JP2002183859A - Abnormality discriminating and reporting apparatus - Google Patents

Abstract

PROBLEM TO BE SOLVED: To report the occurrence of abnormal situation to the outside without wrong recognition even if a user can't directly report the occurrence of abnormal situation. SOLUTION: A discriminating part 7 discriminates the occurrence of abnormal situation according to both the voice information of a voice recognizing part 2 and the measurement result of a pulse sensor 3 and a sweating amount sensor 4, and according to the output of an abnormality occurrence discriminating signal from the discriminating part 7, the occurrence of abnormal situation is reported to the outside from a communicating part 11, whereby the occurrence of abnormal situation can be reported to the outside without direct operation of the user. Since judgment is made according to the user's abnormality of pulse rate and sweating amount, the accuracy of discriminating the occurrence of abnormal situation can be improved to reduce wrong recognition. A release button 10 is pressed to change setting of a detection threshold, so that the accuracy of discriminating the occurrence of abnormal situation can be improved.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an abnormality discriminating and notifying apparatus used in a security system for reporting an abnormality to an outside when an abnormal situation occurs.

[0002]

2. Description of the Related Art Generally, a push-button alarm device, a portable security bell, a surveillance camera, and the like are used to monitor and report occurrence of an abnormal situation such as a crime. Japanese Patent Application Laid-Open No. H11-295141 proposes a notification device that recognizes a voice message when an abnormal situation occurs and reports the occurrence of the abnormality to the outside.

[0003]

In a push-button alarm device, when an abnormal situation such as an abnormal person intruding into a house for the purpose of a crime or the like occurs, the resident or the like of the house can use the alarm device. There is a problem that it is necessary to move to the installed place and press the alarm button, which involves danger. In addition, even if the user has a portable security bell, there is a possibility that the security bell cannot be operated if the limbs, mouth, etc. are restrained or a weapon is directed.

Further, even if a surveillance camera is installed in a house, the surveillance camera can only monitor in a specific direction to which the surveillance camera is directed, and the occurrence of an abnormal situation may not be detected in the blind spot of the surveillance camera. In addition, there is also a problem that it is not easy to constantly monitor the image by the surveillance camera. Furthermore, in a device that uses voice recognition to notify an abnormal situation to the outside, there is a possibility that a message during a normal conversation may be erroneously recognized as having occurred, and it is necessary to provide sufficient voice recognition accuracy. is there.

The present invention has been made in view of the above circumstances, and even if a user falls into a situation where the occurrence of an abnormal situation cannot be reported directly, the user is notified of the occurrence of the abnormal situation without erroneous recognition. An object of the present invention is to provide an abnormality determination notification device that can perform the determination.

[0006]

In order to solve the above-mentioned problems, an abnormality discriminating and notifying apparatus according to the present invention measures a voice recognition means for recognizing specific voice information and measures a physiological response corresponding to a mental state of a user. Determining that an abnormal state has occurred based on a recognition result of voice information by the voice recognition unit and measurement data of the measurement sensor, and outputting an abnormality occurrence determination signal when it is determined that an abnormality has occurred. Means,
Notification means for notifying a user of the occurrence of an abnormality based on the abnormality occurrence determination signal, and a user notified of the occurrence of the abnormality by the notification means, release means for releasing the output of the abnormality occurrence determination signal, Communication means for communicating the occurrence of an abnormality to the outside based on the abnormality occurrence determination signal.

[0007] The abnormality determination reporting device of the present invention further comprises a learning means for changing the detection threshold of the measurement data for determining the occurrence of an abnormality based on the release of the output of the abnormality occurrence determination signal by the release means. It is preferred to have.

[0008] It is preferable that the determination means starts to determine whether there is an abnormality in the measurement data of the measurement sensor when the voice information is recognized by the voice recognition means. Furthermore, when the measurement sensor continues for a predetermined time or more and detects an abnormal state, it may be determined that an abnormality has occurred.

