure 5 shows a case in which the conversion unit 313 converts the probabilities of each basic emotion included in the estimation results of psychological states based on the basic emotion model (indicated by 410), and reflects the converted estimation results in a region on the graph of the Russell annular model.

[0047] However, the conversion method by the conversion unit 313 is not limited to this. For example, the probability of each basic emotion included in the estimation result of psychological state based on the basic emotion model (indicated by 410) may be converted, and the converted estimation result may be reflected in the marks (symbols indicating a single point) on the graph of the Russell annular model.

[0048] Alternatively, the conversion unit 313 selects from the probabilities of each basic emotion included in the estimation result of the psychological state based on the basic emotion model (code 410), • Convert the highest probability, or, • Transform the probabilities of the top n integers (2 ≤ n ≤ 6), The transformed estimation results may be reflected in the region on the graph of the Russell annular model.

[0049] Alternatively, the conversion unit 313 selects from the probabilities of each basic emotion included in the estimation result of the psychological state based on the basic emotion model (code 410), • Convert the highest probability, or, • Transform the probabilities of the top n integers (2 ≤ n ≤ 6), The converted estimation results may be reflected as markers on the graph of the Russell annular model.

[0050] (3) Specific example of processing by the Russell ring estimation unit 322 Figure 6 shows a specific example of processing by the Russell ring estimation unit. As described above, the Russell ring estimation unit 322 estimates the psychological state based on the Russell ring model based on the measurement data B. In Figure 6, reference numeral 610 indicates an example of the estimation result of the psychological state based on the Russell ring model estimated by the Russell ring estimation unit 322. As shown by reference numeral 610, the estimation result of the psychological state based on the Russell ring model is reflected in the marks (reference numeral 611) on the graph of the Russell ring model.

[0051] However, the estimation results of psychological states based on the Russell annular model may be reflected in the region on the Russell annular model graph.

[0052] (4) Specific example of processing by the integration unit 323 1 Figure 7A is the first diagram showing a specific example of processing by the integration unit. As described above, the integration unit 323 integrates the converted estimation result notified by the conversion unit 313 and the estimation result of the psychological state based on the Russell ring model notified by the Russell ring estimation unit 322. In Figure 7A, reference numeral 710 denotes: The converted estimation result (code 511) notified from the conversion unit 313, The estimation result of the psychological state based on the Russell annular model, as notified by the Russell annular estimation unit 322 (indicated by symbol 611), This shows how the determination of whether or not they match or are similar was made on the graph of the Russell annular model (symbol 710).

[0053] Specifically, an example of reference numeral 710 is that the integration unit 323 is, The centroid position of the region reflecting the converted estimation result (code 511) notified from the conversion unit 313, The position of the mark reflecting the estimation result of the psychological state based on the Russell annular model (indicated by 611), as notified by the Russell annular estimation unit 322, This shows how the distance between them was calculated.

[0054] Furthermore, the example of symbol 710 shows that the integration unit 323 determined that the two estimation results were identical or similar because the calculated distance was below the threshold. In this case, the integration unit 323 decides to display the information based on the two estimation results on the Russell annular model graph (i.e., using a classification method appropriate to the Russell annular model) in a display mode that indicates a high level of confidence in the two estimation results.

[0055] In the above explanation, the integration unit 323 uses distance on the Russell annular model graph to determine whether the two estimation results match or are similar, but other parameters may also be used.

[0056] Furthermore, in the above explanation, when calculating the distance on the graph of the Russell annular model, the centroid position of the region reflecting the converted estimation result (indicated by 511) notified by the conversion unit 313 was used, but a representative point other than the centroid position may also be used.

[0057] (5) Specific example of processing by the integration unit 323 2 Figure 7B is a second diagram showing a specific example of processing by the integration unit. The difference from Figure 7A is that the converted estimation result notified by the conversion unit 313 is reflected in the area indicated by reference numeral 711.

[0058] Therefore, in the example of symbol 720, The centroid position of the region reflecting the converted estimation result (code 711) notified from the conversion unit 313, The position of the mark reflecting the estimation result of the psychological state based on the Russell annular model (indicated by 611), as notified by the Russell annular estimation unit 322, The distance between them is greater than or equal to the threshold.

[0059] In this case, the integration unit 323 determines that the two estimation results do not match and are not similar. As a result, the integration unit 323 decides to display the information based on the two estimation results on the Russell annular model graph (i.e., using a classification method appropriate to the Russell annular model) in a display mode that indicates the low reliability of the two estimation results.

[0060] (6) Specific example of processing by the display unit 324 1 Figure 8A is the first diagram showing a specific example of processing by the display unit. As described above, the display unit 324 displays information based on the two estimation results on the Russell annular model graph in the notified display mode.

