JP2018005308A - Wandering control device, wandering control method and wandering control program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a wandering control device that can safely control wandering even for a wanderer who does not follow voice notification.SOLUTION: A wandering control device 12 outputs speech to a wanderer detected, by means of at least one of audio or video when the wanderer is detected in a predetermined region inside a building, determines whether or not the wanderer forgets initial intent to have started wandering based on lapse time from a start of dialogue with the wanderer, and outputs the speech to induce the wanderer to a predetermined induction place by means of the at least one of the audio or the video, when the wanderer is determined to have forgotten the initial intent.SELECTED DRAWING: Figure 1

Description

  The present disclosure relates to a wrinkle suppression device, a wrinkle suppression method, and a wrinkle suppression program for suppressing wrinkles by a deaf person, for example, in a facility such as a nursing facility or a hospital, or in a home living with a family, such as dementia As a cause, the present invention relates to a wrinkle suppression device, a wrinkle suppression method, and a wrinkle suppression program for suppressing the behavior of a person who tries to wander out of an allowable area in which movement is permitted (ie, a wanderer).

  Conventional countermeasures against drowning behavior can be broadly divided into methods that physically prevent wrinkles by restraint and confinement, etc., and methods that detect the whereabouts of the deaf, making it easier to find decoy and generating habits. Each time there was a way for the caregiver to search for and find the deaf person and bring it back.

  However, the former has human rights issues with respect to restraint and confinement, and the latter has a large cost for equipment and systems for controlling access to people in a wide range of allowable areas, as well as nursing care. There was a problem that the burden of bringing back the deaf was not solved.

  Then, after detecting a wrinkle, the caregiver does not intervene, but the method of alerting a widow by voice alert | reporting and suppressing a wrinkle is considered. Patent document 1 is an example of such a conventional wrinkle suppression system.

  In Patent Document 1, as one embodiment, an RFID (Radio Frequency Identification) tag is installed on the sole of a person's shoe, and it is placed on the entrance floor of a living room in a facility to which the present system is applied or on the floor of a passageway. RFID tags are detected by multiple readers, and a voice notification device informs the user, “Now, XX hour (current time). XXX (user name), let's go to bed because it is late at night.” It is possible to suppress wrinkles by voice notification.

Japanese Patent Laid-Open No. 2015-75974

  However, only a prepared voice can be notified from the voice notification device, and there is a problem that there is no wrinkle suppression effect for a deaf person who does not follow the notification content.

  This indication was made in order to solve the above-mentioned subject, and is a wrinkle control device, a wrinkle control method, and a wrinkle control program which can control a wrinkle safely also for those who do not follow voice alert. It is intended to provide.

  A wrinkle suppression device according to one aspect of the present disclosure is a wrinkle suppression device for suppressing wrinkles by a deaf person, and is detected when the decoy is detected in a predetermined area inside a building. Based on the utterance part that outputs the speech to the deaf person by at least one of audio or video and the elapsed time from the start of the dialogue between the deaf person and the utterance part, A behavior intention forgetting determination unit that determines whether or not the user has forgotten, and when the behavior intention forgetting determination unit determines that the deaf person has forgotten the initial intention, the utterance unit is placed at a predetermined guidance location. An utterance for guiding a person is output by at least one of audio and video.

  According to the present disclosure, it is possible to safely suppress wrinkles even for a deaf person who does not follow voice notification.

It is a figure which shows the whole structure of the wrinkle suppression system using the wrinkle suppression apparatus which concerns on one embodiment of this indication. It is a block diagram which shows the structure of an example of the wrinkle suppression apparatus shown in FIG. It is a flowchart which shows an example of the wrinkle suppression process by the wrinkle suppression apparatus shown in FIG.

(Knowledge that became the basis of the present invention)
People with dementia who are prone to 徘徊 behavior often have short-term memory impairment, and the screening test for dementia also includes questions to confirm short-term memory. In addition, a person with dementia who has an intention of wandering has a tendency to become difficult to follow the instructions. In such a case, there is a problem that the effect of suppressing wrinkle is not achieved only by voice notification.

  In order to solve such a problem, the wrinkle suppression device according to one aspect of the present disclosure is a wrinkle suppression device for suppressing wrinkles by a deaf person, wherein the decoy person is in a predetermined region inside the building. Based on the utterance part which outputs the utterance with respect to the said deaf person by at least one of an audio | voice or an image | video when detected, and the elapsed time from the start of the dialogue of the deaf person and the said utterance part, the said deaf person And a behavior intention forgetting determination unit that determines whether or not the initial intention that led to the start of habit has been forgotten, and the utterance unit is determined by the behavior intention forgetting determination unit that the user has forgotten the initial intention In such a case, an utterance for guiding the deaf person to a predetermined guidance location is output by at least one of audio and video.

  With such a configuration, when a deaf person is detected in a predetermined area inside the building, the detected deaf person is uttered, and based on the elapsed time from the start of the dialogue with the deaf person, It is determined whether or not the person has forgotten the initial intention that has led to starting the habit, and when it is determined that the user has forgotten the initial intention, the person can be guided to a predetermined guidance location.

  Therefore, it is possible to realize the forgetting determination of the initial intention using short-term memory impairment by a simple method, and to interact with the deaf person, and by using the short-term memory disorder to the deaf person as time elapses from the start of the dialogue. After forgetting the initial intention, the person can be guided to the designated place of guidance, so that the person can be guided to the guidance place where safety is ensured without damaging the dignity of the person. be able to. As a result, wrinkles can be safely suppressed even for a deaf person who does not follow the voice notification.

  The wrinkle suppression device further includes an utterance estimating unit that estimates the content of the response utterance of the deaf person with respect to the utterance output from the utterance unit, and the behavior intention forgetting determination unit is Based on the content of the response utterance of the estimated deaf person, it may be determined whether the deaf person has forgotten the initial intention.

