JP2011120824A - Sensibility evaluation device, sensibility evaluation method, and sensibility evaluation program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable sensibility to be quantitatively and accurately evaluated. <P>SOLUTION: When different stimulations corresponding to sensibility are given to a subject 50, biological information influencing on the sensibility is detected from a plurality of locations of the subject 50, each of the detected biological information is acquired as learning biological information, main component analysis is carried out based on the correlation matrix for the learning biological information, each of the learning biological information is integrated into one integrated data, the integrated data is plotted in a main component space, and regions corresponding to each of a plurality of sensibilities in main component space are set based on sensibility corresponding to stimulation given to the subject 50 when each of the distribution of the integrated data in the main component space and the biological information for learning being the bases of the integrated data is detected. Each of biological information for evaluating sensibility of the subject is acquired as target biological information and the integrated data obtained by similarly carrying out main component analysis for each of target biological information is plotted in the main component space in which the region is set in each sensibility to evaluate the sensibility of the subject 50. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a sensitivity evaluation apparatus, a sensitivity evaluation method, and a sensitivity evaluation program, and particularly relates to a sensitivity evaluation apparatus, a sensitivity evaluation method, and a sensitivity evaluation program for evaluating sensitivity based on biological information of a subject.

  Conventionally, the sensitivity of a subject has been evaluated using biological information acquired from the subject.

  For example, an apparatus comprising an infrared camera, an infrared device that digitally converts a thermal image from the infrared camera, and a personal computer that measures the amount of facial radiant heat based on data from the infrared device. There has been proposed an apparatus for estimating a person's psychological change by continuously measuring the amount of heat and comparing it with an average value (see, for example, Patent Document 1).

  In addition, the subject's image information is obtained, the subject's psychological state is classified, the subject's physiological information is obtained, and the subject is examined based on the physiological information and the psychological state classification result. There has been proposed a psychological state measuring device that calculates the level of the mental state of a person (see, for example, Patent Document 2).

  In addition, it receives input of a plurality of biological information, performs analysis processing based on the plurality of biological information, and is defined as emotion data such as a 2-axis model, a 3-axis model, a 6-axis model, or the like. An emotion visualization device that outputs a plurality of emotion parameter values has been proposed (see, for example, Patent Document 3). In the technique described in Patent Document 3, for example, a neural network method such as three-layer backpropagation that has been learned or re-learned individually for each user, or a self-organizing map (SOM) is used. Analyzing the emotions of the person who is the source of the biometric information from multiple types of biometric information received as input, and obtaining emotion parameter values as emotion data representing the emotions obtained as a result of the analysis Yes.

JP-A-6-54836 JP 2007-68620 A JP 2005-58449 A

  However, in the techniques of Patent Documents 1 and 2, the body surface temperature is used to evaluate the psychological state of the subject and its level, but as an evaluation standard, the average value or measurement of the body surface temperature of the subject. Since the temperature at the start is used, there is a problem that the sensitivity cannot be accurately evaluated.

  In addition, the technique of Patent Document 3 cannot be said to have sufficient accuracy for quantitatively evaluating sensitivity.

  The present invention has been made to solve the above-described problems, and an object thereof is to provide a sensitivity evaluation apparatus, a sensitivity evaluation method, and a sensitivity evaluation program capable of quantitatively and accurately evaluating sensitivity.

  In order to achieve the above object, the present inventors have conducted intensive studies, and as a result, Kansei evaluation has not been obtained only by comparison between individual parameters as in the prior art, and there are a large number of mutually related factors. It has been found that it can be visualized by new numerical dynamics obtained for the first time by integrating the biometric information of each other, and the present invention has been completed.