[0061] In Figure 8A, reference numeral 810 denotes a graph of the Russell ring model displayed by the display unit 324. Reference numeral 811 indicates how information based on the converted estimation result (reference numeral 511) is displayed on the Russell ring model graph using the same type of mark as reference numeral 611 (however, the mark is such that it can be recognized that the estimation result is based on different measurement data than reference numeral 611). Reference numeral 611 indicates how information based on the estimation result of the psychological state based on the Russell ring model, notified by the Russell ring estimation unit 322, is displayed using the same type of mark as reference numeral 811.

[0062] As shown in Figure 8A, the display unit 324 displays the information based on the estimation result indicated by reference numeral 811 and reference numeral 611 in a display manner that indicates a high level of confidence in the two estimation results.

[0063] In the example of the display of reference numeral 810 in Figure 8A, the case where the two estimation results are displayed using the same type of mark is shown, but the representation format is not limited to this, and different representation formats may be used. For example, as shown in reference numeral 710, information based on the two estimation results may be displayed using a region and a mark.

[0064] Furthermore, while the example of displaying symbol 810 in Figure 8A shows the case where information based on the two estimation results is displayed separately, it is also possible to display information that combines the two estimation results into one. The methods for combining them into one can be broadly categorized into the following two patterns. • A pattern (Pattern 1) in which the estimation result shown by code 611 and the converted estimation result shown by code 511 are treated as equivalent. • A pattern (Pattern 2) in which the converted estimation result shown by code 511 is treated as a supplementary result to the estimation result shown by code 611.

[0065] For Pattern 1, several further cases are possible. For example, suppose both estimation results are reflected in marks on the graph of the Russell annular model. In this case, the display unit 324 may display a mark combining the two estimation results at the midpoint of each mark. Alternatively, suppose one of the two estimation results is reflected in a region on the graph of the Russell annular model, and the other is reflected in a mark on the Russell annular model. In this case, the display unit 324 may display a mark combining the two estimation results at the midpoint between the centroid of the region and the position of the mark. Also, suppose both estimation results are reflected in a region on the graph of the Russell annular model. In this case, the display unit 324 may display the overlapping region of the two regions as a region combining the two estimation results. Or, the display unit 324 may display a mark combining the two estimation results at the midpoint of the centroid of each of the two regions.

[0066] Similarly, for pattern 2, several further cases are possible. For example, suppose both estimation results are reflected in marks on the graph of the Russell annular model. In this case, the display unit 324 may display a combined mark for the two estimation results at the position indicated by the coordinates obtained by weighting the coordinates indicating the positions of each mark and dividing by 2. Alternatively, suppose one of the two estimation results is reflected in a region on the graph of the Russell annular model, and the other is reflected in a mark on the graph of the Russell annular model. In this case, the coordinates indicating the centroid position of one region and the coordinates indicating the position of the other mark may be weighted and added together, and a combined mark for the two estimation results may be displayed at the position indicated by the coordinates obtained by dividing by 2. Alternatively, suppose both estimation results are reflected in regions on the graph of the Russell annular model. In this case, the coordinates indicating the centroid positions of the two regions may be weighted and added together, and a combined mark for the two estimation results may be displayed at the position indicated by the coordinates obtained by dividing by 2.

[0067] In the case of Pattern 2, the configuration may also include displaying marks or regions on the Russell annular model graph at positions that reflect the estimation results of the psychological state based on the Russell annular model (indicated by 611).

[0068] (7) Specific example of processing by the display unit 324 2 Figure 8B is a second diagram showing a specific example of processing by the display unit. The difference from Figure 8A is that the integration unit 323 determined that the reliability of the two estimation results was low.

[0069] Therefore, in the example of reference numeral 820, the display unit 324 displays the information based on the estimation results shown in reference numeral 821 and reference numeral 822 on the graph of the Russell annular model using marks, in a display manner that indicates that the reliability of the two estimation results is low.

[0070] In the example of reference numeral 820, the display unit 324 shows information based on two estimation results in a display manner indicating low reliability, but the method of displaying the two estimation results by the display unit 324 is not limited to this. For example, the display unit 324 may display a message indicating that a highly reliable estimation result could not be obtained. Alternatively, this message may be displayed together with the graph shown in reference numeral 820.

[0071] <Flowchart of the psychological state estimation process> Next, the flow of the psychological state estimation process by the psychological state estimation device 120 will be described. Figure 9 is a flowchart showing the flow of the psychological state estimation process by the psychological state estimation device according to the first embodiment.