  With such a configuration, the content of the deaf person's response utterance to the utterance of the utterance unit is estimated, and based on the estimated content of the deceiver's response utterance, the forgetting determination of the initial intention can be more accurately performed, By interacting with the user himself / herself through dialogue, the user can understand the initial intention that led to the user's understanding and perform a dialogue that draws the user's attention. It is possible to obey the guide to a safe place without damaging the dignity of the user.

  The utterance unit performs a confirmation utterance for confirming the initial intention in an initial utterance to the deaf person, and the utterance estimation unit estimates the content of the response utterance from the deaf person for the confirmation utterance. The behavior intention forgetting determination unit determines whether the deaf person has forgotten the initial intention based on a response utterance from the deaf person with respect to the confirmation utterance estimated by the utterance estimation unit. It may be.

  With such a configuration, the content of the response utterance from the deaf person with respect to the confirmation utterance for confirming the initial intention is estimated, so that the forgetting determination of the initial intention can be performed more accurately.

  The confirmation utterance includes an utterance for encouraging an action different from the action related to the initial intention, and the utterance estimation unit is configured so that the deaf person is based on a response utterance from the deaf person with respect to the confirmation utterance. Estimating whether to follow the confirmation utterance, the behavior intention forgetting determination unit, when it is estimated by the utterance estimation unit that the deaf person follows the confirmation utterance, the deaf person forgot the initial intention You may make it determine.

  With such a configuration, a confirmation utterance for promoting an action different from the action related to the initial intention is performed, and based on the response utterance from the deaf person to the confirmation utterance, it is estimated whether or not the deaf person follows the confirmation utterance. If it is estimated that the user will follow the confirmation utterance, it is determined that the user has forgotten the initial intention. Forgetting the initial intention without reminding the user of the initial intention to start the user Since the determination can be performed more accurately and the deaf person can be guided to a behavior different from the initial intention, the deaf person can be more easily guided to a guide place where safety is ensured.

  In addition, the utterance unit may guide the deaf person to the guidance location by outputting video or audio including instruction content for actually performing the content of the confirmation utterance.

  With such a configuration, since the video or audio including the instruction content for actually performing the content of the confirmation utterance for promoting the action different from the action related to the initial intention is output, Can reliably perform different actions, and can more reliably guide the deaf person to a secure guidance location.

  In addition, when the behavior intention forgetting determination unit determines that the deaf person has forgotten the initial intention, the utterance unit performs a guided utterance that guides the deaf person to the guidance location. You may do it.

  With such a configuration, since the guidance utterance that guides the deaf person to the guide place is performed, the deaf person can be more reliably guided to the guide place where safety is ensured.

  The utterance unit may perform guidance confirmation utterance for confirming whether or not the deaf person is following the guidance when the deaf person's guidance is ongoing.

  With such a configuration, since the guidance confirmation utterance for confirming whether or not the user follows the guidance is performed during the guidance of the user, the guidance to the safe guidance location of the user by the bag suppression device is performed. This can be performed more reliably.

  In addition, the wrinkle suppression device further includes a moving unit for moving inside the building, and the wrinkle suppression device is determined by the behavior intention forgetting determination unit that the deaf person has forgotten the initial intention. In this case, the user may be guided to the guide location by moving while being accompanied by the user using the moving unit.

  With such a configuration, when it is determined that the user has forgotten the initial intention, the user can move while accompanying the user, so that the user can more reliably guide the user to the safe guide location by the device. Can be performed.

  In addition, the wrinkle suppression device further includes a movable part that shows a gesture to the deaf person, and the movable part shows a gesture to the deaf person according to the content of the utterance by the utterance part. You may make it perform operation | movement for this.

  With such a configuration, it is possible to perform utterances with gestures, so that it is possible to attract attention to a deaf person who is poor in hearing or a deaf person who is not interested only by voice notification.

  Further, the utterance unit may output the utterance by voice by outputting information representing the content of the utterance to a voice output unit that outputs voice.

  With such a configuration, it is possible to concretely realize speech utterance, and it is possible to attract attention to a poor user who has poor eyes or a deaf person who is not interested only by video display.

  The utterance unit may output the utterance as a video by outputting information representing the content of the utterance to a video output unit that outputs the video.

  With such a configuration, it is possible to specifically realize the utterance by the video, and it is possible to attract attention to a deaf person who is poor in hearing or a deaf person who does not show interest only by voice notification.

  In addition, the wrinkle suppression device further includes a deaf person detection unit that detects the deaf person based on behavior information related to the deaf person's behavior that is transmitted from an action notification device that is carried or worn by the deaf person. You may do it.

  With such a configuration, the deaf person can be detected from the behavior information related to the deer's behavior, so that the deaf person can be detected more accurately.

  The wrinkle suppression device may further include a deaf person detection unit that detects the deaf person by receiving the image of the predetermined region and analyzing the image.

  With such a configuration, it is not necessary for the deaf person to carry or wear an additional device such as an action notification device, and the decoy can be detected without placing an unnecessary burden on the deaf person.

  Further, the utterance estimation unit may estimate the content of the response utterance by referring to rule information that defines the conversation content.

  With such a configuration, it is possible to understand the content of the utterance of the deaf person by simple means.

  Further, the utterance estimation unit may estimate the content of the response utterance by using a dialogue model generated by a recursive neural network.

  With such a configuration, it is possible to understand the content of the utterance of the deaf person more accurately.

  In addition, the present disclosure can be realized not only as a wrinkle suppression device having the above-described characteristic configuration, but also a wrinkle suppression method that executes characteristic processing corresponding to the characteristic configuration of the wrinkle suppression device. It can also be realized as such. It can also be realized as a computer program that causes a computer to execute characteristic processing included in such a wrinkle suppression method. Therefore, effects similar to those of the above-described wrinkle suppression device can be obtained also in the following other aspects.