It is preferable that the voice recognition means recognizes voice information using an HMM (Hidden Markov Model).

It is preferable that the measurement sensor is one or both of a pulse sensor for detecting a pulse value of the user and a sweat sensor for detecting a sweat amount of the user. Further, the upper limit and the lower limit of the detection threshold of the measurement data by the measurement sensor may be arbitrarily set from the upper limit and the lower limit of a predetermined range.

[0011] Preferably, the communication means is provided with a communication means for making a call between a user and a specific person.

[0012] It is preferable that the communication means is provided with voice guidance means for performing voice guidance of occurrence of an abnormality.

[0013]

(Embodiment 1) Hereinafter, an abnormality determination notification device according to Embodiment 1 of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a schematic configuration of an abnormality discriminating and notifying device 1 according to the first embodiment, and FIG. 2 is a schematic diagram showing a use state of the device.

The abnormality discriminating and notifying device 1 is formed in a small size so that it can be carried and used. For example, as shown in FIG. 2, the device is attached so as to be in direct contact with the skin of the user's wrist or the like. Used as Abnormality determination reporting device 1
As shown in FIG. 1, a voice recognition unit 2 for recognizing voices such as specific words and sound sounds registered in advance, and a pulse sensor 3 for detecting a pulse rate of a user when worn on an arm or the like. Similarly, when worn on an arm or the like, a perspiration sensor 4 for detecting the amount of perspiration of the user, a position sensor 5 for specifying the current position of the user on the ground, An emergency button 6 for notifying an emergency situation to the outside by performing a pressing operation.

The outputs of the voice recognition unit 2, the pulse sensor 3 and the perspiration sensor 4 are given to a determination unit 7 for determining whether or not an abnormal situation has occurred.
Based on the output of the pulse sensor 3 and the output of the perspiration sensor 4, the presence or absence of an abnormal situation is determined. The output of the determination unit 7 is provided to the notification unit 8 and the communication unit 11. The notifying unit 8 includes, for example, a vibrator or the like so as to perform a predetermined operation to notify the user that an abnormality has occurred, and the determining unit 7 determines that an abnormal situation has occurred. Then, by vibrating the arm or the like of the user wearing the abnormality determination notification device 1, the user is notified that the determination unit 7 has determined that the abnormality has occurred. The output of the notification unit 8 is given to a release button 10. When the release button 10 is depressed, the output of the abnormality occurrence signal by the determination unit 7 is stopped, and when the release operation is not performed, it is determined that an abnormal situation has occurred, and the output signal of the determination unit 7 is transmitted to the communication unit 11. Continues to be output to The communication unit 11 communicates the fact that an abnormal situation has occurred to an acquaintance, police, etc., based on the output signal output from the determination unit 7.

The output of the emergency button 6 and the output of the position sensor 5 are given to the communication unit 11, respectively. The emergency button 6 allows the user to operate the emergency button 6 regardless of whether the determination unit 7 determines that an abnormal situation has occurred. By pressing, the occurrence of an abnormal situation is directly communicated to the outside. If the current position of the user is not specified by the position sensor 5, the communication unit 11
Communicates the occurrence of the abnormal situation to a specific person outside the police, etc., assuming that the abnormal situation has occurred. The pressing of the release button 10 is also detected by the learning unit 9. The learning unit 9 changes the threshold value of the determination unit 7 that determines that an abnormal situation has occurred by the pulse sensor 3 and the perspiration sensor 4 by pressing the release button 10.

Next, details of each component of the abnormality determination notification device 1 will be described with reference to the drawings.

FIG. 3 is a block diagram showing a schematic configuration of the speech recognition unit 2. As shown in FIG.

The voice recognition unit 2 includes a microphone 21 for receiving a voice from the outside, a voice analysis unit 22 for performing a voice analysis of the voice signal received by the microphone 21, and a voice model 25 for determining the phonemic sound of the voice signal. , And a matching unit 24 for matching specific words and sounds based on a dictionary 26 registered in advance.