  That is, the sensitivity evaluation apparatus of the present invention includes a plurality of detection means for detecting biological information that affects the sensitivity of the subject from each of a plurality of portions of the subject, and a plurality of sensitivity to the subject. The biometric information detected by the detection means when each of the different stimuli corresponding to each is applied as biometric information for learning, and principal component analysis based on a correlation matrix is performed on the biometric information for learning, Projecting means for integrating each of the learning biometric information into one integrated data and projecting the integrated data onto a principal component space according to the principal component analysis, and the principal component space of the integrated data projected by the projection means Based on the above distribution and the sensibility corresponding to the stimulus given to the subject when each of the learning biometric information based on the integrated data is detected, the plurality of components in the principal component space Areas corresponding to each sensibility A principal component analysis based on a correlation matrix for each of the target biological information, with each of the biological information for setting the setting means for setting and the biological information for evaluating the sensitivity of the subject detected by the detection means as the target biological information And integrating each of the target biological information into one integrated data, projecting the integrated data onto the principal component space in which the area is set by the setting means, and the projected integrated data and the set Evaluation means for evaluating the sensitivity of the subject based on the region.

  According to the sensitivity evaluation apparatus of the present invention, the plurality of detection means detect biological information that affects the sensitivity of the subject from each of the plurality of portions of the subject. Examples of biological information that affects sensitivity include an electroencephalogram, an ocular potential, a heart rate, and a body surface temperature. Examples of detection means for detecting such biological information include an electroencephalogram measuring device, an eye myoelectric potential measuring device, a heart rate measuring device, a digital thermometer, and a thermography device.

  Then, when the projection means gives each of the different stimuli corresponding to each of the plurality of sensitivities to the subject, each of the biological information detected by the detection means is used as learning biological information, and the learning biological information The principal component analysis based on the correlation matrix is performed to integrate each of the learning biological information into one integrated data, and the integrated data is projected onto the principal component space corresponding to the principal component analysis. Thereby, each of the biometric information detected individually is re-synthesized into integrated data which is one index.

  And the setting means was given to the subject when the distribution of the integrated data projected by the projecting means on the principal component space and each of the learning biometric information based on the integrated data were detected. Based on the sensitivity corresponding to the stimulus, a region corresponding to each of the plurality of sensitivity is set in the principal component space. The distribution of the integrated data subjected to the principal component analysis is a region clearly classified for each sensitivity. A principal component space in which regions corresponding to each of a plurality of sensibilities are set as described above is used as a learning model for kansei evaluation.

  When evaluating the sensibility, the evaluation unit uses each piece of biological information for evaluating the sensibility of the subject detected by the detection unit as target biological information, and is based on a correlation matrix for each target biological information. Performs principal component analysis, integrates each target biological information into one integrated data, projects the integrated data onto the principal component space in which the area is set by the setting means, the projected integrated data, and the set area Based on the above, the sensitivity of the subject is evaluated.

  In this way, by projecting integrated data obtained by integrating each piece of biological information by principal component analysis based on a correlation matrix onto the principal component space, it is possible to set a clearly classified region for each sensitivity on the principal component space. Therefore, sensitivity can be evaluated quantitatively with high accuracy.

  The setting means can set an elliptical region corresponding to each of the plurality of sensibilities by a distributed ellipse approximation method.

  The plurality of detection means include a thermography device that detects body surface temperature from the plurality of parts of the subject as the biological information, and the evaluation means includes the body surface detected by the thermography device. The type of sensitivity can be evaluated using temperature. The principal component analysis of biological information revealed that the factor loading vector that extends in the direction separating the region set for each sensibility is due to the body surface temperature. The body surface temperature detected by is used. Thereby, it is possible to evaluate the type of sensitivity in a stress-free manner for the subject.

  And a selection unit for selecting a normal mode and a detailed mode for evaluating the sensitivity in more detail than the normal mode, wherein the plurality of detection units include a plurality of parts of the subject as the biological information. An electroencephalogram detector that detects an electroencephalogram from the evaluation means, and when the normal mode is selected by the selection means, the evaluation means uses the body surface temperature detected by the thermography device to When the type is evaluated and the detailed mode is selected, the type of sensibility using the brain wave detected by the electroencephalogram detector and the body surface temperature detected by the thermography device, and the sensitivity It is possible to evaluate such fluctuations.