[0072] In step S901, the psychological state estimation device 120 acquires measurement data A, which measures the physical changes of the subject 130. Using the acquired measurement data A, the psychological state estimation device 120 estimates the psychological state based on the basic emotion model, thereby obtaining the estimated result of the psychological state based on the basic emotion model.

[0073] In step S902, the psychological state estimation device 120 acquires measurement data B, which measures the physiological changes of the subject 130. Using the acquired measurement data B, the psychological state estimation device 120 estimates the psychological state based on the Russell annular model, thereby obtaining the estimation result of the psychological state based on the Russell annular model.

[0074] In step S903, the psychological state estimation device 120 converts the estimation result of the psychological state based on the basic emotion model into a psychological state based on the Russell circumplex model.

[0075] In step S904, the psychological state estimation device 120 calculates the distance on the Russell annular model graph between the converted estimation result obtained by converting the estimation result of the psychological state based on the basic emotion model and the estimation result of the psychological state based on the Russell annular model. Based on this, the psychological state estimation device 120 determines whether the two estimation results match or are similar. The psychological state estimation device 120 also decides, based on the determination result, whether to display it in a display mode that indicates high confidence or in a display mode that indicates low confidence.

[0076] In step S905, the psychological state estimation device 120 displays the information based on the two estimation results on a Russell annular model graph in the determined display mode.

[0077] <Summary> As is clear from the above description, the psychological state estimation device 120 according to the first embodiment is Using measurement data A, which represents the physical changes of 130 subjects, the psychological state of 130 subjects will be estimated based on a basic emotional model. Using measurement data B, which represents the physiological changes of 130 subjects, the psychological state of 130 subjects will be estimated based on the Russell circumplex model. • Convert the estimated psychological state based on the basic emotion model into a psychological state based on the Russell Circular Model. The information based on the converted estimation results and the estimation results of psychological states based on the Russell annular model is displayed on the Russell annular model graph (according to the classification method of the Russell annular model) in a display manner corresponding to the confidence level of the two estimation results.

[0078] As a result, the psychological state estimation device 120 according to the first embodiment can provide a user interface for estimating the psychological state of an individual subject using multiple types of measurement data and different classification models.

[0079] [Second Embodiment] In the first embodiment described above, the estimated psychological state based on the basic emotion model is converted into a psychological state based on the Russell annular model, and the information based on the two estimated results is displayed on the Russell annular model graph in a display manner corresponding to the level of confidence.

[0080] In contrast, in the second embodiment, the estimation results of the psychological state based on the Russell cyclic model are converted into psychological states based on the basic emotion model, and the information based on the two estimation results is displayed on the graph of the basic emotion model in a display manner according to the level of confidence. The second embodiment will be described below, focusing on the differences from the first embodiment described above.

[0081] <Functional Configuration of the Psychological State Estimation Device> First, the functional configuration of the psychological state estimation device 120 according to the second embodiment will be described. Figure 10 is a diagram showing the details of the functional configuration of the psychological state estimation unit realized in the psychological state estimation device according to the second embodiment. As described above, the psychological state estimation device 120 has a psychological state estimation program installed, and when this program is executed, the psychological state estimation device 120 realizes the psychological state estimation unit 121.

[0082] As shown in Figure 10, in the second embodiment, the psychological state estimation unit 121 includes a data acquisition unit 1011, a basic emotion estimation unit 1012, an integration unit 1013, a data acquisition unit 1021, a Russell circle estimation unit 1022, a conversion unit 1023, and a display unit 1014.

[0083] Of these, the data acquisition unit 1011 and the basic emotion estimation unit 1012 are the same as the data acquisition unit 311 and the basic emotion estimation unit 312 in Figure 3, so their explanation is omitted here. Also, the data acquisition unit 1021 and the Russell ring estimation unit 1022 are the same as the data acquisition unit 321 and the Russell ring estimation unit 322 in Figure 3, so their explanation is omitted here. However, in the second embodiment, the Russell ring estimation unit 1022 is an example of the first estimation unit, the basic emotion estimation unit 1012 is an example of the second estimation unit, and the estimation result estimated by the basic emotion estimation unit 1012 is an example of the second estimation result. Furthermore, in the second embodiment, the measurement device 112 is an example of the first measurement device, and the measurement device 111 is an example of the first measurement device. Moreover, in the second embodiment, the measurement data B measured by the measurement device 112 is an example of the first data, and the measurement data A measured by the measurement device 111 is an example of the second data.

[0084] The conversion unit 1023 converts the estimation result of the psychological state based on the Russell annular model (an example of the first classification model in the second embodiment) into a psychological state based on the basic emotion model (an example of the second classification model in the second embodiment). The conversion unit 1023 notifies the integration unit 1013 of the converted estimation result (an example of the converted first estimation result in the second embodiment).