The speech recognition unit 2 uses an HMM (Hidden Markov Model). The speech analysis unit 22 analyzes the speech signal received by the microphone 21 and converts it into spectrum parameters. Acoustic model 25
Is calculated by the likelihood calculation unit 23 based on the word and the specific words and sound sounds registered in advance in the dictionary 26 are compared with each other to obtain the closest approximation from the words and sound sounds registered in the dictionary 26. Those that are recognized as being present are output to the determination unit 7.

Specific words registered in the dictionary 26 include:
For example, there are "kill" and "quiet",
Specific noises include, for example, gunshots, impact sounds of breaking glass, and the like.

The specific words and sounds registered in the dictionary 26 are given serial numbers in advance, and the voice recognition unit 2 adds a signal indicating that the voice information has been recognized and the serial numbers. Is output to the determination unit 7.

FIG. 4 is a block diagram showing a schematic configuration of the pulse sensor 3. As shown in FIG.

The pulse sensor 3 has a pressure sensor 31 provided on an air band (not shown) which is inflated to be brought into close contact with the user's wrist when air is introduced. The pressure sensor 31 detects a pressure based on the pulse of the closely attached wrist as a pulse signal. Pressure sensor 31
The noise included in the pulse signal detected by the filter 32 is removed by the filter 32. The pulse signal from which noise has been removed is amplified by the amplifier 33. The amplified pulse signal is provided to a waveform shaper 34, and is converted into a pulse signal by waveform shaping. In the waveform shaper 34,
The pulse signal converted into a pulse signal is counted up by a counter 35 and provided to a calculation unit 37. The arithmetic unit 37 counts up the pulse signal output within a predetermined time (for example, one minute) based on the pulse signal counted up by the counter 35 and the time signal output from the timer 36.

As described above, the calculation unit 3 of the pulse sensor 3
At 7, when the number of pulse signals output within a predetermined time is counted up, the count-up value of the pulse signal is
The pulse rate is output to the determination unit 7 as the pulse rate of the user.

FIG. 5 is a block diagram showing a schematic configuration of the perspiration amount sensor 4.

The perspiration sensor 4 has a pair of electrodes A and B, which are in contact with the skin surface of the user at appropriate intervals. A weak current is supplied to one electrode A by a power supply 41. The current of the power supply 41 and the current signal detected by the electrode B are given to the amplifier 42. A weak current flows between the electrodes A and B due to a decrease in the resistance value of the user's skin surface due to perspiration, and when the current is detected by the electrode B, the current is amplified by the amplifier 42 and is passed through the resistor 43. The appropriate voltage signal is output to the discriminating unit 7 as a perspiration amount signal indicating the perspiration amount of the user.
Is output to

The position sensor 5 indicates where the user is currently located, for example, by GPS or PH.
It specifies using the electric wave used for S.

FIG. 6 is a block diagram showing a schematic configuration of the discrimination section 7. As shown in FIG.

The discriminating unit 7 compares a pulse signal output from the pulse sensor 3 with a predetermined threshold value.
And a perspiration rate comparator 72 that compares the perspiration rate signal output from the perspiration sensor 4 with a predetermined threshold value. The outputs of the pulse rate comparator 71 and the perspiration rate comparator 72 are calculated by a calculation unit. 73. The operation unit 73 includes a clock 74
And the output of the voice recognition unit 2 are input. The calculation unit 73 uses the output signal from the voice recognition unit 2 as a trigger signal for starting calculation, and when the output signal from the voice recognition unit 2 is detected, the calculation unit 73
Based on the timer signal input from the clock 74,
Start counting up the timer. The counting is continued until the pulse rate comparator 71 and the sweat amount comparator 72 detect that no abnormal situation has occurred, and the occurrence of the abnormal situation is determined based on the counted up time.