  From the analysis results, the principal component analysis of biological information shows that the factor load vector that is orthogonal to the factor load vector of body surface temperature and that extends in the direction of minute fluctuations in sensitivity is the factor load vector of EEG As a result, when the detailed mode is selected, a minute change in sensitivity is evaluated using the brain wave detected by the brain wave detector. In the normal mode, similarly to the above, it is possible to evaluate the type of sensibility without stress using the body surface temperature detected by the thermographic device. In this way, by selecting the normal mode or the detailed mode, a flexible evaluation result can be obtained as necessary.

  The plurality of different sensitivities can be pleasant, uncomfortable, and contradictory. The sensitivity of “contradiction”, which is neither pleasant nor uncomfortable, or pleasant and uncomfortable, is clearly classified as “pleasant” and “unpleasant” in the principal component space. It can be set as an area.

  In addition, the sensitivity evaluation method of the present invention detects biological information that affects sensitivity by detecting a plurality of detection means from each of the plurality of portions of the subject, and corresponds to each of a plurality of sensitivities different to the subject. Each piece of biological information detected by the detecting means when a stimulus is applied is used as learning biological information, and a principal component analysis based on a correlation matrix is performed on the learning biological information, and each piece of biological information is set to 1 The integrated data is projected onto the principal component space corresponding to the principal component analysis, the distribution of the projected integrated data on the principal component space, and the learning based on the integrated data An area corresponding to each of the plurality of different sensibilities is set in the principal component space based on the sensibility corresponding to the stimulus given to the subject when each of the biological information is detected, and the detection Said test detected by means Each piece of biological information for evaluating the sensibility of the subject is set as target biological information, and a principal component analysis based on a correlation matrix is performed on each piece of the target biological information, and each piece of the target biological information is integrated into one integrated data. Then, the integrated data is projected onto the principal component space in which the area is set by the setting means, and the sensitivity of the subject is evaluated based on the projected integrated data and the set area. is there.

  Further, the sensitivity evaluation program of the present invention obtains the biological information detected by a plurality of detection means for detecting biological information affecting the sensitivity from each of the plurality of parts of the subject. Based on a correlation matrix with respect to the learning biological information, each of the biological information detected by the detecting means when giving a stimulus corresponding to each of a plurality of different sensibilities to the subject as learning biological information Performing principal component analysis, integrating each of the biological information into one integrated data, projecting the integrated data onto a principal component space according to the principal component analysis, integrated data projected by the projection unit Based on the sensitivity corresponding to the stimulus given to the subject when each of the learning biometric information based on the distribution of the principal component space and the integrated data is detected. component A setting means for setting a region corresponding to each of the plurality of different sensibilities in between, and each of the biological information for evaluating the sensitivity of the subject detected by the detection means as the target biological information, the target A principal component analysis based on a correlation matrix is performed on each piece of biological information, and each of the target biological information is integrated into one integrated data, and the integrated data is converted into a principal component space in which a region is set by the setting unit. This is a program for projecting and functioning as an evaluation unit for evaluating the sensitivity of the subject based on the projected integrated data and the set area.

  As described above, according to the sensitivity evaluation apparatus, sensitivity evaluation method, and sensitivity evaluation program of the present invention, integrated data obtained by integrating biological information by principal component analysis based on a correlation matrix is projected onto the principal component space. Thus, since an area clearly classified for each sensitivity can be set on the principal component space, it is possible to obtain an effect that the sensitivity can be quantitatively and accurately evaluated.

It is a block diagram which shows the structure of the sensitivity evaluation apparatus of this Embodiment. It is a flowchart which shows the content of the learning process routine in the sensitivity evaluation apparatus of this Embodiment. It is a figure for demonstrating the detection of the biometric information for learning. It is a figure which shows the stimulus corresponding to the sensitivity given to a subject, when detecting biometric information for learning. It is a graph which shows the result of carrying out the principal component analysis of several biological information. It is a flowchart which shows the content of the sensitivity evaluation process routine in the sensitivity evaluation apparatus of 1st Embodiment. It is a flowchart which shows the content of the sensitivity evaluation process routine in the sensitivity evaluation apparatus of 2nd Embodiment.