[0085] The integration unit 1013 is an example of a determination unit, and it integrates the converted estimation result notified by the conversion unit 1023 with the estimation result of the psychological state based on the basic emotion model notified by the basic emotion estimation unit 1012 (an example of the second estimation result in the second embodiment). Note that integrating the converted estimation result notified by the conversion unit 1023 with the estimation result of the psychological state based on the basic emotion model notified by the basic emotion estimation unit 1012 involves... • Determining whether the two estimation results are the same or similar. If the results are determined to be identical or similar, the information based on the two estimation results will be displayed on the graph of the basic sentiment model (i.e., according to the classification method appropriate to the basic sentiment model) in a manner that indicates that the estimation results are highly reliable. If it is determined that they do not match and are not similar, The information based on the two estimation results is displayed on the graph of the basic sentiment model (i.e., according to the classification method corresponding to the basic sentiment model) in a way that indicates that the estimation result has low reliability, or, Display a message indicating that a highly reliable estimation result could not be obtained. To decide, This includes the following: In other words, the integration unit 323 determines the display method when displaying information based on the two estimation results in a single graph, based on the determination result of whether the two estimation results are the same or similar.

[0086] The integration unit 1013 notifies the display unit 1014 of the converted estimation result notified by the conversion unit 1023, the estimation result of the psychological state based on the basic emotion model notified by the basic emotion estimation unit 1012, and the determined display mode.

[0087] The display unit 1014 is an example of an output unit. • The converted estimation result notified from the conversion unit 1023, • The results of the estimation of psychological state based on the basic emotion model, as notified by the basic emotion estimation unit 1012, The information based on the two estimation results (information based on the two estimation results) is displayed on the graph of the basic sentiment model (i.e., according to the classification method corresponding to the basic sentiment model) in the notified display format (an example of the output format).

[0088] <Specific examples of processing performed by each part of the psychological state estimation unit> Next, we will explain specific examples of processing performed by each part of the psychological state estimation unit 121 (here, the conversion unit 1023, the integration unit 1013, and the display unit 1014).

[0089] (1) Specific example of processing by the conversion unit 1023 Figure 11 is a second diagram showing a specific example of processing by the conversion unit. As described above, the conversion unit 1023 converts the estimation result of the psychological state based on the Russell annular model into a psychological state based on the basic emotion model. In Figure 11, the symbol 1120 shows an example of the converted estimation result, in which the estimation result of the psychological state based on the Russell annular model (symbol 1110) (symbol 1111) is converted into a psychological state based on the basic emotion model.

[0090] As shown in symbol 1120, the estimation result of the psychological state based on the Russell annular model (symbol 1111) is converted to a psychological state based on the basic emotion model, and the converted estimation result is reflected in the graph of the basic emotion model.

[0091] Note that the symbol 1120 in Figure 11 is an example of the estimated result after conversion, and generally, when the estimated results of psychological states based on the Russell circumplex model are converted to psychological states based on the basic affective model, the following tendencies are observed. If the estimated psychological state based on the Russell annular model falls within the first quadrant of the Russell annular model graph, the transformed estimated state will have a higher probability of being "joyful," "surprised," or "sad" in the basic emotion model graph. If the estimated psychological state based on the Russell Circular Model falls within the second quadrant of the Russell Circular Model graph, the transformed estimated state has a higher probability of being "anger," "disgust," "fear," "contempt," "surprise," or "sadness" in the basic emotion model graph. • If the estimated psychological state based on the Russell annular model falls within the third quadrant of the Russell annular model graph, the transformed estimated result will have a higher probability of being "sad" in the basic emotion model graph. • If the estimated psychological state based on the Russell annular model falls within the fourth quadrant of the Russell annular model graph, the transformed estimated result has a higher probability of being "joyful" or "neutral" in the basic emotion model graph.

[0092] Furthermore, the example in Figure 11 shows a case in which the conversion unit 1023 converts the estimation result of the psychological state based on the Russell annular model (indicated by 1110) into the probability of each basic emotion in the graph of the basic emotion model.

[0093] However, the conversion method by the conversion unit 1023 is not limited to this. For example, the conversion unit 1023 may convert the estimation result of the psychological state based on the Russell annular model (indicated by 1110) into the probability of any one of the basic emotions on the graph of the basic emotion model.

[0094] Alternatively, the estimation results of psychological states based on the Russell annular model (symbol 1110) may be converted into the probabilities of the top n basic emotions (integers 2 ≤ n ≤ 6) on the graph of the basic emotion model.