  A wrinkle suppression method according to another aspect of the present disclosure is a wrinkle suppression method in a wrinkle suppression device for suppressing wrinkles by a wanderer, when the wrinkle is detected in a predetermined area inside a building. Was the detected utterance to the deaf person output by at least one of audio or video and forgot the initial intention that the deaf person started to deaf based on the elapsed time from the start of the dialogue with the deaf person? If it is determined that the user has forgotten the initial intention, an utterance for guiding the user to a predetermined guidance location is output by voice or video.

  A wrinkle suppression program according to another aspect of the present disclosure is a wrinkle suppression program for causing a computer to function as a wrinkle suppression device for suppressing wrinkles by a deaf person. When the deaf person is detected in the area, the detected utterance for the deaf person is output by at least one of audio and video, and based on the elapsed time from the start of the dialogue with the deaf person, It is determined whether or not the initial intention that led to the start of the habit has been forgotten, and if it is determined that the deaf person has forgotten the initial intention, an utterance for guiding the deaf person to a predetermined guidance location is expressed by voice or video. The process of outputting by at least one is executed.

  Needless to say, the above-described computer program can be distributed via a computer-readable non-transitory recording medium such as a CD-ROM or a communication network such as the Internet.

  Moreover, you may comprise as a system which disperse | distributed a part of component of the wrinkle suppression apparatus which concerns on one embodiment of this indication, and the other component to several computers.

  Note that each of the embodiments described below shows a specific example of the present disclosure. Numerical values, shapes, components, steps, order of steps, and the like shown in the following embodiments are merely examples, and are not intended to limit the present disclosure. In addition, among the constituent elements in the following embodiments, constituent elements that are not described in the independent claims indicating the highest concept are described as optional constituent elements. In all the embodiments, the contents can be combined.

(Embodiment)
Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. FIG. 1 is a diagram illustrating an overall configuration of a wrinkle suppression system using a wrinkle suppression device according to an embodiment of the present disclosure. The wrinkle suppression system 11 illustrated in FIG. 1 includes a wrinkle suppression device 12, a wrinkle detector 13, and a behavior notification device 14.

  The behavior notification device 14 is carried or attached to the person HB and transmits behavior information regarding the behavior of the person HB. The behavior information is, for example, information indicating the current position of the person HB, information indicating the route the person HB has moved to the present time, and includes information for identifying the person HB. The behavior notification device 14 is composed of, for example, an RF (Radio Frequency) tag, and is attached to the body (for example, ankle) of the person HB or the shoe sole. In this case, the behavior notification device 14 communicates with the eyelid detector 13 by short-range wireless communication.

  The scissors detector 13 is installed in a predetermined area inside the building and detects a scissors. The wrinkle detector 13 is disposed at a predetermined position such as near the exit ED of a building such as a nursing facility, a hospital, or a home, and detects the person HB as a widow. The wrinkle detector 13 is composed of, for example, an RF tag reader, and determines that a deaf person has been detected when the presence of the action notification device 14 is detected around itself by short-range wireless communication.

  In addition, when it is determined that the widow has been detected, for example, the wrinkle detector 13 transmits the widow information that is information about the widow to the wrinkle suppression device 12 connected by a wired or wireless communication line. . Here, the deaf person information includes at least information indicating that the deaf person has been detected. The user information may further include the behavior information described above, that is, information indicating the current position of the user, information indicating a route the user has traveled to the present, information for identifying the user, and the like. .

  The wrinkle suppression device 12 performs a predetermined operation for suppressing wrinkles, for example, an utterance SP, on the person HB determined to be a deaf person, and the person HB performs a response utterance AS to the utterance SP. Specifically, the wrinkle suppression device 12 performs a predetermined operation such as dialogue, and when it is determined that the person HB determined to be a widow has forgotten the action intention (initial intention) that has led to wrinkles, The person HB is guided to a predetermined guiding place where is secured. Here, the predetermined operation is, for example, an utterance or a video and / or audio reproduction for forgetting a behavioral intention that the deceased has started to make deception using a short-term memory disorder. Details of a series of operations performed by the wrinkle suppression device 12 will be described later.

  For example, the wrinkle suppression device 12 includes a mobile robot having a user interface, and a humanoid robot can be used. In this case, the wrinkle suppression device 12 includes a head simulating a human head, two hands simulating a human hand, two legs simulating a human foot, and a body simulating a human torso. Parts.

  With the configuration described above, in the wrinkle suppression system 11, the wrinkle detector 13 detects the action notification device 14 attached to the person HB when the person HB is going out of the building ED with some action intention. Then, the person HB is detected as a deaf person. When it is detected that the person HB is a person, the wrinkle detector 13 transmits that fact to the wrinkle suppression device 12. When it is detected that the person HB is a person, the wrinkle suppression device 12 performs a predetermined action such as talking to the person (person HB).

  FIG. 2 is a block diagram illustrating a configuration of an example of the wrinkle suppression device 12 illustrated in FIG. 1. The wrinkle suppression device 12 shown in FIG. 2 includes a person information receiving unit 21, a voice acquisition unit 22, a voice recognition unit 23, an utterance estimation unit 24, a control unit 25, an action intention forgetting determination unit 26, an utterance sentence generation unit 27, and an utterance. A DB (database) 28, a voice synthesis unit 29, a voice output unit 30, a video generation unit 31, a video output unit 32, a moving unit 33, and a movable unit 34 are provided.

  The control unit 25 controls the wanderer information receiving unit 21, the utterance estimation unit 24, the action intention forgetting determination unit 26, the utterance sentence generation unit 27, the moving unit 33, and the movable unit 34, thereby suppressing the wrinkle suppression described below. Control the operation of the device 12.

  In FIG. 2, the illustration is simplified, and two legs, an actuator that drives these members, a drive circuit, and the like are illustrated as a moving portion 33, and two hands and these members are shown. An actuator, a drive circuit, and the like for driving are illustrated as a movable portion 34. The audio output unit 30 is provided near the mouth of the head of the humanoid robot, and the video output unit 32 is provided near the chest of the torso of the humanoid robot.