  Hereinafter, embodiments of a sensitivity evaluation apparatus according to the present invention will be described in detail with reference to the drawings.

  As shown in FIG. 1, the sensitivity evaluation apparatus 10 according to the first embodiment is configured to input various information including a detection unit 12 that detects biological information from a subject, a keyboard, a mouse, and the like. An operation unit 13, a display unit 14 for displaying a sensitivity evaluation result, and a computer 16 for executing sensitivity evaluation processing are provided.

  The computer 16 includes a CPU 20 that controls the sensitivity evaluation apparatus 10 as a whole, a ROM 22 as a storage medium that stores various programs such as learning processing and sensitivity evaluation processing described later, a RAM 24 that temporarily stores data as a work area, and various information Includes a hard disk (HDD) 26, an input / output port (I / O port) 28, a network interface (network I / F) 30, and a bus connecting them. A detection unit 12, an operation unit 13, and a display unit 14 are connected to the I / O port 28.

  The detecting unit 12 detects an electroencephalogram detector that detects an electroencephalogram from electrodes attached to a plurality of parts of the subject's scalp, and detects an ocular myoelectric potential from electrodes attached to the subject's right eyebrow and under the eye. Analyzes ocular myoelectric potential detector, heart rate detector that detects heart rate, digital thermometer that detects body surface temperature from thermocouple attached to a predetermined part of the subject, and images taken by thermography camera and thermography camera And a thermographic device that detects the body surface temperature of the subject. Data detected by the detection unit 12 is input to the computer 16 as digital data.

  Here, according to the international standard electroencephalogram electrode arrangement 10-20 method, 19 poles are arranged on the scalp of the subject. The digital thermometer is placed on the subject's wrist and the upper part of the clavicle.

  An electroencephalogram detector, an ocular potential detector, a heart rate detector, and a digital thermometer detect biological information by attaching electrodes or the like to a subject, and these are referred to as contact-type detectors. On the other hand, the thermography device does not require an electrode or the like to be attached to the subject, and this is called a non-contact detector.

  Next, with reference to FIG. 2, a learning processing routine in the sensitivity evaluation apparatus 10 according to the first embodiment will be described. In the present embodiment, a case where the types of sensibility are “pleasant”, “uncomfortable”, and “contradiction” will be described.

  In step 100, biometric information detected when various images and sounds are given to the subject as stimuli corresponding to each of “pleasant”, “unpleasant”, and “contradiction” is acquired as biometric information for learning. . Specifically, as shown in FIG. 3, the subject 50 is placed in front of a screen 52 that presents an image to the subject 50 and a speaker 54 that presents sound. About 30 edited by using information such as still images, images, music, environmental sounds, and the like that are expected and avoided by the subject 50 provided in advance by the subject 50 and other information that is expected to induce sensibility. Minute video and audio are presented to the subject 50 from the screen 52 and the speaker 54. The presented video and audio are edited so that the stimulus corresponding to “Pleasure”, the stimulus corresponding to “Uncomfortable”, and the stimulus corresponding to “Contradiction” are switched in units of several tens of seconds to several tens of seconds.

  As shown in FIG. 4, the stimulus corresponding to “pleasant” presents a video that is preferred from the screen 52 and a voice that is preferred from the speaker 54. The stimulus corresponding to “discomfort” presents a video to be avoided from the screen 52 and a voice to be avoided from the speaker 54. The stimulus corresponding to the “contradiction” presents a video that is preferred from the screen 52 and a voice to be avoided from the speaker 54 or a video that is to be avoided from the screen 52 and a voice that is preferred from the speaker 54. is there.

  As shown in FIG. 3, the scalp of the subject 50 is attached with a plurality of electrodes of the electroencephalogram detector 12a. Further, a thermography camera of the thermography device 12b is installed at a position where the subject 50 can be photographed, and the subject 50 is photographed. Furthermore, although illustration is omitted, the subject 50 is equipped with an ocular potential detector, a heart rate detector, and a digital thermometer. When the stimulus as described above is given to the subject 50, the biological information of the subject 50 is measured for 30 minutes by the detection unit 12. The detection interval of each biological information is set to an appropriate time interval depending on the characteristics of the biological information and the characteristics of the detector. For example, time intervals such as every 2 milliseconds for brain waves and every 10 seconds for body surface temperature can be set. Each piece of biological information detected in this way is acquired.

