JP2000194252A - Ideal action support device, and method, system, and recording medium therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make it possible for an examinee him/herself to bring own speech and behavior close to those of an ideal person as a model very easily without always minding a video camera and a microphone, and completely confirming recorded own speech and behavior from the beginning to the end and to confirm the own unsuitable speech and behavior on the real time. SOLUTION: This ideal action support method derives and stores (step S502-S504) parameters based on a voice, a dynamic scene, and a physiologic data according to the speech and behavior of an examinee, and when the device is set to an evaluation mode (step S505), it compares the parameter concerning the speech and behavior of the examinee derived and stored with the parameter concerning the speech and behavior of the model character set in the set mode (step S506), and based on the comparison result, the device generates an advice concerning the improvement of the speech and behavior for approaching to the model character to output it (step S507-S508).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an ideal behavior supporting apparatus for recognizing a subject's behavior and giving appropriate instructions to the subject, thereby improving the behavior of the subject, and an ideal behavior support apparatus therefor. The present invention relates to a method and a system thereof, and a storage medium.

[0002]

2. Description of the Related Art Conventionally, in order for a subject to observe his / her own words and actions objectively, a video camera is set up by himself / herself and an action is taken in front of the video camera to record his / her own actions. Then, it was necessary to play back the video with a video playback device or the like later and check his / her own action.

[0003] If a subject wants to view audio information such as his / her own speech objectively, he or she installs a recording device such as a microphone and a tape recorder by himself, and speaks around the microphone. It was necessary to check his voice by recording his voice on a tape recorder or magnetic disk and playing it back later.

[0004]

However, when recording an image of a person who is taking an action or recording the content of his / her own voice by using such a method, the subject himself / herself must use his / her own voice. There is a high possibility that a video camera or a microphone installed in front of the subject is always conscious of actions or remarks, and there is a disadvantage that it is difficult to record the unconscious speech of the subject.

[0005] In particular, since the behavior outside the imaging range of the video camera and the voice outside the sound collection range of the microphone are not recorded, the subject can use the video camera or the video camera installed when taking an action or speaking. I had to be very careful not to stay away from the microphone.

[0006] In addition, while a video playback device or a tape recorder is playing back video or audio, the subject must always pay attention to his or her own reproducible behavior.
When the recorded contents were long, the subject was imposed a great burden.

[0007] In addition, since the subject cannot compare and contrast the behavior of the ideal person as a model with his / her own behavior, how can the subject approach his / her behavior to that of the ideal person. I didn't know how to know.

[0008] In addition, when a subject makes an inappropriate speech, the recorded video or sound can be confirmed when the recorded video or sound is reproduced. However, the subject himself could not confirm inappropriate behavior.

[0009] In view of the above problems, the present invention provides a method in which the subject himself or herself is always conscious of a video camera or a microphone.
It is very easy to bring your behavior close to the ideal model person without having to check the entire recorded behavior from the beginning to the end, and to see your inappropriate behavior in real time. It is an object of the present invention to provide an ideal behavior support device, a method and a system thereof, and a storage medium that can be confirmed.

[0010]

In order to achieve the above object, an ideal action support device according to the present invention comprises a storage means for storing first information relating to the behavior of a first person, and a second person. Input means for inputting second information relating to the behavior of the person;
Comparing means for comparing the first information stored in the storage means with the second information input by the input means;
Notification means is provided for notifying the second person of advice regarding improvement of behavior to approach the first person based on a comparison result by the comparison means.

[0011] In order to achieve the above object, an ideal behavior support method according to the present invention comprises an input step of inputting first information relating to the behavior of a first person, and a second step stored in a storage means. A comparing step of comparing the second information on the behavior of the person with the first information input in the input step,
And a notifying step of notifying the first person of advice regarding improvement in behavior to approach the second person based on a result of the comparison in the comparing step.

In order to achieve the above object, an ideal behavior support system according to the present invention is an ideal behavior support system comprising a portable terminal device and an information processing device,
The portable terminal device has input means for inputting first information related to the behavior of a certain person, and transmission means for transmitting the first information input by the input means to the information processing apparatus, The information processing device includes a storage unit that stores various people and information about the behavior of the various people, and a selection that selects, from the storage unit, a specific person and second information about the behavior of the specific person. Means, receiving means for receiving the first information transmitted from the portable terminal device, first information received by the receiving means,
Comparing means for comparing the second information selected by the selecting means with the second information, based on a comparison result by the comparing means,
And notifying means for notifying the certain person of advice regarding improvement in behavior for approaching the specific person.

[0013] In order to achieve the above object, a storage medium according to the present invention comprises an input module for inputting first information relating to the behavior of a first person, and a second module stored in storage means. A comparison module for comparing the second information on the behavior of the person with the first information input by the input module, and the first module based on a comparison result by the comparison module. A computer program having a notification module for notifying a person of advice regarding improvement of behavior in approaching the second person is stored.

[0014]

Embodiments of the present invention will be described below with reference to the drawings.

(First Embodiment) As a first embodiment of the present invention, a case will be described in which a subject actually uses a wristwatch-type portable information terminal having a function of supporting ideal behavior while also serving as a wristwatch. I do.

[Outline of Wristwatch-Type Portable Information Terminal] First, a wristwatch-type portable information terminal according to the first embodiment will be described in detail with reference to FIGS.

FIG. 1 is an external view of the front side of a wristwatch-type portable information terminal according to the first embodiment.

Reference numeral 1 denotes an ultra-small and wide-angle CCD camera which is used to image a subject and is detachable from the main body of the portable information terminal.
The video is sent to the main body wirelessly. Reference numeral 2 denotes a microphone, which is used to collect a subject's voice and surrounding sounds. Reference numeral 3 denotes a speaker, which is used to notify a test subject of a warning sound, a warning message, and the like. Reference numeral 4 denotes a liquid crystal display, which normally displays the current time in a clock mode.

Reference numeral 5 denotes a clock adjustment button. By operating this button, the current time displayed on the liquid crystal display 4 is adjusted. A mode button 6 is pressed to switch from the clock mode to another mode. Reference numeral 7 denotes a select button, which is displayed on the liquid crystal display 4 and depressed when any candidate needs to be determined. Reference numeral 8 denotes an arrow button, which is displayed on the liquid crystal display 4, and when there are a plurality of candidates and one of them is selected, a button having a direction corresponding to the moving direction at that time is pressed.

Reference numeral 9 denotes a data card slot, into which various data cards can be inserted.

FIG. 2 is an external view of the back side of the wristwatch-type portable information terminal according to the first embodiment.

Reference numeral 10 denotes a perspiration sensor, which is used to measure the amount of perspiration of the subject. Reference numeral 11 denotes a heart rate blood pressure sensor, which is used to measure a subject's heart rate, maximum blood pressure, minimum blood pressure, and the like.