  Therefore, the wrinkle suppression device 12 can be walked on two legs by the moving unit 33 and can move inside the building. In addition, the wrinkle suppression device 12 outputs various utterances by voice and / or video by the voice output unit 30 and / or the video output unit 32 while performing various gestures for the deaf person by the movable unit 34. Can do. In addition, the structure of the wrinkle suppression apparatus 12 is not specifically limited to said example, It may be made to move using a wheel etc. instead of two legs, and various changes are possible.

  The deaf person information receiving unit 21 detects the deaf person based on the action information related to the deer's behavior transmitted from the action notification device 14 carried or attached to the deaf person. Specifically, the deaf person information receiving unit 21 receives deaf person information, which is information related to the deaf person, from the deaf detector 13, detects the deaf person based on the behavior information included in the deaf person information, and performs control. Through the unit 25, the detection result and the deaf person information are output to the utterance sentence generation unit 27 and the like.

  The utterance sentence generation unit 27 generates an utterance sentence for outputting an utterance to the detected deaf person by at least one of audio and video when a deaf person is detected in a predetermined area inside the building. In the utterance DB 28, utterance content appropriate for each of a plurality of persons that can be detected as a deaf person (for example, a utterance sentence for calling, an utterance sentence for chat, an utterance sentence for confirmation utterance, an utterance for guided utterance) Sentence, utterance sentence for guidance confirmation utterance, etc.) are stored in advance in a database format.

  Specifically, when the utterance is detected based on the utterance information, the utterance sentence generation unit 27 identifies the utterance based on the utterance information, and utters an utterance sentence suitable for the identified utterance. It is generated with reference to the DB 28 (for example, an utterance sentence for calling) and output to the speech synthesizer 29. Note that the configuration of the utterance sentence generation unit 27 is not particularly limited to the above example, and the utterance DB 28 is omitted and the utterance sentence is generated without referring to the utterance DB 28, or the call sentence stored in the external server It is possible to make various changes such as referring to the utterance sentence, the utterance sentence for chat, the utterance sentence for confirmation utterance, the utterance sentence for guidance utterance, or the utterance sentence for guidance confirmation utterance.

  The voice synthesizer 29 synthesizes a voice corresponding to the uttered sentence and outputs the synthesized voice to the voice output unit 30. The audio output unit 30 includes, for example, a speaker and outputs the audio synthesized by the audio synthesis unit 29.

  Further, the utterance sentence generation unit 27 may output the generated utterance sentence to the video generation unit 31. In this case, the video generation unit 31 generates a character string image in a text format corresponding to the uttered sentence and outputs it to the video output unit 32. The video output unit 32 is composed of, for example, a display and outputs a character string image generated by the video generation unit 31.

  In this way, in the present embodiment, when a deaf person is detected in a predetermined area inside the building, an utterance for the detected deaf person is output by at least one of audio or video to Start talking and call the deaf person as a confirmation utterance to confirm the initial intention.

  The voice acquisition unit 22 is composed of a microphone or the like, and when the deaf person makes a response utterance in response to the utterance such as the above call, the sound of the response utterance from the deaf person is acquired and output to the voice recognition unit 23. To do. The voice recognition unit 23 recognizes the voice of the response utterance, generates text data including a character string, and outputs the text data to the utterance estimation unit 24. The utterance estimation unit 24 estimates the content of the response utterance from the text data, and outputs the estimation result to the action intention forgetting determination unit 26 and the utterance sentence generation unit 27 via the control unit 25.

  At this time, the utterance sentence generation unit 27 refers to the utterance DB 28, generates an utterance sentence for the deaf person according to the estimated content of the response utterance, and outputs the utterance sentence to the speech synthesis unit 29. The voice synthesizer 29 synthesizes a voice corresponding to the utterance sentence and outputs the synthesized voice to the voice output part 30. The voice output part 30 reads out the utterance sentence generated by the utterance sentence generator 27 and utters the utterance to the deaf person. I do.

  For example, when talking to the deaf person is started using the voice output unit 30 or the like, and there is a response utterance from the deaf person, the utterance estimation unit 24 uses the voice acquisition unit 22 or the like to reply from the deaf person The utterance is acquired and the content of the response utterance is estimated, and the utterance sentence generation unit 27 generates an utterance sentence related to the estimated content of the response utterance and outputs it to the speech synthesizer 29. The voice output unit 30 uses the voice synthesized by the voice synthesis unit 29 to utter a chat or the like related to the content of the response utterance in order to continue the dialogue with the deaf person. Here, the content of the utterance estimated by the utterance estimation unit 24 includes the action intention (initial intention) that led the user to start the habit.

  The estimation of the utterance content and the generation of the utterance sentence according to the estimated utterance content can be realized, for example, by referring to the rule information defining the conversation content or by generating the conversation model by a recursive neural network. .

  That is, the utterance estimation unit 24 may store rule information defining conversation contents in an internal memory in advance, and estimate the contents of response utterances by referring to the rule information. Further, the utterance estimation unit 24 may estimate the content of the response utterance by using a dialogue model generated by a recursive neural network. Similarly, the utterance sentence generation unit 27 stores rule information that defines an utterance sentence corresponding to the estimated utterance content in an internal memory in advance, and refers to this rule information to thereby utter an utterance corresponding to the estimated utterance content. A sentence may be generated. Further, the utterance sentence generation unit 27 may generate an utterance sentence corresponding to the estimated utterance content by using a dialogue model generated by a recursive neural network.