  Next, in step 102, the acquired biological information is processed. The brain wave is acquired as a 500 Hz digital signal, subjected to a 50 Hz low-pass filter process, and subjected to a fast Fourier transform (FFT) every 4.096 seconds, and the power spectrum is converted into frequency components (δ wave: 1-4, θ wave: 4- 8, α wave: 8-13, β wave: 13-40 [Hz]). The ocular myoelectric potential is acquired as a 500 Hz digital signal, and a power spectrum analysis is performed every 4 seconds. The heart rate is analyzed by power spectrum analysis of a high frequency component (HF: 0.15 Hz or more) and a low frequency component (LF: 0.04 to 0.15 Hz) by spectral analysis of RR interval (peak value interval) fluctuation. Deriving HF and LF / HF. The body surface temperature is calculated by linear interpolation between detection values at intervals of 10 seconds, for example, and matched with the time width (4 seconds) of processing such as brain waves.

  Next, in step 104, among the 30 minutes in which the biometric information is detected, a representative who has a sensibility of “pleasant”, “unpleasant” or “contradiction” is confirmed by a questionnaire response of 50 subjects after the fact. Each one minute of biological information is extracted, a principal component analysis based on a correlation matrix is performed on the extracted biological information, and each piece of biological information is integrated into one integrated data. And plot (project) them in the first and second principal component spaces of the principal component analysis as shown in FIG.

  In addition, you may make it use the biometric information of all the acquired time, without extracting the representative part of biometric information. Further, the state of the subject 50 who is detecting the biological information, that is, the state of the subject 50 when the stimulus is applied is recorded using a camera or the like, and each representative one minute is extracted. At this time, not only the questionnaire responses of 50 subjects but also the recorded state of the subject 50 may be confirmed, and the biological information of the time when the emotional expression is recognized may be extracted.

  Next, in step 106, an elliptical region for each sensitivity is set in the principal component space by the distributed ellipse approximation method. Specifically, the variance of the plotted integrated data on the principal component space is approximated by a variance ellipse. That is, a principal component analysis is performed using a variance-covariance matrix on the integrated data plot for each sensitivity, and the first eigenvector positive / negative direction obtained by multiplying the average of the distribution of the integrated data by the 1/2 power value of the first principal component eigenvalue. To obtain a dispersion ellipse whose major axis is both ends of the second principal component eigenvalue and the minor axis of both ends of the second eigenvector positive / negative direction multiplied by the ½ power value of the second principal component eigenvalue. To do. The dispersion ellipse approximated in this way is set in the principal component space as an area for each sensitivity, and the process ends. The principal component space in which an elliptical region is set for each sensitivity is called a sensitivity evaluation learning model.

  FIG. 5 is an example of an elliptical area for each sensitivity set in the principal component space by the processing according to the above routine. It can be seen that the elliptical areas of “pleasant”, “unpleasant” and “contradiction” are clearly classified on the principal component space. The relationship between the factor loading shown in FIG. 5 and the principal component space will be described later.

  Next, with reference to FIG. 6, a sensitivity evaluation processing routine in the sensitivity evaluation device 10 of the first exemplary embodiment will be described.

  In step 120, a plurality of pieces of biological information detected from the subject 50 are acquired as target biological information. The biological information is detected by the same process as step 100 of the learning process except that no known stimulus is given to the subject 50. The detection unit 12 may use all of the electroencephalogram detector, the ocular potential detector, the heart rate detector, the digital thermometer, and the thermography device used in the learning process, or select and use some of them. Also good. Particularly, since one electroencephalogram detector, one digital thermometer, and one thermography device can detect biological information of a plurality of parts of the subject 50, one each may be used. In addition, although mentioned later for details, it is preferable to use combining an electroencephalogram detector and a digital thermometer or a thermography apparatus.

  Next, in step 122, the acquired biological information is processed as in step 102 of the learning process.