FIG. 3 is a hardware connection diagram inside the wristwatch portable information terminal according to the first embodiment.

Reference numeral 301 denotes a CPU which performs various controls on each hardware. 302 is a RAM,
Used as a work area during processing by the CPU 301,
Various data are temporarily stored. The internal configuration of the RAM 302 will be described in detail with reference to FIG. Reference numeral 303 denotes a ROM, which stores a computer program for implementing the processing according to the first embodiment. Reference numeral 304 denotes an HDD, which stores various databases necessary for processing according to the first embodiment. The internal configuration of the HDD 304 will be described in detail with reference to FIG.

Reference numeral 305 denotes a camera control unit.
The operation of the CD camera 1 is controlled. In addition, CCD camera 1
When using the camera separately from the main body, the camera control unit 30
Information is exchanged between the camera 5 and the CCD camera 1 via a communication interface (not shown).

Reference numeral 306 denotes a microphone control unit.
Control the operation of. A speaker control unit 307 controls the operation of the speaker 3. A display control unit 308 controls the operation of the liquid crystal display 4. A button control unit 309 controls various buttons such as a clock setting button 5, a mode button 6, a select button 7, and an arrow button 8. Reference numeral 310 denotes a sensor control unit 310 which controls the operation of the perspiration sensor 10 and the heart rate blood pressure sensor.

Reference numeral 311 denotes a data card control unit, which is a data card 312 inserted into the data card slot 9.
Control related to When the data card 312 is inserted into the data card slot 9, the RAM 302 or the HDD 3
04 to the data card 312, and the data in the wristwatch-type portable information terminal can be externally ported.

Reference numeral 313 denotes a clock control unit which controls a built-in clock 314. And 315 is
A bus that interconnects the above hardware.

FIG. 4 is a memory map inside the RAM 302.

Reference numeral 3020 denotes a model parameter storage area which stores a model person described later and parameter values of each evaluation item of the person. Reference numeral 3021 denotes a voice storage area for storing voice. Reference numeral 3022 denotes an image data storage area for storing image data. 3023
Is a physiological data storage area for storing various physiological data. Reference numeral 3024 denotes a voice conversion character string storage area which stores a character string when voice is converted to a character string.
Reference numeral 3025 denotes a quantified data storage area for storing data quantified from the subject's behavior. A subject parameter storage area 3026 stores parameter values obtained based on the subject's behavior.

FIG. 5 is a memory map inside the HDD 304.

Reference numeral 3040 denotes a model setting database, which stores a plurality of model persons, which will be described later, and parameter values of each person in each evaluation item. 30
Reference numeral 41 denotes an honorific operation knowledge database, which stores a group of information describing human relationships necessary for operating honorific words for each conversation scene. Reference numeral 3042 denotes an honorific grammar knowledge database, which stores a group of information in which the rewritable, respectable, modest, and polite forms of the rewritable word are described for each word. Reference numeral 3043 denotes a compliment database in which phrases for complimenting the other party are stored.

Reference numeral 3044 denotes a whole-body model image database, which stores images of various posture models of the whole human body. Reference numeral 3045 denotes a face model database in which images of various contour models of a human face are stored. Reference numeral 3046 denotes an eye part image database in which various model images of the eye part of the human face are stored.
Reference numeral 3047 denotes a mouth part image database which stores various model images of a mouth part of a human face.

Reference numeral 3048 denotes an operation voice-related database which stores information in which a human operation and its voice are set. Reference numeral 3049 denotes a physical condition determination knowledge database, which stores a group of information for determining the physical condition of the subject by determining the physical condition, posture, movement, facial expression, and the like of the subject in combination. Have been.

Reference numeral 3050 denotes an advice-giving knowledge database, which stores information in which the absolute value of the difference between the behavior of the model person and the subject and the advice to the subject are set. Reference numeral 3051 denotes a phrase database that stores words of the best phrase and words of the worst phrase, and is rewritable.

[Setting of Ideal Model Person] Next, for such a wristwatch-type portable information terminal,
The procedure for setting an ideal person to be a model of a subject will be described.

First, the mode button 6 is pressed until the model setting mode is set. In the model setting mode, “model setting” is displayed at the upper left of the liquid crystal display 4 as shown in FIG.

Hereinafter, the subsequent operation procedure will be described with reference to FIG.

First, in the model setting mode,
When the down arrow button in the arrow buttons 8 is pressed, two categories of “model person” and “original person” are displayed under “model setting” on the liquid crystal display 4. When "model person" is selected, one of the preset persons is selected, and when "original person" is selected, a completely original person ideal for the subject can be set. For now, here, "model person" is selected.

When "model person" is selected, two categories of "male" and "female" are displayed. Here, when the "women" is selected by operating the arrow buttons 8, categories such as "historical figures", "entertainers", "other celebrities", and "fictional figures" are displayed on the right side of "woman". Is displayed. Here, when you select "History person", the right side displays the choices of specific people such as "Seijin no Shonen", "Murasaki Shikibu", "Nightingale", "Jeanne d'Arc", "Mother Teresa" Is done. Here, when the down arrow button in the arrow buttons 8 is pressed several times, “Mother Teresa” is selected.

Next, when the user releases the arrow button 8 with "Mother Teresa" selected, a display as shown in FIG. 8 appears. This is a display that prompts the subject whether the selection target is “Mother Teresa” or not, “OK”,
Choices of “cancel”, “option”, and “help” are displayed, and “OK” is selected in the initial state.
Here, if the selection target as the model person is “Mother Teresa”, the subject presses the select button 7 as it is. By pressing the select button 7, the ideal model person is set as “Mother Teresa”. After the setting is confirmed, the mode automatically returns to the clock mode, and the current time is displayed again on the liquid crystal display.

When the subject wants to cancel, the user may press the right arrow button of the arrow buttons 8 once to select "cancel" and then press the select button 7;
When it is desired to change the value of a parameter given in advance for “Mother Teresa”, the user may press the right arrow button twice, select “option”, and press the select button 7. The processing when this “option” is selected will be described later in detail. To request help, the user may select "Help" one position to the right.

The information on the model person displayed in FIG. 7 is stored in the model setting database 3 in the HDD 304.
040 is stored in advance. That is, the “Mother Teresa” selected this time and the information related thereto are stored in the model setting database 3040 in advance. Of course, information may be obtained from outside via the data card 3312.

In the model setting database 3040, evaluation items and parameters as shown in Table 1 below are set in advance for the behavior of each model person.

[0045]

[Table 1]

In other words, the evaluation items relating to the behavior of each person can be roughly divided into “speech”, “personal”,
It is related to three states of "body state". For example, "speaking style" includes "volume", "range", "speed", "smoothness", "attentiveness", "honorifics", and "praise degree""," Coolness "," Strength of pushing ", and" Good listening ". Of course, what is shown in Table 1 is only an example of the evaluation items, and there may be evaluation items related to “how to laugh” and “how to eat”.