  The behavior intention forgetting determination unit 26 determines whether or not the user has forgotten the initial intention and outputs the determination result to the control unit 25. The determination of whether or not the initial intention has been forgotten is performed based on, for example, whether or not a certain time has elapsed since the start of the dialogue with the user. In this case, the action intention forgetting determination unit 26 acquires the conversation content with the deaf person from the utterance estimation unit 24 via the control unit 25, acquires the start time of the dialog using an internal timer, etc. When a certain time has elapsed from the start, it is determined that the user has forgotten the initial intention. Such a determination is based on the characteristic that the user forgets the initial intention by continuing the dialogue for a certain period of time (for example, 5 minutes, preferably 10 minutes, more preferably 20 minutes). .

  Specifically, the behavior intention forgetting determination unit 26 guides the deaf person to a predetermined guidance place where a confirmation utterance for promoting an action different from the action related to the action intention or a safety is ensured after starting the dialogue. When a fixed time has elapsed without performing the guided utterance, or when a fixed time has elapsed since the previous guided utterance, it is determined that the user has forgotten the initial intention, and the determination result is output to the control unit 25. .

  In addition, the determination method of the intention whether or not the initial intention has been forgotten is not particularly limited to the above example. Instead of the determination of the passage of time or in addition to the determination of the passage of time, for example, You may perform based on the content of the utterance by a deaf person. In this case, when the content of the response utterance estimated by the utterance estimation unit 24 is irrelevant to the initial intention, the behavior intention forgetting determination unit 26 may determine that the user has forgotten the initial intention.

  In the initial utterance to the deaf person, the utterance sentence generation unit 27 generates an utterance sentence for confirmation utterance for confirming the initial intention, and the voice output unit 30 confirms the utterance confirmed by the voice synthesizer 29. , The utterance estimation unit 24 estimates the content of the user's response utterance with respect to the confirmation utterance, and the behavior intention forgetting determination unit 26 determines that the content of the response utterance from the user with respect to the confirmation utterance is irrelevant to the initial intention. It may be determined that the user has forgotten the initial intention.

  Next, when a certain period of time has elapsed from the start of the conversation and the behavior intention forgetting determination unit 26 determines that the deaf person has forgotten the initial intention, the utterance sentence generation unit 27 displays the determination result via the control unit 25. Acquired, referring to the utterance DB 28, generates an utterance sentence for confirmation utterance for encouraging an action different from the action related to the initial intention, and outputs it to the speech synthesizer 29. The voice synthesizing unit 29 synthesizes a voice corresponding to the utterance sentence for the confirmation utterance and outputs the synthesized voice to the voice output unit 30, and the voice output unit 30 outputs the confirmation utterance by voice.

  Here, the utterance sentence generation unit 27 may output the generated utterance sentence for confirmation utterance to the video generation unit 31. In this case, the video generation unit 31 generates a text-format character string image corresponding to the utterance sentence for the confirmation utterance and outputs it to the video output unit 32. The video output unit 32 outputs the confirmation utterance as a video.

  When the deaf person makes a response utterance in response to the confirmation utterance, the voice acquisition unit 22 acquires the response utterance voice from the deaf person in response to the confirmation utterance and outputs the voice to the voice recognition unit 23. Recognizes the voice of the response utterance, generates text data composed of a character string, and outputs the text data to the utterance estimation unit 24. The utterance estimation unit 24 estimates whether or not the user follows the confirmation utterance from the text data that has been voice-recognized, and outputs the estimation result to the utterance sentence generation unit 27 via the control unit 25.

  Further, the utterance estimation unit 24 outputs the estimation result to the behavior intention forgetting determination unit 26 via the control unit 25, and the behavior intention forgetting determination unit 26 follows the confirmation utterance in which the deaf person promotes different actions. If it is estimated, it may be determined that the deaf person has forgotten the initial intention.

  When it is determined that the deaf person follows the confirmation utterance, the utterance sentence generation unit 27 refers to the utterance DB 28, generates a utterance sentence for the guidance utterance that guides the deaf person to the guidance place where safety is ensured, and performs speech. The data is output to the combining unit 29. The voice synthesizer 29 synthesizes the voice corresponding to the utterance sentence for the guided utterance and outputs the synthesized voice to the voice output unit 30, and the voice output unit 30 outputs the guided utterance by voice.

  Here, the utterance sentence generator 27 may output the generated utterance sentence for the guided utterance to the video generator 31. In this case, the video generation unit 31 generates a text-format character string image corresponding to the utterance sentence for the guided utterance and outputs it to the video output unit 32. The video output unit 32 outputs the guided utterance as a video.

  As described above, the wrinkle suppression device 12 guides to a guidance place where safety is ensured if the deaf person obeys the confirmation utterance, and re-interacts if he does not follow the confirmation utterance. After continuing, try the confirmation utterance again.

  Next, when the deaf person makes a response utterance in response to the above-described guided utterance, the voice acquisition unit 22 acquires the voice of the response utterance from the deaf person in response to the induced utterance, and outputs the voice to the voice recognition unit 23. The recognition unit 23 recognizes the voice of the response utterance, generates text data including a character string, and outputs the text data to the utterance estimation unit 24. The utterance estimation unit 24 estimates whether or not the user follows the guided utterance from the speech-recognized text data, and outputs the estimation result to the control unit 25.

  When it is determined that the deaf person follows the guided utterance, the control unit 25 controls the moving unit 33 so as to move to the guidance location where safety is ensured. In this manner, the wrinkle suppression device 12 uses the moving unit 33 when it is determined by the behavior intention forgetting determination unit 26 that the user has forgotten the initial intention and when the user is determined to follow the guidance by the guided utterance. The person is guided to the guidance place by moving while accompanying the person.

  The movable part 34 shows gestures to the deaf person using two hands, etc., and performs, for example, a shoulder hitting operation to the deaf person and a guide operation to the guide place. Specifically, the movable unit 34 is controlled by the control unit 25 and performs an operation for gesturing the deaf person according to the content of the utterance by the audio output unit 30 or the video output unit 32.