[0095] (2) Specific example of processing by the integration unit 1013 1 Figure 12A is a third figure showing a specific example of processing by the integration unit. As described above, the integration unit 1013 integrates the converted estimation result notified by the conversion unit 1023 and the estimation result of the psychological state based on the basic emotion model notified by the basic emotion estimation unit 1012. In Figure 12A, reference numeral 1220 denotes: The converted estimation result (code 1120) notified from the conversion unit 1023, • The estimation result of the psychological state based on the basic emotion model (code 1210) notified from the basic emotion estimation unit 1012, This shows how the system determined whether the items matched or were similar.

[0096] An example of code 1220 is: The probability of each basic emotion included in the converted estimation result (code 1120) notified from the conversion unit 1023, The probability of each basic emotion included in the estimation result of the psychological state based on the basic emotion model (code 1210) notified by the basic emotion estimation unit 1012, This indicates that the similarity was determined to be greater than a predetermined threshold. In this case, the integration unit 1013, • A display method that indicates a high level of confidence in the two estimation results, • Information based on two estimation results (for example, an estimation result obtained by summing two estimation results) • Display on a graph of the basic emotion model (i.e., according to the classification method corresponding to the basic emotion model). To decide.

[0097] In the above explanation, the determination of whether two estimation results match or are similar is based on whether the similarity is greater than a threshold. However, the method for determining whether two estimation results match or are similar is not limited to this.

[0098] Furthermore, in the above explanation, the integration unit 1013 uses the probability of each basic emotion to determine whether the two estimation results match or are similar. However, the probabilities used for the determination are not limited to these, and the determination of whether the two estimation results match or are similar may be made based on whether the basic emotion with the highest probability matches. Alternatively, the determination of whether the two estimation results match or are similar may be made based on whether the basic emotion with a probability above a predetermined threshold matches.

[0099] (3) Specific example of processing by the integration unit 1013 2 Figure 12B is a fourth figure showing a specific example of processing by the integration unit. The difference from Figure 12A is that the converted estimation result notified by the conversion unit 1023 is significantly different from the estimation result shown by reference numeral 1210 in Figure 12A.

[0100] Therefore, in the example of reference numeral 1240, The converted estimation result (code 1230) notified from the conversion unit 1023, • The estimation result of the psychological state based on the basic emotion model (code 1210) notified from the basic emotion estimation unit 1012, The similarity is below a predetermined threshold.

[0101] In this case, the integration unit 1013 determines that the two estimation results do not match and are not similar. As a result, the integration unit 1013 decides to display the information based on the two estimation results on the graph of the basic sentiment model (i.e., using a classification method appropriate to the basic sentiment model) in a display mode that indicates the low reliability of the two estimation results.

[0102] (4) Specific example of processing by the display unit 1014 1 Figure 13A is a third figure showing a specific example of processing by the display unit. As described above, the display unit 1014 displays information based on the converted estimation result notified by the conversion unit 1023 and the estimation result of the psychological state based on the basic emotion model notified by the basic emotion estimation unit 1012 on the graph of the basic emotion model in the notified display mode.

[0103] In Figure 13A, reference numeral 1310 indicates an example of a graph of the basic sentiment model displayed by the display unit 1014. As shown by reference numeral 1310 in Figure 13A, the display unit 1014 displays information based on two estimation results (the estimation result obtained by summing the two estimation results) on the graph of the basic sentiment model in a display manner that indicates a high level of confidence in the two estimation results.

[0104] In the example of the display of symbol 1310 in Figure 13A, the two estimation results are displayed in the same format by summing them on the graph of the basic emotion model. However, they may be displayed in different formats. Alternatively, they may be displayed separately for each basic emotion in different formats without summing them.

[0105] Furthermore, in the example of the display of symbol 1310 in Figure 13A, the probabilities of each basic emotion were added together when displaying the two estimation results on the basic emotion model graph. However, it is also possible to display only the estimation result obtained by adding together the maximum probabilities of each of the two estimation results on the basic emotion model graph.

[0106] Furthermore, while the example of displaying symbol 1310 in Figure 13A shows the case where information based on two estimation results is displayed separately on the graph of the basic emotion model, it is also possible to display information that combines the two estimation results into one. The methods for combining them into one can be broadly divided into the following two patterns. • A pattern (Pattern 1) in which the estimation result shown by symbol 1210 in Figure 12A and the converted estimation result shown by symbol 1120 in Figure 12A are treated as equivalent. • A pattern (Pattern 2) in which the converted estimation result shown at code 1120 in Figure 12A is treated as a supplementary result to the estimation result shown at code 1210 in Figure 12A.