  Next, in step 124, as in step 104 of the learning process, the biometric information obtained and processed is subjected to principal component analysis based on the correlation matrix, and each piece of biometric information is integrated into one integrated data. The integrated data is plotted on a sensitivity evaluation learning model (for example, FIG. 5) generated by the learning process.

  Next, in step 126, the sensitivity of the subject 50 is evaluated based on the plotted integrated data and the region set in the principal component space indicated by the sensitivity evaluation learning model, and the evaluation result is displayed on the display unit 14. indicate. For example, a graph as shown in FIG. 5 is output as an evaluation result, or “pleasant”, “unpleasant”, and “contradiction” are determined depending on which sensibility ellipse area the plotted integrated data is included in or close to. Either of them can be output. Further, the integrated data based on the biological information detected sequentially may be displayed on the graph shown in FIG. 5 every moment to evaluate the change in sensitivity.

  As described above, according to the sensibility evaluation apparatus 10 of the first embodiment, by plotting integrated data obtained by integrating each piece of biological information by principal component analysis based on a correlation matrix in a principal component space, Since an area clearly classified for each sensitivity can be set on the space, the sensitivity can be quantitatively evaluated with high accuracy.

  Next, a sensitivity evaluation apparatus 210 according to the second embodiment will be described. The second embodiment is different from the first embodiment in that a normal mode and a detailed mode in which sensitivity evaluation is performed in more detail than the normal mode can be selected. In addition, since the structure of the sensitivity evaluation apparatus 210 of 2nd Embodiment is the same as the structure of the sensitivity evaluation apparatus 10 of 1st Embodiment, it attaches | subjects the same code | symbol and abbreviate | omits description.

  First, the principle of the sensitivity evaluation process in the sensitivity evaluation device 210 of the second embodiment will be described.

  As shown in FIG. 5, the integrated data and settings plotted in the principal component space are compared with the factor loading amounts of each parameter contributing to the principal component space against the factor loading graph displayed as a vector extending from the average center. When the region for each sensitivity is considered, it can be seen that the factor load vector that extends in the direction of separating the region for each sensitivity is due to the body surface temperature detected by the thermography device 12b. On the other hand, it can be seen that the factor load vector extending in the direction orthogonal to the factor load vector of the body surface temperature is the factor load vector of the electroencephalogram. This is because it is effective to use body surface temperature as biometric information for the evaluation of sensibility (type of sensibility) in the order of minutes, such as separation expression of “pleasant”, “unpleasant” and “contradiction”. It is shown that it is effective to use an electroencephalogram as biological information for evaluation of fine sensibility (fine fluctuations in sensibility).

  Therefore, in the sensitivity evaluation apparatus 210 of the second embodiment, the normal mode and the detailed mode can be selected by setting the sensitivity evaluation of the minute order as the normal mode and the evaluation of the sensitivity as the second order as the detailed mode.

  Next, a sensitivity evaluation processing routine in the sensitivity evaluation apparatus 210 according to the second embodiment will be described with reference to FIG. Note that the learning process of the second embodiment is the same as the learning process of the first embodiment (FIG. 2), and thus the description thereof is omitted.

  In step 200, it is determined whether the normal mode or the detailed mode is selected as the sensitivity evaluation mode. This determination is made by selecting either the normal mode or the detailed mode by the operation unit 13 and setting it, and reading this setting. If the normal mode is selected, the process proceeds to step 202. If the detailed mode is selected, the process proceeds to step 212.

  In step 202, body surface temperatures detected from a plurality of parts of the subject 50 by the thermography device 12b are acquired as target biological information. Since the thermography device 12b is a non-contact detector, it can perform stress-free detection on the subject 50.

  Next, in step 204, principal component analysis based on a correlation matrix is performed on the acquired biological information, each of the biological information is integrated into one integrated data, and the integrated data is generated by the learning process. Plot to the learning model for evaluation.