  In addition, when the user's guidance is continuing, the utterance sentence generation unit 27 refers to the utterance DB 28 and generates an utterance sentence for the guidance confirmation utterance for confirming whether or not the user follows the guidance. To the speech synthesizer 29. The voice synthesizing unit 29 synthesizes voice corresponding to the utterance sentence for the guidance confirmation utterance and outputs the synthesized voice to the voice output unit 30. The voice output unit 30 outputs the guidance confirmation utterance by voice.

  Here, the utterance sentence generation unit 27 may output the generated utterance sentence for the guidance confirmation utterance to the video generation unit 31. In this case, the video generation unit 31 generates a text string image corresponding to the utterance sentence for the guidance confirmation utterance and outputs it to the video output unit 32. The video output unit 32 outputs the guidance confirmation utterance as a video. To do.

  As described above, when it is determined from the guidance confirmation utterance that the deaf person obeys the guidance utterance, the heel suppression device 12 continues the guidance to the guidance location where safety is ensured and determines that the guidance utterance is not followed. If this happens, try again after continuing the conversation again.

  In the wrinkle suppression device 12 configured as described above, the utterance sentence generation unit 27, the utterance DB 28, the speech synthesis unit 29, and the video generation unit 31 correspond to an example of the utterance unit, and the action intention forgetting determination unit 26 sets the action intention. It corresponds to an example of a forgetting determination unit, the voice acquisition unit 22, the voice recognition unit 23, and the utterance estimation unit 24 correspond to an example of an utterance estimation unit, the control unit 25 and the movement unit 33 correspond to an example of a movement unit, The control unit 25 and the movable unit 34 correspond to an example of a movable unit, the audio output unit 30 corresponds to an example of an audio output unit, the video output unit 32 corresponds to an example of a video output unit, and the visitor information receiving unit 21. Corresponds to an example of a deaf person detection unit.

  FIG. 3 is a flowchart illustrating an example of wrinkle suppression processing by the wrinkle suppression device illustrated in FIG. 1.

  As shown in FIG. 3, the wrinkle suppression device 12 is first in a standby state (S31), and determines whether or not a deaf person has been detected at a predetermined interval (S32). As a method for detecting the deaf person, for example, the deaf person information receiving unit 21 receives the behaviour information including the action information from the beak detector 13 by receiving the behaviour information transmitted from the behaviour detector 13. This is done by determining whether or not it has been done.

  As a result of the determination, if the deaf person is not detected (No in S32), the wrinkle suppression device 12 returns to the standby state (S31) and continues to determine whether or not the deaf person has been detected (S32).

  On the other hand, when a deaf person is detected (Yes in S32), the wrinkle suppression device 12 performs a dialogue with the deaf person using the voice output unit 30 or the like (S33). As a dialogue with the deaf person, for example, the wrinkle suppression device 12 first utters a talk to the deaf person, and then performs a confirmation utterance for confirming the initial intention of the deaf person. After that, the wrinkle suppression device 12 utters chats or the like and continues the dialogue with the widow. In the dialogue with the deaf person, the wrinkle suppression device 12 may wait in advance around the position where the deaf person is detected, or may move from a predetermined standby position to the position where the deaf person is detected. Good.

  Next, the behavior intention forgetting determination unit 26 of the wrinkle suppression device 12 determines whether or not a certain time has elapsed since the start of the conversation with the widow (S34). As a method for determining whether or not a certain period of time has elapsed, for example, the behavior intention forgetting determination unit 26 obtains the start time of the conversation from an internal timer or the like, and whether the difference between the start time and the current time exceeds a predetermined value. This is done by determining whether or not.

  If it is not determined that the predetermined time has elapsed (No in S34), the process returns to step S33, and the wrinkle suppression device 12 continues the conversation with the deaf person by speaking such as chat.

  On the other hand, when it is determined that a certain time has elapsed (Yes in S34), the behavior intention forgetting determination unit 26 determines that the deaf person has forgotten the initial intention, and the wrinkle suppression device 12 uses the audio output unit 30 or the like. Thus, a confirmation utterance for promoting an action different from the action related to the initial intention is performed (S35), and the utterance estimation unit 24 determines whether or not the deaf person follows the confirmation utterance (S36). For example, the utterance estimation unit 24 estimates the content of the user's response utterance to the confirmation utterance, and the content of the estimated utterance indicates consent to the confirmation utterance. This is done by determining whether or not.

  When it is determined that the deaf person does not follow the confirmation utterance (No in S36), the process returns to step S33, and the wrinkle suppression device 12 uses the voice output unit 30 or the like to perform a dialogue with the deaf person by utterance such as chat. continue.

  On the other hand, when it is determined that the deaf person follows the confirmation utterance (Yes in S36), the wrinkle suppression device 12 uses the voice output unit 30 or the like to perform a guided utterance that guides the deaf person to a safe guidance place. The deaf person is guided to the deaf place where safety is ensured (S37). As a guide method for the deaf person, for example, the wrinkle suppression device 12 uses the moving unit 33 or the like to accompany the deaf person to a predetermined guide place where safety is ensured.

  Here, the guidance method for the deaf person is not particularly limited to the above example, and the wrinkle suppression device 12 uses the voice output unit 30 or the like to specifically execute the content of the confirmation utterance. May be performed by video or audio, or the movable unit 34 may be used to perform a support operation for actually performing the content of the guided utterance.

  In addition, when performing guidance to a predetermined guidance location, the utterance sentence generation unit 27 may perform utterance that teaches how to reach the predetermined guidance location from the current position by voice or video (screen display). As described above, as a guidance method for the deaf person, an operation according to the ability of the wrinkle suppression device 12 is performed.