[0107] For Pattern 1, several further cases are possible. For example, the two estimated results are added together and divided by 2, and the resulting estimate is displayed on the graph of the basic sentiment model. Alternatively, the maximum probabilities of each of the two estimated results are added together and divided by 2, and the resulting estimate is displayed on the graph of the basic sentiment model.

[0108] Similarly, for pattern 2, several further cases are possible. For example, the two estimation results could be weighted and added together, and the result obtained by dividing by 2 could be displayed on the graph of the basic sentiment model. Alternatively, the maximum probabilities of each of the two estimation results could be weighted and added together, and the result obtained by dividing by 2 could be displayed on the graph of the basic sentiment model.

[0109] In the case of Pattern 2, the configuration may be such that only the graph of the basic emotion model is displayed, reflecting the estimated result of the psychological state based on the basic emotion model (code 1210 in Figure 12A). Alternatively, the configuration may be such that only the graph of the basic emotion model is displayed, reflecting the highest probability among the estimated results of the psychological state based on the basic emotion model (code 1210 in Figure 12A).

[0110] (5) Specific example of processing by the display unit 1014 2 Figure 13B is a fourth figure showing a specific example of processing by the display unit. The difference from Figure 13A is that the integration unit 1013 determined that the reliability of the two estimation results was low.

[0111] Therefore, in the example of reference numeral 1320, the display unit 1014 displays information based on the estimation result shown in reference numeral 1210 and the estimation result shown in reference numeral 1230 on the graph of the basic sentiment model in a display mode that indicates that the reliability of the two estimation results is low.

[0112] In the example of reference numeral 1320, the display unit 1014 shows the information based on the two estimation results in a display manner indicating low reliability, but the method of displaying the information based on the two estimation results by the display unit 1014 is not limited to this. For example, the display unit 1014 may display a message indicating that a highly reliable estimation result could not be obtained. Alternatively, this message may be displayed together with the graph of reference numeral 1320.

[0113] <Flowchart of the psychological state estimation process> Next, the flow of the psychological state estimation process by the psychological state estimation device 120 according to the second embodiment will be described. Figure 14 is a flowchart showing the flow of the psychological state estimation process by the psychological state estimation device according to the second embodiment.

[0114] In step S1401, the psychological state estimation device 120 acquires measurement data A, which measures the physical changes of the subject 130. Using the acquired measurement data A, the psychological state estimation device 120 estimates the psychological state based on the basic emotion model, thereby obtaining the estimated result of the psychological state based on the basic emotion model.

[0115] In step S1402, the psychological state estimation device 120 acquires measurement data B, which measures the physiological changes of the subject 130. Using the acquired measurement data B, the psychological state estimation device 120 estimates the psychological state based on the Russell annular model, thereby obtaining the estimation result of the psychological state based on the Russell annular model.

[0116] In step S1403, the psychological state estimation device 120 converts the estimation result of the psychological state based on the Russell ring model into a psychological state based on the basic emotion model.

[0117] In step S1404, the psychological state estimation device 120 calculates the similarity between the converted estimation result obtained by transforming the estimation result of the psychological state based on the Russell annular model and the estimation result of the psychological state based on the basic affection model. Based on this, the psychological state estimation device 120 determines whether the two estimation results match or are similar. The psychological state estimation device 120 also decides whether to display the result in a way that indicates high confidence or in a way that indicates low confidence based on the determination result.

[0118] In step S1405, the psychological state estimation device 120 displays the information based on the two estimation results on the graph of the basic emotion model in the determined display mode.

[0119] <Summary> As is clear from the above description, the psychological state estimation device 120 according to the second embodiment is Using measurement data A, which represents the physical changes of 130 subjects, the psychological state of 130 subjects will be estimated based on a basic emotional model. Using measurement data B, which represents the physiological changes of 130 subjects, the psychological state of 130 subjects will be estimated based on the Russell circumplex model. • Convert the estimated psychological state based on the Russell Circular Model into a psychological state based on the Basic Emotion Model. The information based on the converted estimation results and the estimation results of the psychological state based on the basic emotion model is displayed on the graph of the basic emotion model (according to the classification method of the basic emotion model) in a display manner that corresponds to the reliability of the two estimation results.

[0120] As a result, the psychological state estimation device 120 according to the second embodiment can provide a user interface for estimating an individual's psychological state using multiple types of measurement data and different classification models.

[0121] [Third Embodiment] In the embodiments described above, the case in which measuring device 111 measures physical changes and measuring device 112 measures physiological changes was explained. However, the system configuration of the psychological state estimation system 100 is not limited to this, and either of the two measuring devices 111 and 112 may be configured to measure physical changes. Alternatively, either of the two measuring devices 111 and 112 may be configured to measure physiological changes.