A certain value is set for each evaluation item. For example, in the case of “volume”, “a large”, “b slightly large”, “c normal”,
Any value of “d is slightly smaller” or “e is smaller” is given. Of course, for other items, similarly, any value is given as a parameter.

Of course, these values for the model person are given as superficial images and the like, and do not always correspond to the real image of the person. If you feel uncomfortable with the values of the parameters you are using,
The parameter may be changed.

In such a case, however, an ideal person to be a model is selected, and when a model confirmation screen as shown in FIG. 8 is displayed, the select button 7 is not pressed and the arrow button 8 is pressed. Is pressed twice, to select "option", and then the select button 7 is pressed. Then, a parameter setting screen as shown in FIG. 9 appears on the liquid crystal display 4. Thereafter, the value of the parameter given in advance may be changed according to the instruction on the liquid crystal display 4. In the case of FIG. 9, “volume” is selected as an example, and the value of “d is slightly lower” is given in advance, so if you want to make a change, press the arrow button 8 to change the setting. I just need.

When the user enters the model setting mode and selects "original person" instead of "model person", a parameter setting screen similar to that shown in FIG. 9 appears, and the user can freely set values relating to his or her ideal behavior. Can be set to

In the above-described manner, an ideal person serving as a model of a subject and parameters of the person in each evaluation item can be set in the wristwatch-type portable information terminal.

Needless to say, even after the model has been set once, the model can be reset at any time by returning to the model setting mode.

[Outline of Processing by Ideal Behavior Support Program] Next, as a result of a subject wearing a wristwatch-type portable information terminal as described above actually acting or speaking, An outline of the process of whether advice for supporting ideal behavior is output from the terminal will be described.

FIG. 10 is a flowchart of the ideal behavior support processing according to the first embodiment.

A computer program for realizing the flowchart in FIG. 10 is stored in the ROM 303 and is processed by the CPU 301.

First, in step S501, it is determined whether or not the mode of the wristwatch type portable information terminal is the model setting mode. In the case of the model setting mode, since the model setting is still being performed, the process waits until the mode is changed. If the mode is not the model setting mode, it means that the model setting has already been performed.
In step S502, a voice, a moving image, physiological data, and the like obtained by the subject's behavior are stored in the RAM 302.

Next, the process proceeds to step S503, in which the values of the parameters of each evaluation item are derived based on the voice, moving image, and physiological data obtained in step S502.
At 504, the value of the derived parameter is
It is stored in M302.

Then, in step S505, it is determined whether or not the mode of the wristwatch-type portable information terminal has changed to the evaluation mode. If the mode has not been changed to the evaluation mode, it is still necessary to accumulate information.
Returning to 502, a voice, a moving image, physiological data, and the like obtained by the subject's behavior are stored in the RAM 302. When the mode is changed to the evaluation mode, the next step S5
The process proceeds to step 06 onward to perform a process of evaluating the subject's behavior.

First, in step S506, the parameters of each evaluation item of the model person set by the method described above are compared with the parameters of each evaluation item of the subject stored in the RAM 302 in step S504. Then, the process proceeds to step S507, where advice for the subject is generated based on the comparison result obtained in step S506, and finally, in step S508, step S50 is performed.
The advice generated in step 7 is output, and the series of processing ends.

In step S502, even when all of the voice, moving image, and physiological data cannot be obtained from the subject's behavior, the subsequent processing can be performed using only the obtained information.

[Regarding Recording and Parameterization of Subject's Behavior] Next, the processing in steps S502 and S503 in the flowchart shown in FIG. 10 will be described in more detail.

In these steps, in short, when the subject wearing the wristwatch-type portable information terminal acts or speaks, the behavior is recorded, and the speech is recorded. A process of converting the recorded voice into parameter values of the evaluation items as shown in Table 1 is performed.

In the following, description will be made in three cases, namely, "speaking style", "self-care", and "body state".

<How to Speak> FIG. 11 is a functional block diagram of a process relating to “how to speak” in the CPU 301.

Referring to FIG. 11, the processing relating to "how to speak" according to the first embodiment will be described in the order of processing.

First, the sound input from the microphone 2 is R
The voice stored in the voice storage area 3021 in the AM 302 is stored in the voice storage area 3021.
Is recognized by the voice separation function 80
According to 2, the voice is separated into three voices: voice a of the subject himself, voice b of a person other than the subject, and voice c other than the voice. Then, the separated voice a of the subject and the voice b of the person other than the subject are analyzed by the voice analysis function 803.
As a result, values such as the volume, the sound range, the speed, and the amplitude of the wavelength of the voice a of the subject and the voice b of the person other than the subject are obtained. Then, the parameter setting function 804 obtains, based on the result concerning the voice a of the subject, 1-1 volume, 1-2 range, 1-3 speed, 1-4 smoothness, 1-smoothness,
8 is set, and a parameter of 1-10 is good from the value of the voice b of the person other than the subject.
The parameter of -9 pressing strength is set. Note that the other voice c is sent to the operation sound detection function 1021 described later, and is not used for parameter derivation here.

Further, with respect to the voice a of the subject, after being separated by the voice separating function 802, the interpersonal conversation ability analysis function 805 analyzes whether or not the subject uses a conversation expression suitable for the interpersonal conversation. Then, from the analysis result, parameters of 1-5 politeness, 1-6 honorifics, and 1-7 praise are set by the parameter setting function 804.

The interpersonal conversation ability determination function 805 will be described in more detail with reference to FIG.

First, the voice a of the subject himself is converted into character information by the language processing function 901. This character information is understood by the utterance understanding function 902, and the understood contents are analyzed by the dialogue management function 903 to generate response contents. This response content is converted into a sentence with a correct expression by the honorific generation function 904.

Next, a character string containing the subject's voice a converted into character information by the language processing function 901 and a correct expression understood by the utterance understanding function 902 and converted by the honorific word generation function 904 are provided. The evaluation function 905 compares and evaluates the character string having the utterance as the content.

The processing by the evaluation function 905 is performed by the HDD 30
4 and a human relations knowledge database 30 describing the human relationships necessary to use the honorifics for each conversation scene.
41 and an honorific grammar knowledge database 304 in which the rewritable, respectable, humble and polite forms of the rewritable word are described for each word.
2 and a praise database 3043 that describes phrases that praise the other party.

More specifically, the number of honorific words and the number of praise words appearing in the utterance of the subject are counted, and errors such as nonuse or misuse of honorific words, and mistakes between honorific words and humble words are found. By counting them, the interpersonal conversation ability of the subject is evaluated.

The parameter setting function 804 sets the parameters of 1-5 politeness, 1-6 honorifics, and 1-7 praise, based on the contents analyzed through such a series of processing. .