  Next, the wrinkle suppression device 12 determines whether or not the user is following the guidance when continuously guiding the user by speaking or moving (S38). As a method for determining whether or not to continue following the guidance, for example, the wrinkle suppression device 12 uses the voice output unit 30 or the like to perform a guidance confirmation utterance for confirming whether or not the deaf person is following the guidance. Utters a response utterance in response to the guidance confirmation utterance, the utterance estimation unit 24 acquires the response utterance from the deaf person using the voice acquisition unit 22 and the like, estimates the content of the response utterance, and the estimated utterance It is performed by determining whether or not the content of the content indicates consent to the guidance confirmation utterance, or by determining whether or not the dialogue with the deaf person can be continued.

  The method for determining whether to continue following the guidance is not particularly limited to the above example, and the wrinkle suppression device 12 is provided with a detector for detecting a deaf person, such as a human sensor, so that the wrinkle suppression device 12 is guided. When performing the movement for this purpose, it may be performed by determining whether or not a deaf person is detected at a position within a predetermined range from the position of the wrinkle suppression device 12.

  If it is determined that the deaf person is not following the guidance (No in S38), the process returns to step S33, and wrinkle suppression device 12 performs the dialogue with the deaf person again by speaking such as chatting and continues the subsequent processing. .

  On the other hand, when it is determined that the deaf person continues to follow the guidance (Yes in S38), the wrinkle suppression device 12 determines whether the guidance is finished (S39). As a method for determining whether or not the guidance has ended, for example, the wrinkle suppression device 12 performs the determination by determining whether or not the user has been guided to the guidance location.

  If the guidance has not ended (No in S39), the process returns to step S37, and the wrinkle suppression device 12 continues the guidance to the deaf person. On the other hand, when the guidance is completed (Yes in S39), the kite suppression device 12 ends the kite suppression process and returns to the vicinity of the exit of the building, for example, in order to suppress the next kite.

  Next, a specific example of the dialogue with the deaf person by the wrinkle suppression device 12 according to the present embodiment will be described. In the following example, the wrinkle suppression device 12 is described as a mobile robot.

  First, the wrinkle suppression device 12 utters a call to the person HB determined to be a widow. The call utterance is, for example, “Go out? Where are you going?”.

  Next, the wrinkle suppression device 12 acquires a response utterance from the person HB to the utterance of the call. For example, the wrinkle suppression device 12 acquires an utterance such as “I have to go home and cook rice” as a response utterance. Thereby, the wrinkle suppression device 12 understands that the initial intention of the widow is “to go home and cook rice”.

  Thereafter, wrinkle suppression device 12 continues the dialogue with the widow. In continuing the dialogue, for example, the wrinkle suppression device 12 utters further questions regarding the user's response utterance, such as “Which is the house?” Or utterance of a chat unrelated to the user's response utterance. Etc.

  Then, after a certain period of time has elapsed from the start of the conversation with the deaf person, wrinkle suppression device 12 performs a confirmation utterance to promote an action different from the initial intention. For example, utter a utterance such as “Speaking of that, it ’s time to start watching the program that we are always watching. Let ’s go to the cafeteria.”

  In response to the above confirmation utterance, for example, when a response utterance such as “I have to go to see it sooner?” Is acquired from the user, the device suppresses the determination that the user follows the confirmation utterance. The guidance utterance based on the content is performed, and the heel control device 12 guides the heel to the dining room while accompanying the heel. On the other hand, when a response utterance that rejects the content of the confirmation utterance or the guided utterance or a response utterance having content that is irrelevant to the confirmation utterance or the guided utterance is acquired from the deaf person, the wrinkle suppression device 12 may Decide not to follow and continue conversations such as chat.

  In the example shown in FIG. 3 described above, the wrinkle suppression device 12 determines that the initial intention has been forgotten after a certain period of time, and performs confirmation utterance or guided utterance, but whether or not the initial intention has been forgotten. This determination is not limited to this example, and may be performed based on the utterance content.

  For example, at any timing after the dialogue for confirming the initial intention of the deaf person, the wrinkle suppression device 12 uses “Soon 12 as a confirmation utterance for prompting an action different from the action related to the initial intention”. You may be allowed to make an utterance about the daily activities of the deaf person, such as “It's time for lunch.” If a response utterance such as “Oh, is that time?” Is acquired from the deaf person as a response utterance to follow this confirmation utterance, the behavior intention forgetting determination unit 26 must return to the house and cook rice. May have been forgotten, and the wrinkle suppression device 12 may continue to perform a guided utterance such as “Let's go to the cafeteria”.

  As described above, in the wrinkle suppression device 12 according to the present embodiment, the user can forget the initial intention of the user by directly acting on the user and, after that, the user can be guided. Because you don't feel that your will has been restrained, you can suppress jealousy without compromising human dignity.

  As described above, the wrinkle suppression device according to one aspect of the present disclosure has been described based on the above embodiment, but the present disclosure is not limited to the above embodiment. Unless it deviates from the gist of the present disclosure, forms in which various modifications conceived by those skilled in the art have been made in the present embodiment or a combination of components in different embodiments are also included in the scope of the present disclosure. It is.

  For example, in the present embodiment, the deaf person is detected by communication between the wrinkle detector 13 installed in a predetermined area inside the building and the action notification device 14 attached to the person HB. The method may be other methods. For example, instead of the eyelid detector 13 and the person information receiving unit 21, image data obtained by photographing a predetermined area inside a building is acquired by a camera installed at a predetermined position, and facial image recognition is performed on the image data. It is also possible to use a deaf person detection unit that detects a deaf person depending on whether or not a predetermined person has been detected. In addition, video data shot by a camera installed at a predetermined position is acquired, and by video analysis, for example, a deaf person is detected based on whether or not a specific person moves back and forth in the same place. A deaf person detection unit may be used.

  Further, in the present embodiment, the wrinkle detector 13 and the wrinkle suppression device 12 are separated from each other, but the wrinkle detection device 12 may be mounted on the wrinkle suppression device 12. In this case, the wrinkle suppression device 12 stands by at a predetermined position, for example, near the exit ED of the building, and the wrinkle detector 13 mounted on the wrinkle suppression device 12 is It is possible to detect wrinkles based on whether or not the action information related to the person's action is received and the transmitted action information can be received.