[0122] When both of the two types of measuring devices 111 and 112 are configured to measure physical changes, the psychological state estimation unit 121 estimates the psychological state based on six basic emotion models using, for example, measurement data A measured by measuring device 111. Furthermore, the psychological state estimation unit 121 estimates the psychological state based on seven basic emotion models using, for example, measurement data B measured by measuring device 112. In addition, the psychological state estimation unit 121 converts the estimation results of the psychological state based on the six basic emotion models into psychological states based on seven basic emotion models and displays them on a graph of the seven basic emotion models in a display mode corresponding to the level of confidence. Alternatively, the psychological state estimation unit 121 converts the estimation results of the psychological state based on seven basic emotion models into psychological states based on six basic emotion models and displays them on a graph of the six basic emotion models in a display mode corresponding to the level of confidence.

[0123] In the example above, we described a case where both of the two measuring devices 111 and 112 measure physical changes, and the psychological state estimation unit 121 estimates the psychological state based on the basic emotion model. However, the same applies to a case where both of the two measuring devices 111 and 112 measure physiological changes, and the psychological state estimation unit 121 estimates the psychological state based on the Russell ring model.

[0124] [Fourth Embodiment] In the embodiments described above, cases were explained in which the psychological state estimation unit 121 estimates either or both of the psychological state based on the basic emotion model and the psychological state based on the Russell ring model. However, the combination of the two models is not limited to these. Models other than the basic emotion model, or models other than the Russell ring model, may also be used in combination.

[0125] Furthermore, in each of the above embodiments, when estimating psychological states based on two types of models, the case in which the estimation result of the psychological state based on one model is converted to the psychological state based on the other model was described. However, the method of converting the estimation result of the psychological state is not limited to this.

[0126] For example, the estimated psychological state results based on the first model and the estimated psychological state results based on the second model may be converted to psychological states based on the third model, respectively. The third model referred to here is a model that has a classification method different from both the first and second models.

[0127] Furthermore, the conversion procedure for converting to a psychological state based on the third model is arbitrary. For example, the estimated result of the psychological state based on the first model may be converted to a psychological state based on the second model, and then the converted estimated result of the psychological state based on the second model and the estimated result of the psychological state based on the second model may be converted together to a psychological state based on the third model. Alternatively, the estimated result of the psychological state based on the second model may be converted to a psychological state based on the first model, and then the converted estimated result of the psychological state based on the first model and the estimated result of the psychological state based on the first model may be converted together to a psychological state based on the third model.

[0128] Furthermore, the phrase "converting them all into a psychological state based on the third model" used here means: • Using separate transformation units to transform each estimation result at the same time, or • Using the same transformation unit to transform each estimation result, or • Integrate and convert the respective estimation results. It includes one of the following.

[0129] [Other embodiments] The conversion content used by the conversion unit 313 or conversion unit 1023 when converting the estimation result in the first or second embodiment described above may be changed, for example, according to the age, gender, personality classification, tendencies, etc., of the subject 130.

[0130] Furthermore, the method used by the integration unit 323 or integration unit 1013 in the first or second embodiment to determine whether the two estimation results match or are similar may be changed, for example, according to the age, gender, personality classification, tendencies, etc., of the subject 130.

[0131] Furthermore, the format in which the display unit 324 or display unit 1014 displays the two estimation results in the first or second embodiment described above may be changed, for example, according to the age, gender, personality classification, tendencies, etc., of the subject 130.

[0132] Furthermore, if the conversion content, judgment method, or expression format is changed according to the age, gender, personality classification, tendencies, etc. of the subject 130, the psychological state estimation unit 121 may have a selection unit that pre-selects the age, gender, personality classification, tendencies, etc. of the subject 130.

[0133] Furthermore, although the display unit 324 or display unit 1014 in the first or second embodiment described above is described as displaying information based on two estimation results, the information based on the two estimation results may also be output by an output method other than display.

[0134] Furthermore, in the first embodiment described above, the details of the conversion method used by the conversion unit 313 to convert the estimation results of the psychological state based on the basic emotion model into the psychological state based on the Russell ring model were not mentioned. However, the conversion unit 313 may, for example, use a machine learning model for the conversion.

[0135] Similarly, in the second embodiment described above, the details of the conversion method used by the conversion unit 1023 to convert the estimation results of psychological states based on the Russell ring model into psychological states based on the basic emotion model were not mentioned. However, the conversion unit 1023 may, for example, use a machine learning model for the conversion.