FIG. 13 is a functional block diagram of the "body-care" process in the CPU 301.

Hereinafter, with reference to FIG. 13, the processing relating to “self-care” according to the first embodiment will be described in the order of processing.

First, at the place where the camera 1 is installed, when the subject is sufficiently visible from the wristwatch-type portable information terminal,
Use it while attached to the information terminal, and when it is not in a position where the subject can be seen sufficiently, remove it from the information terminal and install it in a position where the subject can be grasped. The video in that case is wirelessly sent from the camera 1 to the information terminal.

As described above, such a camera 1 is very small, so that the subject hardly needs to be aware of the existence thereof. Subject can be captured. Of course, a function of tracking the subject may be provided so that the camera 1 is always moved so as to face the subject.

The moving image captured by the camera 1 is stored in an image storage area 3022 in the RAM 302.
The motion detection function 1001 of the CPU 301 detects the posture and the motion of the subject as a whole from the moving image stored in the image storage area 3022.
11 detects the facial expression of the subject. The information on the posture, the motion, and the expression of the subject obtained by the motion detecting function 1001 and the facial expression detecting function 1011 are stored in the parameter setting function 804 and the physiological condition detecting function 1 described later.
301.

The posture and the motion of the subject obtained by the motion detecting function 1001 are the same as those of the voice separating function 10 described above.
21 and sent to the operation sound detection function 1021, where the operation sound detection function 1021
Detects only the voice related to the motion of the subject and sends it to the parameter setting function 804.

The parameter setting function 804 sets the parameters of the evaluation items related to “personal care”.

FIG. 14 shows the operation detecting function 1 in FIG.
001, facial expression detection function 1011 and operation sound detection function 102
1 is described in further detail.

First, the operation detecting function 1001 will be described.

First, a difference image is generated by the difference image generation function 1002 from the image stored in the image storage area 3022. Next, the labeling function 1003 labels the area of the moving object. Then, an edge extraction function 1004 newly performs edge extraction from the original image for each of the labeled regions. next,
From the obtained edge image, the contour information generation function 10
05, the contour information is obtained and stored in the whole body model database 3044 in the HDD 304, for example, as shown in FIG.
The model matching function 1006 performs matching using a three-dimensional model as shown in FIG. 5 to determine the posture of the subject.

Here, as shown in FIG. 16, the bending angles of the joints such as the movements of the hands, feet, neck, etc., the rotation in the body axis direction, the bending of the waist, and the like are represented by the angles in the three-dimensional space. No, 2
It is derived as an angle when projected on a dimensional plane (image). In FIG. 16, the angle from θa to θf is derived according to the motion of the subject. Here, the reason for deriving the angle when projected on the two-dimensional plane is that only one camera is installed, and if a plurality of cameras are installed, It is also possible to derive an angle in a three-dimensional space.

From the angles thus derived, the hand,
The magnitude of the motion of each part of the body such as the feet, neck, torso, and waist can be obtained. Furthermore, the speed of the operation can be obtained from the correlation between the time from the start to the end of the operation and the derived angle, and the change in the operation per unit time can be calculated sequentially to obtain the operation speed. Can be obtained. Here, the wristwatch-type portable information terminal may be provided with an acceleration sensor to measure the acceleration.

Then, based on the data derived through such a series of processing, the parameter setting function 804 sets the size of the 2-1 operation, 2-2 speed, 2-3 elegance (smoothness), Is set.

Next, the expression detection function 1011 will be described.

By detecting the posture by the model collation function 1006 described above, it is possible to assume where in the acquired image the face of the subject exists.

Therefore, first, from the image stored in the image storage area 3022, the head is estimated from the posture detected by the model collation function 1006, and the face part extraction function 1012
Generate a partial image. Next, the hue / saturation image generation function 1013 creates a hue / saturation image for the partial image, and detects a flesh color area. Further, in order to determine whether the detected area is a face, the face part extraction function 1014 performs matching based on the face model stored in the face model database 3045 in the HDD 304. And
By using the position and color information, the face part extraction function 1015 searches for a characteristic face part such as an eye or a mouth.

FIG. 17 is a schematic diagram showing the image data of each part of the face, eyes, and mouth extracted by the above-described method.

In the case of a dark room, an infrared camera is used. That is, color information cannot be used. Therefore, detection using shape information and luminance information is performed.

Since the eyes and the mouth are extracted in this manner, the movements of the eyes and the mouth are observed by the eye part observing function 1016 and the mouth part observing function 1017.

The eye part observing function 1016 is a function for observing the subject's eyes open, closed, angry, laughing, crying, and the like. Specifically, since the position of the eyes is known by the face part extraction function 1015, a rectangular partial image trimmed to include the eyes is created from the image in the image storage area 3022, and the eyes in the HDD 304 are stored. Partial image database 3046
The emotional state of the eyes in the created partial image is determined with reference to the eye partial image data.

The mouth part observing function 1017 is a function for observing the subject's mouth talking, angry, laughing, crying, yawning, and the like. Specifically, since the position of the mouth is known by the face part detection function 1015, a rectangular partial image trimmed to include the mouth is generated from the image in the image storage area 3022, and the mouth in the HDD 304 is generated. The emotional state of the mouth in the created partial image is determined with reference to the lip shape model in the partial image database 3047.

The expression determining function 1018 comprehensively expresses the subject's facial expression based on the emotional state of the eyes determined by the eye part observing function 1016 and the emotional state of the eyes determined by the mouth part observing function 1017. Is determined.

The parameter setting function 804 sets the parameters of the 2-5 expression based on the data derived through such a series of processing.

If the eye or mouth portion does not exist in the image, or if the eye or mouth is hidden by something else and cannot be detected, only the processing of that portion is not performed.

Finally, the operation sound detection function 1021 will be described.

The voice stored in the voice storage area 3021 in the RAM 302 is separated by the voice separation function 802 into three voices: voice a of the subject himself, voice b of a person other than the subject, and voice c other than the voice. As described above, in the operation sound detection function 1021, first, the operation sound extraction function 802 stores the operation sound related database in the HDD 304 in which the human operation and the sound related thereto are stored in association with each other. By comparing and contrasting the voice c other than the person separated by the voice separating function 802 with the motion of the subject detected by the motion detecting function 1001 using 3048, the voice c other than the separated person c Then, a voice related to the motion of the subject is extracted. Then, the operation sound analysis function 1023 measures the loudness and sound quality of the extracted sound.

Then, the parameter setting function 804 sets a parameter of 2-4 quietness based on data obtained through such a series of processing.

<Body State> FIG. 18 is a functional block diagram of the “body state” process in the CPU 301.

Hereinafter, the processing relating to the “body state” according to the first embodiment will be described in accordance with the processing order with reference to FIG.