FIG. 7 shows an example of a method for determining the presence or absence of a pulse abnormality by the pulse rate comparator 71. For example, when the detected pulse rate is equal to or more than a predetermined upper threshold K1 as shown in the area A of FIG. 7 or when it is less than the predetermined lower threshold K2 as shown in the area B in FIG. Outputs a signal “1” indicating that the user's pulse rate is abnormal and an abnormal state has occurred. On the other hand, when the pulse rate is between the lower threshold K2 and the upper threshold K1, a signal "0" indicating that the pulse rate of the user is normal and no abnormal situation has occurred is output. Output. It should be noted that K1 indicating the upper threshold of the pulse rate and K2 indicating the lower threshold can be set within predetermined ranges indicated by C and D, respectively, as shown in FIG. Are changed by the learning unit 9, respectively. The range in which the pulse rate is normal is determined as a pulse value obtained by adding a constant value to a normal average pulse value as an upper limit value, and a pulse value obtained by subtracting a constant value from the normal pulse value as a lower limit value. You may do so.

FIG. 8 is a graph showing an example of a method for determining the presence or absence of abnormal sweating.

As the user continues to be nervous, the amount of sweating of the user increases, and a signal "1" is output. If the signal is less than the threshold L, it is determined that the sweating is normal, A signal "0" indicating that the state is not abnormal is output.

The arithmetic unit 73 performs a sum operation of the signals output from the pulse rate comparator 71 and the sweat amount comparator 72, and counts up the timer signal when either signal is "1". When both signals output “0”, the counting up of the timer signal is stopped. Then, when the counted-up time exceeds a preset fixed time ΔT, the occurrence of an abnormal situation is determined, and an abnormal situation occurrence signal is output to the notification unit 8.

In FIG. 9, the horizontal axis represents time t (sec), and the vertical axis represents presence or absence of voice recognition (a), pulse rate (b),
A time chart illustrating generation of a voice recognition signal, temporal variation of a pulse rate and a perspiration amount, and a discrimination operation based on these values as a perspiration amount (c) and presence / absence of an abnormality occurrence signal (d) is shown. ing.

As shown in FIG. 9 (a), at time t1, the speech recognition unit 2 recognizes a specific word or sound "present", and as shown in FIGS. 9 (b) and 9 (c). If any of the pulse rate and the amount of perspiration exceeds the respective detection thresholds and this state continues for a predetermined fixed time ΔT, as shown in FIG. 4D, at the time t2 when the time ΔT has elapsed from the time t1 Then, the occurrence of an abnormal situation is determined by the arithmetic unit 73, and an abnormal situation occurrence signal is output.

It should be noted that a specific word or sound is recognized by the voice recognition unit 2 at times other than the time of t1, but since no abnormality is detected in the pulse rate and the amount of sweat at that time, the calculation is performed. The abnormal situation occurrence signal by the unit 73 is not output.

On the other hand, even if the pulse rate and the amount of perspiration exceed the detection threshold, if a specific word or sound is not detected by the voice recognition unit 2 at that time, the arithmetic unit 73 sends a trigger signal from the voice recognition unit 2 to the arithmetic unit 73. Is not output, and also in this case, the abnormal situation occurrence signal from the arithmetic unit 73 is not output.

FIG. 10A is a block diagram showing the configuration of the learning section.

The learning section 9 includes a ROM 91 for storing respective detection thresholds of the user's pulse rate and sweat rate, which are used as criteria for the determination section 7 to determine whether or not an abnormal situation has occurred.
And a count 92 for incrementing or decrementing the detection threshold value set in. FIG.
(B) shows a pulse rate upper limit table as an example of the detection threshold value stored in the ROM 91.