  Next, in step 206, based on the plotted integrated data and the region set in the principal component space indicated by the sensitivity evaluation learning model, the plotted integrated data is “pleasant”, “unpleasant”, and “contradiction”. By determining which region belongs to or closest to which region, the type of sensitivity of the subject 50 is evaluated, the process proceeds to step 210, and the evaluation result is output to the display unit 14. .

  On the other hand, if it is determined that the detailed mode is selected and the process proceeds to step 212, the body surface temperature detected from the plurality of parts of the subject 50 by the thermography device 12b and the brain wave detector Brain waves detected from a plurality of parts of the person 50 are acquired as target biological information.

  Next, in step 214, the acquired biological information is subjected to principal component analysis based on a correlation matrix, and each piece of biological information is integrated into one integrated data, and the integrated data is generated by the learning process. Plot to the learning model for evaluation.

  Next, in step 216, based on the plotted integrated data and the region set in the principal component space indicated by the sensitivity evaluation learning model, the plotted integrated data is “pleasant”, “unpleasant”, and “contradiction”. By determining which region belongs to or closest to which region, the type of sensitivity of the subject 50 is evaluated, and minute variations in the sensitivity of the subject 50 are evaluated. The minute fluctuation of sensitivity is a change in the direction in which the factor loading vector of the electroencephalogram extends in the elliptical region for each sensitivity or in the vicinity thereof. For example, a change in the factor load vector direction of the EEG within the elliptical area of the sensitivity “pleasant” may indicate the degree (level) of “pleasant”, but simply the intensity of the “pleasant” degree It does not necessarily represent high or low. Therefore, here, the change in sensitivity in the direction in which the factor load vector of the electroencephalogram is extended is treated as representing a finer sensitivity fluctuation in a certain sensitivity. And it transfers to step 210 and outputs an evaluation result to the display part 14. FIG.

  As described above, according to the sensitivity evaluation apparatus 210 of the second embodiment, the sensitivity evaluation based on the body surface temperature that is effective for evaluating the sensitivity of the minute order is set to the normal mode, and effective for the evaluation of the sensitivity of the second order. When the normal mode is selected as the detailed mode, the sensitivity type can be evaluated in a stress-free state. When the detailed mode is selected, a fine mode is selected. Since the change in sensitivity can be evaluated, a flexible evaluation result can be obtained as necessary.

  In the second embodiment, the case where the normal mode and the detailed mode can be selected has been described. However, only one of the modes may be configured.

  In the second embodiment, the case where both the body surface temperature detected by the thermographic device 12b and the electroencephalogram detected by the electroencephalogram detector 12a are used in the detailed mode is described. However, only the electroencephalogram is used. Also good.

  In the first and second embodiments, the case where brain waves, eye muscle potentials, heartbeats, and body surface temperature are used as the biological information for learning has been described. However, it is not necessary to use all the biological information. Moreover, it is not limited to these biological information, You may use the other biological information which affects a sensitivity. However, the direction of the factor loading vector in the principal component analysis is used in combination with the biological information in which the direction is separated for each type of sensibility and the biological information in which the direction of sensibility is in the direction perpendicular to the direction Is preferred.

  In the first and second embodiments, the case where the evaluation result is output and displayed on the display unit has been described. However, the evaluation result is output to an external device connected via the network I / F 30. May be. For example, an evaluation result may be output to an external device such as an air conditioner or a lighting device control device, and temperature control or dimming control may be performed according to the sensitivity of the subject 50 based on the evaluation result. Good.

  In the first and second embodiments, the case where the region for each sensitivity is set using the distributed ellipse approximation method has been described. However, the present invention is not limited to this. For example, a method of setting a boundary line in which the principal component space is classified for each sensitivity may be used.