First, the amount of perspiration detected by the perspiration sensor 10 and the heart rate, systolic blood pressure and diastolic blood pressure data detected by the heart / blood pressure sensor 11 are stored in the physiological data storage area 3023 in the RAM 302.

Next, in addition to the physiological data stored in the physiological data storage area 3023, the physiological condition judging function 1301 of the CPU 301 can determine the subject's position obtained by the above-described motion detecting function 1001 and facial expression detecting function 1011. Using the physical condition judgment knowledge database 3049 in the HDD 304 in which the posture, the motion, and the expression are added and the various data are associated with the physical condition, the degree of the tension and the fatigue of the subject are comprehensively determined. To judge.

In particular, regarding the degree of fatigue, it is often difficult to know only from physiological data such as the amount of sweat, heart rate, blood pressure and the like. Therefore, in the first embodiment, it is necessary to determine how much the subject blinks. Detecting the degree of yawning and recording the number of yawns facilitates the detection of the degree of fatigue.

In the case of blinking, the eye portion observation function 1017
It is sufficient to simply detect whether the eyes are open or closed, but in the case of a yawn, usually, a human covers his or her mouth with his / her hand, so the motion detection function 1001 detects the movement of the arm, At the same time, it is necessary to recognize the shape of the mouth by the mouth part observation function 1018.

Further, at this time, the motion detection function 1001 detects the direction of the face, the movement of the neck, the movement of the limbs, etc., so that "hands on the hips", "slaps on the shoulders", "stretches" and "stretches". It is also possible to detect unconscious movements of the body such as "relax the stiffness of the neck." In particular, since these operations are characterized by performing the same operation repeatedly, by recording the size, number of times, etc. of each operation, the state of the degree of fatigue can be more easily grasped. .

Then, the parameter setting function 804 sets parameters of 3-1 degree of tension and 3-2 degree of fatigue.

[Regarding Comparison of Behavior between Subject and Model Person and Output of Advice] Next, the processing in steps S506 to S508 in the flowchart shown in FIG. 10 will be described in more detail.

In these steps, in short, the parameter values set for the subject wearing the wristwatch-type portable information terminal are compared with the parameter values already set as those corresponding to the ideal model person. Then, an advice is created based on the comparison result, and a process of outputting the advice is performed.

At this time, it is better to accumulate a large amount of data obtained from the speech of the subject, so that a more accurate comparison process can be performed. Therefore, after a series of speech such as a presentation has been completed, the user has returned home. It is desirable that the subject himself / herself slowly looks at the comparison result, such as later.

More specifically, when the subject presses the mode button 6 several times to put the wristwatch-type portable information terminal into the evaluation mode, the CPU 301 performs the processing, and the liquid crystal display 4 displays the comparison result and the comparison result. Is displayed based on the information for the subject to approach the ideal person.

FIG. 19 shows an example of the display of the comparison result and the advice at that time.

In FIG. 19, items such as “loudness of voice”, “velocity of voice”, “smoothness of words”, and “wording” are provided as notes of “how to speak”. The degree of achievement is displayed as a percentage based on the difference between the parameter related to the subject's behavior and the parameter related to the ideal person's behavior, and, at the same time, advice on how to improve the behavior in order for the subject to approach the ideal person is provided. Is displayed. Further, at the top of the display, a score as an overall evaluation is displayed.

Here, it is needless to say that by operating the arrow buttons 8, it is possible to see not only "how to speak" but also advice regarding "personal care" and "physical condition".

[0116] The advice displayed here is stored in advance in the advice-added knowledge database 3050 in the HDD 304. of course,
The advice itself may be rewritten so as not to make the subject feel uncomfortable.

Further, as shown in FIG. 19, in addition to the advice on how to make the subject approach the ideal person based on the parameter comparison result, the evaluation function 905 evaluates the subject. From errors such as misuse or misuse of honorifics, or mistakes between honorifics and humble words,
Incorrect and correct answer of subject's utterance expression, indication of undesired utterance content and desirable correct answer example, and comprehensive evaluation and correctness of user's utterance derived from score of error type of utterance expression and error distance of utterance content A hint or the like for conducting a conversation may be advised.

Of course, such advice is displayed not only on the liquid crystal display 4 but also on the speaker 3.
More specifically, it may be output as an electronic sound or a human voice, or a wristwatch-type portable information terminal may be provided with a vibrator function and transmitted to a subject with a signal of a specific rhythm.

Further, as shown in FIG.
When the best phrase and the worst phrase are stored in advance in the phrase database 3051 of the subject, and the subject puts the wristwatch-type portable information terminal into the evaluation mode in order to see the evaluation of his / her own behavior of the day, The utterance contents may be compared with the phrases in the phrase database 3051 to display “the best phrase of the day” and “the worst phrase of the day” on the liquid crystal display 4.

The “best phrase of the day” and the “worst worst phrase of the day” can be automatically set based on the evaluation result by the evaluation function 905 in addition to the subject storing the phrase in the phrase database 3051 in advance. It is also possible.

Further, as shown in FIG. 21, when a personal appearance or a facial expression farthest from the ideal person's personal appearance or facial expression occurs, it is updated and stored in the image storage area 3022 in the RAM 302. When the test subject puts the wristwatch-type portable information terminal in the evaluation mode in order to see the evaluation of his / her behavior for a day, the stored image is called from the image storage area 3022, and is displayed on the liquid crystal display 4.
As a “today's worst action”, a quick time movie image may be displayed.

As described above, according to the first embodiment, the subject himself / herself always considers the video camera and the microphone, and confirms all the recorded actions from the beginning to the end. Even without doing so, it is very easy to bring one's behavior closer to the ideal person who becomes the model.

(Second Embodiment) In the first embodiment, the wristwatch-type portable terminal device is provided with only one camera. However, by providing a plurality of cameras, other devices other than the subject himself / herself can be provided. Even a person can be photographed with a camera.

A case where a plurality of cameras are provided in a wristwatch-type portable information terminal as described above will be described as a second embodiment of the present invention.

In the second embodiment, each of the cameras is used to photograph the subject and the other person, and the movement and expression of the person other than the subject can be read. The determining function 1007 determines the motion of each person, and the facial expression determining function 1018 determines the facial expression of each person. In addition, based on the voice b of a person other than the subject himself or herself separated by the voice separation function 802, the voice analysis function 803 analyzes the content of each person's remark.

As a result, it is possible to determine the behavior of a person other than the subject. For example, whether the audience is yawning or blinking when the subject is giving a presentation, i.e., they are becoming sleepy, less focused, angry, Whether nodding, laughing, or convinced can be determined.

Here, the presentation of the subject is
If you include the degree of appeal to the audience as a new evaluation item for the ideal person, you can set parameters by predicting the audience response due to the behavior of the ideal person, and based on the actual audience response By comparing the parameter value with the parameter value, it is possible to indirectly determine how close the subject's behavior is to the ideal person's behavior.