When the notification unit 8 notifies the user that the determination unit 7 has determined that an abnormal situation has occurred, the user is in a calm state and the determination of the occurrence of the abnormal situation by the determination unit 7 is incorrect. In this case, the user presses the release button 10 to reset the output of the abnormal condition determination signal by the determination unit 7, but at the same time as the reset of the determination signal, for example, the upper limit value of the pulse rate detection threshold value Is determined to be inappropriate, the counter 92 increments the address 1 of the pulse rate upper limit table, and sets the pulse rate stored at the address 2 as the upper limit value of the pulse rate detection threshold.
Thereby, the pulse rate upper limit is updated.

When the release button 10 is not pressed by the user, the upper limit of the pulse rate detection threshold is determined to be appropriate and the pulse rate stored in the address 1 of the pulse rate upper limit table by the counter 92 is determined. Keep the number as the upper limit.

Further, the address stored in the pulse rate lower limit table is decremented by the same procedure, thereby updating the pulse rate lower limit. Similarly, the detection thresholds for the upper limit and the lower limit of the amount of perspiration are also updated by incrementing or decrementing the address of the table stored in the ROM 91, respectively.

If the learning section 9 is provided with a reset button for resetting the counter 92, the counter 92 can be returned to the initial setting state, and by repeating the above-described learning operation, the learning section 9 is suitable for a new user. The threshold can be set.

Next, the operation of the abnormality determination notification device 1 according to the first embodiment having the above-described configuration will be described with reference to the flowchart shown in FIG.

First, in step S1, the emergency button 6
Judge the presence or absence of pressing, and if the pressing operation is performed,
A notification signal is output to the communication unit 11, and if the pressing operation has not been performed, the voice recognition unit 2 determines in step S <b> 2
The presence or absence of voice recognition of specific words, sounds, etc. is determined. In step S2, if the voice of the specific word or the like is not recognized, the process returns to step S1. to decide. If the pulse is abnormal in step S3, a predetermined time ΔT
If the pulse abnormality continues over the period, the notification processing signal is output to the communication unit 11. On the other hand, if the pulse abnormality has not continued for the predetermined time ΔT, the process returns to step S3.

If the pulse is not abnormal in step S3, it is determined in step S4 whether the user has abnormal sweating. If it is determined in step S4 that the perspiration is abnormal, a notification signal is output to the communication unit 11 in step S5 if the perspiration is continued for a predetermined time ΔT. On the other hand, if the sweating abnormality does not continue for the predetermined time ΔT, the process returns to step S3.

In step S4, if there is no abnormal sweating, it is determined that there is no abnormal occurrence, and the process returns to step S1.

As described above, the abnormality discriminating and notifying apparatus 1 according to the first embodiment includes the speech information recognized by the speech recognizing unit 2 and the pulse rate sensor 3 and the sweat rate sensor 4 of the user's pulse rate or sweat rate. Based on the measurement result, the determination unit 7 determines the occurrence of an abnormal condition, and the communication unit 11 notifies the outside of the occurrence of the abnormal condition. Therefore, the occurrence of the abnormality is determined without the direct operation of the user. It becomes possible to report. Moreover, the occurrence of an abnormal situation is not determined based only on the audio information.
Since the judgment is made based on the fluctuation of the physiological response such as the user's pulse rate and the amount of perspiration, it is possible to improve the accuracy of discriminating the occurrence of an abnormal situation, and to greatly increase the probability that an abnormal occurrence is erroneously recognized. Can be reduced. Further, when the discriminating unit 7 makes an erroneous recognition, the output signal for notifying the occurrence of an abnormal situation output from the discriminating unit 7 is stopped by the release button 10 and the detection threshold value such as the pulse rate set in the learning unit 9 is set. Is changed, the accuracy of determining the occurrence of an abnormal situation can be improved.

In a situation similar to the occurrence of a crime, for example, when the user watches a TV suspense program, the voice recognition unit 2 uses the voice of the TV to register a specific word or Even if the user recognizes noise and finds an abnormality in the user's pulse rate and sweating amount, the duration of the user's nervous state in a situation where an abnormal situation actually occurs is determined by TV.
Since the time during which the tension state continues under a situation similar to an abnormal situation such as a program is clearly different, the probability that the determination unit 7 erroneously recognizes the occurrence of the abnormal situation is further reduced.