[0136] Furthermore, although the above embodiments described a case in which one psychological state estimation device 120 implements the psychological state estimation unit 121, multiple psychological state estimation devices may be configured so that each implements a function of the psychological state estimation unit 121.

[0137] It should be noted that the present invention is not limited to the configurations shown in the above embodiments, including combinations with other elements. These aspects can be modified without departing from the spirit of the present invention and can be appropriately determined according to their application. [Explanation of Symbols]

[0138] 100: Psychological State Estimation System 111: Measuring device 112: Measuring device 120: Psychological State Estimation Device 311: Data Acquisition Unit 312: Basic emotion estimation part 313: Conversion Unit 321: Data Acquisition Unit 322: Russell annular estimation unit 323: Integration Department 324:Display section 1011: Data acquisition unit 1012: Basic emotion estimation part 1013: Integration Department 1014: Display section 1021: Data acquisition unit 1022: Russell annular estimation unit 1023: Conversion section

Claims

1. A first estimation unit that estimates the psychological state of the subject based on a first classification model using first data obtained from the subject, A second estimation unit estimates the psychological state of the subject based on a second classification model using second data obtained from the subject, A conversion unit converts the first estimation result estimated by the first estimation unit into the psychological state of the subject based on a second classification model, An output unit outputs information based on the first estimation result converted by the conversion unit and the second estimation result estimated by the second estimation unit, using a classification method corresponding to the second classification model. A psychological state estimation device having the following features.

2. The first data is data measuring the physical changes of the subject, and the second data is data measuring the physiological changes of the subject. The psychological state estimation device according to claim 1.

3. The first classification model is the basic emotion model, and the second classification model is the Russell circumplex model. The psychological state estimation device according to claim 2.

4. The first data is data measuring the physiological changes of the subject, and the second data is data measuring the physical changes of the subject. The psychological state estimation device according to claim 1.

5. The first classification model is the Russell annular model, and the second classification model is the basic emotion model. The psychological state estimation device according to claim 4.

6. The system further includes a determination unit that determines whether the first estimation result converted by the conversion unit matches the second estimation result. The output unit outputs in an output manner corresponding to the determination result by the determination unit. The psychological state estimation device according to claim 3.

7. The system further includes a determination unit that determines whether the first estimation result converted by the conversion unit matches the second estimation result. The output unit is, If the determination unit determines that they match, it outputs in an output mode indicating a high level of confidence. If the determination unit determines that they do not match, it outputs an output indicating low confidence. The psychological state estimation device according to claim 1.

8. The output unit outputs information based on the first estimation result converted by the conversion unit and information based on the second estimation result estimated by the second estimation unit, respectively. The psychological state estimation device according to claim 7.

9. The output unit outputs information that combines the first estimation result converted by the conversion unit and the second estimation result estimated by the second estimation unit. The psychological state estimation device according to claim 7.

10. The output unit combines the first estimation result converted by the conversion unit and the second estimation result estimated by the second estimation unit using the same weights. The psychological state estimation device according to claim 9.

11. The output unit combines the first estimation result converted by the conversion unit and the second estimation result estimated by the second estimation unit using different weights. The psychological state estimation device according to claim 9.

12. The determination method by the determination unit, or the output mode by the output unit, is changed according to the age, gender, personality classification, and tendencies of the subject. The psychological state estimation device according to claim 6.

13. The system further includes a selection unit for selecting at least one of the following: the age, gender, personality classification, or tendency of the subject. The psychological state estimation device according to claim 12.

14. A first measuring device that measures the subject and obtains the first data, A second measuring device for measuring the subject and obtaining second data, A psychological state estimation device according to any one of claims 1 to 13 and A psychological state estimation system that possesses the following characteristics.

15. A first estimation step involves using first data obtained from measuring the subject to estimate the subject's psychological state based on a first classification model, A second estimation step involves using second data obtained from measuring the subject to estimate the subject's psychological state based on a second classification model, A conversion step that converts the first estimation result estimated in the first estimation step into the psychological state of the subject based on a second classification model, An output step that outputs information based on the first estimation result converted in the conversion step and the second estimation result estimated in the second estimation step, using a classification method corresponding to the second classification model. A method for estimating psychological states performed by a computer.

16. A first estimation step involves using first data obtained from measuring the subject to estimate the subject's psychological state based on a first classification model, A second estimation step involves using second data obtained from measuring the subject to estimate the subject's psychological state based on a second classification model, A conversion step that converts the first estimation result estimated in the first estimation step into the psychological state of the subject based on a second classification model, An output step that outputs information based on the first estimation result converted in the conversion step and the second estimation result estimated in the second estimation step, using a classification method corresponding to the second classification model. A psychological state estimation program designed to be executed by a computer.

Images