As described above, according to the second embodiment, by analyzing the behavior of another person other than the subject himself, it is possible to bring his / her behavior closer to the ideal model person. It will be easier.

(Third Embodiment) In the first embodiment, the value of the parameter corresponding to the ideal person is set to a certain value regardless of the situation. However, different parameters are required depending on the situation, such as giving a presentation, working at a company desk, eating, or sleeping. That is, it is preferable that a lot of remarks be made during a presentation, but it is preferable that there be few remarks except for telephone conversations when working at a company desk.

In view of the above points, in the third embodiment of the present invention, it is possible to set a status mode such as “presentation mode”, “work mode”, “eat mode”, “sleep mode”, etc. A parameter specific to the situation mode is set in the evaluation item so that the above-described problem can be dealt with. The setting of the situation mode is performed by the subject himself / herself operating the wristwatch-type portable information terminal immediately before giving a presentation, immediately before starting his or her work at the office desk, just before eating, just before going to bed.

As described above, according to the third embodiment, the evaluation method is changed depending on the situation, so that the subject can be used as a model regardless of the situation. You can bring your behavior closer to the ideal person.

(Fourth Embodiment) In the first embodiment described above, the subject uses the mode button 6 of the wristwatch-type portable information terminal.
, Unless the information terminal is in the evaluation mode,
I cannot know the evaluation of my behavior. However, it is natural that the subject wants to know the evaluation of his / her behavior in real time.

In order to cope with such a situation, in the fourth embodiment, the wristwatch-type portable information terminal is set not in the watch mode but in the evaluation mode in advance while the subject is speaking. While the subject is making a speech, the comparison result is output one by one.Also, a warning can be given immediately for the speech of the subject who should be warned that greatly exceeds the allowable value from the set target. .

More specifically, since the parameters relating to the behavior of the subject are set by the CPU 301 while the subject is making the behavior, the value of the parameter is immediately compared with the value of the parameter of the ideal person. If the two parameters are different, the speaker 3
Therefore, a sequential advice is generated as a human voice, and an electronic sound is generated as a warning sound. Of course, if a vibrator function is provided in the wristwatch-type information terminal, advice or warning can be notified to the subject by vibration without generating sound from the speaker 3.

As described above, various warning methods are conceivable, and the subject can freely select the watch by pressing the mode button 6 several times to set the wristwatch-type information terminal in the warning setting mode. be able to.

FIG. 22 shows an example of the warning setting screen.

In FIG. 22, after the information terminal is set in the warning setting mode, first, “sound”, “voice”, “character”,
"Sound" is selected from "vibes", and "wake-up" and "rully" are also selected.

As described above, according to the fourth embodiment, the subject can confirm in real time that his / her behavior differs from that of an ideal person.

(Fifth Embodiment) In the first embodiment, advice and warning are given based on how much the subject's behavior differs from the ideal person's behavior. Regardless of who the ideal person is, there are always behaviors that are judged inappropriate. It is too violent, berserk, and bad habit.

[0140] More specifically, if the voice is too loud at mealtime, the handling of tableware is rough, the use of honorifics is completely different during conversation, the manner of walking in the street is rough, or the hair is frequently touched, For example, chewing a nail or eating too much while dieting.

If such inappropriate words and behaviors are stored in advance in the HDD 304 of the wristwatch-type portable information terminal, a warning can be given as soon as the subject makes the words and phrases, so It is possible to correct inappropriate behaviors without knowing and correct habits.

FIG. 23 shows an example of the warning at that time.

In the fifth embodiment, even when the wristwatch-type portable information terminal is in the normal clock mode, detection of inappropriate speech is performed sequentially, and when the subject makes inappropriate speech, Immediately, a warning message is displayed on the liquid crystal display 4 so as to hide the display of the clock.

Of course, the warning method at this time is as follows.
Various methods as shown in the fourth embodiment are possible.

As described above, according to the fifth embodiment, the subject can check his / her inappropriate behavior in real time.

(Sixth Embodiment) In the first embodiment, the wristwatch-type portable information terminal performs all the processing independently. However, the number of ideal persons is large, and the evaluation items are diversified. In the case where the time for recording the behavior of the subject is extended for a long time, only the RAM 302 or the HDD 304 in the wristwatch-type portable information terminal needs to store all the information. Insufficient capacity.

Of course, if a large-sized storage device is incorporated in a wristwatch-type portable information terminal, it is possible to cope with that, but in this case, there is a problem that the information terminal body becomes large and heavy.

Therefore, in the sixth embodiment, the information terminal only accumulates data, does not perform any information processing, and the external information processing apparatus performs parameter comparison processing and result notification processing. By doing so, the burden on the information terminal itself was reduced.

FIG. 24 is a system configuration diagram for explaining the sixth embodiment.

First, the subject prepares a data card 2402 for data storage and puts it on the wristwatch-type portable information terminal 2.
It is inserted into the data card slot 401 and the information terminal is carried around. Then, every time the subject makes a certain behavior, voice data, image data, physiological data, and the like relating to the behavior are accumulated one after another in the RAM of the information terminal.

When the subject returns home, the subject performs a process of setting an ideal person as a model toward the information processing device 2404 installed in the house. This may be done before leaving the house. The detailed contents are as described above. Then, when the setting of the ideal person is completed, the subject enters the wristwatch-type portable information terminal 240.
1 and pull out the data card 2402 from the card reader 2403 attached to the information processing device 2404.
And inputting an evaluation result by an input device such as a keyboard. Then, as a result of the parameter comparison processing as described above, the advice is output by text or voice. In this way, at the subject's house, behavior analysis and comprehensive evaluation can be performed on a daily basis.

At this time, the data card 24 for storing data is used.
02, not a wristwatch-type portable information terminal 24
01 to the information processing device 2404 by wireless communication using radio waves or infrared rays, or data may be sent by wired communication using a telephone line or the Internet.

As described above, according to the sixth embodiment, all the necessary evaluation processing is performed by the external information processing apparatus.
The weight can be reduced.

[0154]

As described above, according to the present invention,
It is extremely easy for the subject to express his or her own behavior to the model ideal person without always being aware of the video camera or microphone and checking the entire beginning and end of his recorded behavior. This makes it possible to bring the user closer to the user and check his / her inappropriate behavior in real time.

[Brief description of the drawings]

FIG. 1 is an external view of a front side of a wristwatch-type portable information terminal according to a first embodiment.

FIG. 2 is an external view of the back side of the wristwatch-type portable information terminal according to the first embodiment.

FIG. 3 is a hardware connection diagram inside the wristwatch portable information terminal according to the first embodiment.

FIG. 4 is a diagram showing a memory map inside a RAM 302;

FIG. 5 is a diagram showing a memory map inside the HDD 304.

FIG. 6 is a diagram illustrating a display example of a model setting mode.