(Embodiment 2) Hereinafter, an abnormality determination notification device 1 according to Embodiment 2 of the present invention will be described with reference to the drawings.

As shown in FIG. 12, the abnormality discriminating and notifying device 1 has a function of ascertaining the current state of the user from a remote place.
A communication unit 12 having a microphone for grasping the voice of the user and a speaker for transmitting a voice from outside to the user is connected to the communication unit 11. Other configurations are the same as those of the abnormality determination notification device 1 shown in FIG.

The abnormality determination notification device 1 according to the second embodiment is
When the user is in a situation where the user can talk with an outside person by providing the communication unit 12, the outside person can directly talk with the user and accurately monitor the situation of the user remotely. Can be grasped. Further, when the user cannot speak, the microphone provided in the communication unit 12 allows the user to recognize the situation by voice.

(Embodiment 3) Hereinafter, an abnormality determination reporting device according to Embodiment 3 of the present invention will be described with reference to the drawings.

As shown in FIG. 13, the abnormality discrimination notification device has the same communication unit 12 as the abnormality discrimination device of the second embodiment.
Is connected to the communication unit 11, and furthermore, when an abnormal situation is reported to the outside by the communication unit 11, a voice guidance unit for transmitting a message such as “Are you all right?” To the user from the outside. 13 is connected to the communication unit 12. The response button 14 for the user to respond to the message issued by the voice guidance unit 13 is
3 is connected.

In the abnormality discriminating and notifying apparatus 1 according to the third embodiment, when the occurrence of an abnormality is reported from the communication section 11, a message such as "Is it okay?" Is generated. When the user presses the response button 14 in response to this message, the transmission of the abnormality occurrence signal is stopped, and when the user does not press the response button 14, the user is faint, etc. It is determined that the abnormal situation continues, and the report is continued to the outside such as the police.

Each of the abnormality discriminating and notifying apparatuses 1 of the first to third embodiments described above can construct a network in each member by wearing each of members registered in advance such as family members and acquaintances. Then, the occurrence of the abnormality and its contents and location are notified to all the members constituting the member.

[0059]

As described above, the discriminating means determines the occurrence of an abnormal situation based on both the recognition of the voice information by the voice recognizing means and the measurement result by the measuring sensor, and the abnormal situation is externally detected by the communication means. Since the occurrence is reported to the outside, the occurrence of an abnormal situation can be reported to the outside without the direct operation of the user. In addition, since the occurrence of an abnormal situation is not determined based only on the voice information but based on the fluctuation state of the physiological response such as the pulse rate and the amount of sweating of the user, the accuracy of determining the occurrence of the abnormal situation is improved. Erroneous recognition can be reduced. Further, when the discriminating means erroneously recognizes, by the releasing means,
While stopping the output signal to notify the occurrence of an abnormal situation,
Since the learning means can change the setting of the detection threshold such as the pulse rate, the accuracy of determining the occurrence of an abnormal situation can be improved.

In addition, if the outside means can directly talk to the user by the communication means and the user cannot speak, the microphone is used to notify the user of the situation. Can understand.

Further, when an abnormality occurs in the voice guidance means,
When a message such as "Are you OK?" Is displayed to the user, and the response button is not pressed in response to this message, an abnormal situation occurs such as the user being unconscious. Is reported to the outside, such as the police, and an abnormal situation can be detected early.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of an abnormality determination notification device according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram illustrating an example of a usage state of the abnormality determination notification device according to the first embodiment of the present invention.

FIG. 3 is a block diagram illustrating a schematic configuration of a speech recognition unit.

FIG. 4 is a block diagram showing a schematic configuration of a pulse rate sensor.