DESCRIPTION OF SYMBOLS 10,210 Sensitivity evaluation apparatus 12 Detection part 12a EEG detector 12b Thermography apparatus 13 Operation part 14 Display part 16 Computer

Claims (7)

A plurality of detection means for detecting biological information that affects the sensitivity of the subject from each of the plurality of portions of the subject;
Each of the biometric information detected by the detection means when giving each of different stimuli corresponding to each of a plurality of sensibilities to the subject as biometric information for learning is used as a correlation matrix for the biometric information for learning. Performing projection based on the principal component analysis, integrating each of the learning biometric information into one integrated data, and projecting the integrated data onto a principal component space according to the principal component analysis;
The stimulus given to the subject when the distribution of the integrated data projected by the projection means on the principal component space and each of the learning biometric information based on the integrated data are detected. Setting means for setting a region corresponding to each of the plurality of sensibilities in the principal component space based on the corresponding sensibilities;
Using each of the biological information for evaluating the sensitivity of the subject detected by the detection means as target biological information, principal component analysis based on a correlation matrix is performed on each of the target biological information, and the target Each of the biological information is integrated into one integrated data, the integrated data is projected onto the principal component space in which the area is set by the setting unit, and based on the projected integrated data and the set area, An evaluation means for evaluating the sensitivity of the subject;
Kansei evaluation device.   The sensitivity evaluation apparatus according to claim 1, wherein the setting means sets an elliptical region corresponding to each of the plurality of sensitivities by a distributed ellipse approximation method. The plurality of detection means include a thermography device that detects body surface temperature from a plurality of parts of the subject as the biological information,
The sensitivity evaluation apparatus according to claim 1, wherein the evaluation unit evaluates the type of sensitivity using a body surface temperature detected by the thermography apparatus. Selecting means for selecting a normal mode and a detailed mode for evaluating the sensitivity in more detail than the normal mode;
The plurality of detection means further includes an electroencephalogram detector that detects electroencephalograms from a plurality of parts of the subject as the biological information,
When the normal mode is selected by the selection unit, the evaluation unit evaluates the type of sensitivity using the body surface temperature detected by the thermographic device, and when the detailed mode is selected. The sensibility evaluation apparatus according to claim 3, wherein the type of sensibility and minute fluctuations of the sensibility are evaluated using the electroencephalogram detected by the electroencephalogram detector and the body surface temperature detected by the thermography device.   The sensitivity evaluation apparatus according to claim 1, wherein the plurality of different sensitivities are defined as pleasant, unpleasant, and contradiction. Detecting biological information affecting sensibility from each of a plurality of parts of the subject by a plurality of detection means,
Each of the biological information detected by the detection means when a stimulus corresponding to each of a plurality of different sensibilities is given to the subject is used as learning biological information. Perform component analysis, integrate each of the biological information into one integrated data, project the integrated data onto the principal component space according to the principal component analysis,
The sensitivity corresponding to the stimulus given to the subject when the distribution of the projected integrated data in the principal component space and each of the learning biometric information based on the integrated data are detected Based on the principal component space, a region corresponding to each of the plurality of different sensibilities,
Using each of the biological information for evaluating the sensitivity of the subject detected by the detection means as target biological information, principal component analysis based on a correlation matrix is performed on each of the target biological information, and the target Each of the biological information is integrated into one integrated data, the integrated data is projected onto the principal component space in which the area is set by the setting unit, and based on the projected integrated data and the set area, A sensitivity evaluation method for evaluating the sensitivity of the subject. Computer
Obtaining means for obtaining the biological information detected by a plurality of detecting means for detecting biological information affecting sensibility from each of the plurality of parts of the subject;
Each of the biological information detected by the detection means when a stimulus corresponding to each of a plurality of different sensibilities is given to the subject is used as learning biological information. Projecting means for performing component analysis, integrating each of the biological information into one integrated data, and projecting the integrated data onto a principal component space according to the principal component analysis;
The stimulus given to the subject when the distribution of the integrated data projected by the projection means on the principal component space and each of the learning biometric information based on the integrated data are detected. Setting means for setting regions corresponding to each of the plurality of different sensitivities in the principal component space based on the corresponding sensibilities, and biological information for evaluating the sensibility of the subject detected by the detection means Each of the target biometric information is subjected to principal component analysis based on a correlation matrix for each of the target biometric information, each of the target biometric information is integrated into one integrated data, and the integrated data is set as the setting means A sensitivity evaluation for functioning as an evaluation means for evaluating the sensitivity of the subject based on the integrated data thus projected and the set area. Program.

Images