FIG. 7 is a diagram showing a process of setting a model in a model setting mode.

FIG. 8 is a diagram illustrating an example of a model person selection confirmation screen.

FIG. 9 is a diagram showing a parameter setting screen of a model person.

FIG. 10 is a flowchart of an ideal behavior support process according to the first embodiment.

FIG. 11 is a functional block diagram of processing relating to “how to speak” in a CPU 301.

FIG. 12 is a functional block diagram of an interpersonal conversation ability determination function 805.

FIG. 13 is a functional block diagram of processing relating to “personal care” in the CPU 301.

FIG. 14 is a detailed functional block diagram of processing relating to “personal care”;

FIG. 15 is a diagram showing an example of a three-dimensional model stored in a whole body model database 3044.

FIG. 16 is a conceptual diagram related to detection of a motion of a subject.

FIG. 17 is a conceptual diagram related to detection of a subject's face, eyes, and mouth.

FIG. 18 is a functional block diagram of processing relating to “body state” in the CPU 301.

FIG. 19 is a diagram illustrating a display example of a comparison result and advice in the first embodiment.

FIG. 20 is a diagram showing an example of a result display for displaying a best phrase and a worst phrase.

FIG. 21 is a diagram showing a result display example of displaying a worst action quick time movie.

FIG. 22 is a diagram showing an example of a warning setting screen according to the fourth embodiment.

FIG. 23 is a diagram showing an example of a warning display according to the fifth embodiment.

FIG. 24 is a configuration diagram of an ideal behavior support system according to a sixth embodiment.

[Explanation of symbols]

 Reference Signs List 1 CCD camera 2 Microphone 3 Speaker 4 Liquid crystal display 5 Clock setting button 6 Mode button 7 Select button 8 Arrow button 9 Data card slot 10 Sweating sensor 11 Heart rate blood pressure sensor 301 CPU 302 RAM 303 ROM 304 HDD 305 Camera control unit 306 Microphone control unit 307 Speaker control unit 308 Display control unit 309 Button control unit 310 Sensor control unit 311 Data card control unit 312 Data card 313 Clock control unit 314 Clock 315 Bus

Claims (45)