FIG. 5 is a block diagram showing a schematic configuration of a sweat amount sensor.

FIG. 6 is a block diagram illustrating a schematic configuration of a determination unit.

FIG. 7 is a schematic diagram showing an algorithm for determining whether or not a pulse is abnormal by a determination unit.

FIG. 8 is a schematic diagram showing an algorithm in which a determination unit determines whether there is an abnormality in the amount of perspiration.

FIG. 9 is a graph illustrating the determination of the occurrence of an abnormality in the determination unit based on the presence or absence of voice recognition and a change in the pulse or the amount of perspiration.

FIG. 10A is a block diagram illustrating a schematic configuration of a learning unit, and FIG. 10B is an example of a table stored in a ROM.

FIG. 11 is a flowchart illustrating an operation of the abnormality determination notification device according to the first embodiment.

FIG. 12 is a block diagram illustrating a schematic configuration of an abnormality determination notification device according to a second embodiment.

FIG. 13 is a block diagram illustrating a schematic configuration of an abnormality determination notification device according to a third embodiment.

[Description of Signs] 1 Abnormality discrimination reporting device 2 Voice recognition unit 3 Pulse sensor 4 Sweating sensor 5 Position sensor 6 Emergency button 7 Discrimination unit 8 Notification unit 9 Learning unit 10 Release button 11 Communication unit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G10L 15/14 G10L 3/00 535Z 15/00 551S 15/28 F-term (Reference) 5C086 AA22 AA26 CA04 CA09 CA15 CB07 CB20 CB26 CB27 DA02 DA08 EA11 EA40 EA41 5C087 AA02 AA03 AA37 BB64 BB65 BB74 DD05 DD24 EE05 EE06 EE08 EE18 FF04 FF16 GG08 GG19 GG31 GG35 GG57 GG66 5D015 GG01 H0223 KK01

Claims (9)

[Claims] 1. A voice recognition means for recognizing specific voice information; a measurement sensor for measuring a physiological response corresponding to a mental state of a user; a recognition result of the voice information by the voice recognition means and the measurement sensor Discriminating means for discriminating the occurrence of an abnormal state based on the measurement data, and outputting an abnormal occurrence discrimination signal when it is judged that an abnormal occurrence has occurred; Means, a user notified of the occurrence of an abnormality by the notification means,
An abnormality determination notification device comprising: a release unit configured to release an output of an abnormality occurrence determination signal; and a communication unit that externally communicates occurrence of an abnormality based on the abnormality occurrence determination signal. 2. The learning device according to claim 1, further comprising a learning unit configured to change a detection threshold of the measurement data for determining whether an abnormality has occurred, based on cancellation of the output of the abnormality occurrence determination signal by the cancellation unit. Abnormality notification device. 3. The abnormality determination notification device according to claim 1, wherein the determination unit starts determining whether there is an abnormality in the measurement data of the measurement sensor when the voice information is recognized by the voice recognition unit. . 4. The abnormality determination notification device according to claim 3, wherein when the measurement sensor continues for a predetermined time or more and detects an abnormal state, it is determined that an abnormality has occurred. 5. The speech recognition unit according to claim 1, wherein the speech information is recognized using an HMM (Hidden Markov Model).
The abnormality discrimination notification device according to the above. 6. The abnormality determination and notification device according to claim 1, wherein the measurement sensor is one or both of a pulse sensor for detecting a pulse value of the user and a sweat sensor for detecting a sweat amount of the user. . 7. The abnormality determination notification device according to claim 6, wherein an upper limit and a lower limit of a detection threshold value of the measurement data by the measurement sensor are arbitrarily set from an upper limit value and a lower limit value of a predetermined range. 8. The abnormality determination notification device according to claim 1, wherein said communication means includes a communication means for making a call between a user and a specific person. 9. The abnormality determination notification device according to claim 1, wherein the communication unit includes a voice guidance unit that performs voice guidance of occurrence of an abnormality.