[Claims] 1. A storage means for storing first information relating to the behavior of a first person, an input means for inputting second information relating to the behavior of a second person, and a first information stored in the storage means. And comparing means of the second person and the second information input by the input means, based on a comparison result by the comparing means, for approaching the second person to the first person. And a notifying means for notifying of advice regarding improvement of the behavior of the person. 2. The method according to claim 1, wherein the first information and the second information are:
The ideal behavior support device according to claim 1, further comprising at least one of information on voice, information on body movement, and information on physiological information. 3. The input means includes: voice input means for inputting voice of the second person; image input means for inputting an image of the second person; and physiological information of the second person. 2. The apparatus according to claim 1, further comprising at least one of physiological information input means for inputting.
The ideal behavior support device as described. 4. The ideal behavior support device according to claim 1, wherein the notification unit outputs the advice as a voice. 5. The ideal behavior support device according to claim 1, wherein the notification unit outputs the advice as a vibration. 6. The ideal behavior support device according to claim 1, wherein the notification unit displays the advice as characters. 7. The ideal behavior support device according to claim 1, further comprising an instruction unit for instructing the comparison unit to start comparison. 8. The first information and the second information,
The input means includes: voice input means for inputting voice of the second person; conversion means for converting voice input by the voice input means into character information; and 2. The ideal behavior support device according to claim 1, further comprising deriving means for deriving information on the wording of the second person based on the converted character information. 9. The first information and the second information,
The input means includes: voice input means for inputting voice of the second person; recognition means for recognizing a feature of voice input by the voice input means; 2. The ideal behavior support device according to claim 1, further comprising deriving means for deriving information on the emotion or character of the second person from the recognized voice feature. 10. The first information and the second information include information on a facial expression, wherein the input means includes: an image input means for inputting an image of the second person; and the image input means. Face extracting means for extracting a face portion of the second person from the image input by the following, based on the face portion extracted by the face extracting means,
The ideal behavior support device according to claim 1, further comprising deriving means for deriving information on the facial expression of the second person. 11. The first information and the second information include information related to emotion or physical condition, the input unit includes: a physiological information input unit that inputs physiological information of the second person; 2. The ideal behavior supporting device according to claim 1, further comprising: deriving means for deriving information on the emotion or physical condition of the second person based on the physiological information input by the physiological information inputting means. . 12. The first information and the second information include information relating to emotion or physical condition, the input unit includes: an image input unit that inputs an image of the second person; Physiological information input means for inputting physiological information of a person; an image input by the image input means; and physiological information input by the physiological information input means, based on the physiological information input by the physiological information input means. The ideal behavior support device according to claim 1, further comprising: a deriving unit that derives information on emotion or physical condition. 13. The ideal behavior support device according to claim 1, further comprising a change unit that changes the first information. 14. A storage means for storing first information relating to the behavior of a first person, and second information relating to the behavior of a second person and the behavior of surrounding persons with respect to the behavior of the second person. An input unit for inputting; a comparing unit for comparing the first information stored in the storage unit with the second information input by the input unit; A notifying means for notifying the second person of advice relating to improvement of the behavior for approaching the first person. 15. The first information and the second information include information relating to emotion, and the input unit includes: a voice for inputting voices of the second person and a person around the second person. Input means, separating means for separating the voice input by the voice input means into those of the second person, and those of people around the second person, Recognizing means for recognizing voice features separated from those of the persons around the second person; and deriving information about emotions of the people around the second person from the voice features recognized by the recognizing means. The ideal behavior supporting device according to claim 14, further comprising a deriving unit that performs the deriving operation. 16. The first information and the second information include information on emotion, and the input unit inputs an image of the second person and an image of a person around the second person. Image input means; and an image input by the image input means;
Discriminating means for discriminating the facial expression and body movement of a person around the person, based on the facial expression and body movement of the person around the second person determined by the discriminating means, 15. The ideal action support device according to claim 14, further comprising deriving means for deriving information about emotions of persons around the second person. 17. A storage unit for storing first information relating to a behavior to be warned, an input unit for inputting second information relating to a behavior of a person, a first information stored in the storage unit, Ideal action support, comprising: comparison means for comparing the second information input by the input means; and warning means for warning the person based on the comparison result by the comparison means. apparatus. 18. The ideal behavior support device according to claim 17, wherein the warning unit issues a warning as a voice. 19. The ideal behavior support device according to claim 17, wherein the warning unit issues a warning as vibration. 20. The ideal behavior support device according to claim 17, wherein said warning means issues a warning as a character. 21. An input step of inputting first information relating to a first person's behavior, a second information relating to a second person's behavior stored in a storage means, and a second information input in the input step. A comparing step of comparing the first information with the first information; and a notifying step of notifying the first person of advice regarding improvement in behavior to approach the second person based on the comparison result in the comparing step. An ideal behavior support method characterized by comprising: 22. The information according to claim 21, wherein the first information and the second information include at least one of information on voice, information on body movement, and information on physiological information. 21. The ideal behavior support method according to 21. 23. The input step, comprising: a voice input step of inputting voice of the first person; an image input step of inputting an image of the first person; and inputting physiological information of the first person. 22. The ideal behavior supporting method according to claim 21, comprising at least one of a physiological information inputting step of inputting. 24. The ideal action support method according to claim 21, wherein the advice is output as a voice in the notification step. 25. The ideal action support method according to claim 21, wherein the advice is output as a vibration in the notification step. 26. The ideal action support method according to claim 21, wherein the advice is displayed as characters in the notification step. 27. The ideal action support method according to claim 21, further comprising an instruction step of instructing a start of the comparison in the comparison step. 28. The method according to claim 28, wherein the first information and the second information include information on use of words, wherein the input step includes: a voice input step of inputting a voice of the first person; A conversion step of converting the input voice into character information; and a derivation step of deriving information on the wording of the first person based on the character information converted in the conversion step. The ideal behavior support method according to claim 21. 29. The first information and the second information include information relating to emotion or personality, wherein the input step includes: a voice input step of inputting a voice of the first person; and the voice input step. And a derivation step of deriving information on the emotion or character of the first person from the features of the voice recognized in the recognition step. The ideal behavior support method according to claim 21. 30. The first information and the second information include information related to a facial expression, wherein the input step includes an image input step of inputting an image of the first person; and the image input step. A face extraction step of extracting a face portion of the first person from the image input in step (a), and information on a facial expression of the first person is derived based on the face portion extracted in the face extraction step. The ideal behavior supporting method according to claim 21, further comprising a deriving step. 31. The first information and the second information include information related to emotion or physical condition, wherein the input step includes a physiological information input step of inputting physiological information of the first person; A deriving step of deriving information relating to the emotion or physical condition of the first person based on the physiological information input in the physiological information inputting step.
1. The ideal behavior support method according to 1. 32. The first information and the second information include information on emotion or physical condition, wherein the inputting step includes an image inputting step of inputting an image of the first person; A physiological information inputting step of inputting physiological information of the person, an image input in the image inputting step, and the physiological information input in the physiological information inputting step, based on the physiological information input in the physiological information inputting step. 22. The ideal behavior supporting method according to claim 21, further comprising a deriving step of deriving information on emotion or physical condition. 33. The ideal action support method according to claim 21, further comprising a change step of changing the second information. 34. An inputting step of inputting first information relating to the behavior of a first person and the behavior of surrounding persons with respect to the behavior of the first person, and the second person's behavior stored in a storage means. The second about behavior
And a comparison step of comparing the second information input in the input step with the second information. In the comparison step, the first person approaches the second person based on the comparison result in the comparison step. A notification step of notifying advice on improvement of behavior for the user. 35. The first information and the second information include information relating to emotion, and the inputting step includes: a voice for inputting voices of the first person and a person around the first person. An input step, a separation step of separating the voice input in the voice input step into those of the first person and those of persons around the first person, and A recognition step of recognizing a feature of a voice separated from that of a person around the one person; and deriving information about emotions of the person around the first person from the features of the voice recognized in the recognition step. 35. The ideal behavior supporting method according to claim 34, further comprising: 36. The first information and the second information include information on emotion, and the inputting step includes inputting an image of the first person and an image of a person around the first person. An image input step; a discrimination step of discriminating a facial expression and a body movement of a person around the first person from the image input in the image input step; and the first discrimination discriminated in the discrimination step. 35. A deriving step of deriving information on emotions of a person around the first person based on facial expressions and body movements of the person around the person. Action support method. 37. An input step of inputting first information relating to a person's behavior, a second information relating to a warning behavior stored in a storage unit, and the first information input in the input step. An ideal action support method, comprising: a comparing step of comparing; and a warning step of giving a warning to the person based on a result of the comparison in the comparing step. 38. The ideal behavior support method according to claim 37, wherein in the warning step, a warning is issued as a voice. 39. The ideal action support method according to claim 37, wherein in the warning step, a warning is issued as vibration. 40. The ideal action support method according to claim 37, wherein in the warning step, a warning is issued as a character. 41. An ideal behavior support system comprising a portable terminal device and an information processing device, wherein the portable terminal device comprises: input means for inputting first information relating to a certain person's behavior; and the input means A transmitting unit that transmits the first information input by the information processing device to the information processing device, wherein the information processing device includes: a storage unit configured to store information on various persons and information on behaviors of the various people; Selecting means for selecting, from the storage means, a specific person and second information relating to the behavior of the specific person; receiving means for receiving the first information transmitted from the portable terminal device; Comparing means for comparing the first information received by the means with the second information selected by the selecting means, based on a comparison result by the comparing means, An ideal behavior support system, comprising: a notification unit configured to notify a user of an advice regarding improvement of behavior to approach a certain person. 42. An ideal behavior support system comprising a portable terminal device and an information processing device, wherein the portable terminal device has first insertion means for inserting a removable external storage means. Input means for inputting first information relating to a person's behavior, and storage means for storing information input by the input means in an external storage means inserted into the first insertion means; The apparatus includes: a second insertion unit for inserting a detachable external storage unit; and first information relating to a certain person's behavior stored in the external storage unit inserted into the second insertion unit. Reading means for reading, various people, storage means for storing information on the behavior of the various people, and a specific person and second information on the behavior of the specific person are selected from the storage means. Selecting means; comparing means for comparing the first information read by the reading means with the second information selected by the selecting means; and the person based on a comparison result by the comparing means. Notification means for notifying the user of advice regarding improvement of behavior for approaching the specific person. 43. An input module for inputting first information relating to a first person's behavior, a second information relating to a second person's behavior stored in storage means, and an input by the input module. A comparison module for comparing the input first information with the first information, and an improvement in the behavior of the first person toward the second person with respect to the first person based on a comparison result obtained by the comparison by the comparison module. A storage medium storing a computer program having a notifying module for notifying advice on the information. 44. A storage unit for storing first information relating to an ideal behavior, an input unit for inputting second information relating to a certain person's behavior, the first information stored in the storage unit, Comparing means for comparing the second information input by the input means; and, based on a result of the comparison by the comparing means, notifying the person of advice on improvement of the behavior to approach the ideal behavior. An ideal action support device comprising: 45. An input step of inputting first information relating to a certain person's behavior, a second information relating to an ideal behavior stored in a storage unit, and a first information input in the input step. A comparing step of comparing, based on a comparison result in the comparing step, a notification step of notifying the person of advice regarding improvement of behavior to approach the ideal behavior. Ideal behavior support method to do.