  In the present embodiment, wrinkle suppression device 12 is a mobile robot, but it may be a stationary robot or a stationary display that displays a guided utterance or the like as an image. Further, when the wrinkle suppression device 12 is a mobile or stationary robot, the shape may not be a human type, and may be an animal type or an arbitrary character type. In addition, a wrinkle suppression device that interacts with the user using a microphone or a speaker embedded in a predetermined location around the detection position of the user and that does not involve a target that the user can visually recognize as a partner of the communication may be used. .

  In addition, the content of the utterance in the conversation between the deaf person and the deering apparatus may be based on the deer's profile, hobbies, preferences, life data, vital data, and the like. In this case, for example, wrinkle suppression device 12 receives necessary data through communication with a server device that stores these data, and generates an utterance that includes at least part of the received content. This makes it possible to continue the dialogue that attracts the attention of the deaf.

  In the present embodiment, the time required for forgetting the initial intention is passed mainly by continuing the voice conversation. However, the wrinkle suppression device 12 is used together with or instead of the voice conversation. The user's attention may be drawn by displaying video data in the video output unit 32 which is a display provided, or by a physical operation of the movable unit 34 provided in the wrinkle suppression device 12. At this time, it is also effective to attract attention to change the height and angle of the video output unit 32 and the movable unit 34 according to the eyes of the deaf person.

  Further, in this embodiment, the wrinkle is suppressed by forgetting the initial intention, but the wrinkle may be suppressed by recognizing that the initial intention has already been solved by sound or video. For example, when the initial intention is “go home and cook rice”, the wrinkle suppression device 12 displays on the display visible to the deaf person a video indicating that “the family is already eating rice at home” You may make it reproduce | regenerate the condition "the family is already eating rice at home" by transmitting the instruction information for outputting a sound. Instead of transmitting information as audio or video on the display, it may be recognized that the purpose has already been solved in virtual reality.

  The wrinkle suppression device, the wrinkle suppression method, and the wrinkle suppression program according to the present disclosure can safely suppress wrinkles even for a deaf person who does not follow voice notification. It is useful as a wrinkle suppression method and a wrinkle suppression program.

DESCRIPTION OF SYMBOLS 11 Haze suppression system 12 Haze suppression apparatus 13 Haze detector 14 Behavior notification apparatus 21 Person information receiving part 22 Voice acquisition part 23 Voice recognition part 24 Speech estimation part 25 Control part 26 Action intention forgetting determination part 27 Speech sentence generation part 28 Utterance DB
29 voice synthesis unit 30 audio output unit 31 video generation unit 32 video output unit 33 moving unit 34 movable unit

Claims (17)

A wrinkle suppression device for suppressing wrinkles by a deaf person,
An utterance unit that outputs an utterance to the detected deaf person by at least one of audio or video when the deaf person is detected in a predetermined area inside the building;
An action intention forgetting determination unit that determines whether or not the deaf person has forgotten the initial intention that led to the deception based on the elapsed time from the start of the conversation between the deaf person and the utterance unit,
The utterance unit outputs an utterance for guiding the deaf person to a predetermined guidance place by at least one of audio or video when the deliberate decision of the action intention forgetting is determined to have forgotten the initial intention. To
Wrinkle suppression device. A speech estimation unit for estimating the content of the response speech of the deaf person with respect to the speech output from the speech unit;
The behavior intention forgetting determination unit determines whether the deaf person has forgotten the initial intention based on the content of the response utterance of the deaf person estimated by the utterance estimation unit.
The wrinkle suppression device according to claim 1. The utterance unit performs a confirmation utterance to confirm the initial intention in an initial utterance for the deaf person,
The utterance estimation unit estimates the content of a response utterance from the deaf person to the confirmation utterance,
The behavior intention forgetting determination unit determines whether or not the user has forgotten the initial intention based on the response utterance from the user for the confirmation utterance estimated by the utterance estimation unit.
The wrinkle suppression device according to claim 2. The confirmation utterance includes an utterance for prompting an action different from the action related to the initial intention,
The utterance estimation unit estimates whether or not the deaf person follows the confirmation utterance based on the response utterance from the deaf person to the confirmation utterance,
The behavior intention forgetting determination unit determines that the deaf person has forgotten the initial intention when it is estimated by the utterance estimation unit that the deaf person follows the confirmation utterance;
The wrinkle suppression device according to claim 3. The utterance unit guides the deaf person to the guidance location by outputting video or audio including instruction content for actually performing the content of the confirmation utterance,
The wrinkle suppression device according to claim 4. The utterance unit, when it is determined by the behavior intention forgetting determination unit that the deaf person has forgotten the initial intention, for the deaf person, to perform a guided utterance to guide the deaf person to the guidance location,
The wrinkle suppression device according to any one of claims 1 to 5. The utterance unit performs a guidance confirmation utterance for confirming whether or not the deaf person is following the guidance when the deaf person's guidance is ongoing,
The wrinkle suppression device according to any one of claims 1 to 6. A moving part for moving inside the building;
When the behavior intention forgetting determination unit determines that the deaf person has forgotten the initial intention, the wrinkle suppression device moves the user while accompanying the deaf person using the moving unit. Guiding to the guidance location,
The wrinkle suppression device according to any one of claims 1 to 7. It further comprises a movable part that shows gestures to the deaf person,
The movable part performs an operation for gesturing the deaf person according to the content of the utterance by the utterance part.
The wrinkle suppression device according to any one of claims 1 to 8. The utterance unit outputs the utterance by voice by outputting information representing the content of the utterance to a voice output unit that outputs voice.
The wrinkle suppression device according to any one of claims 1 to 9. The utterance unit outputs the utterance as a video by outputting information representing the content of the utterance to the video output unit that outputs the video.
The wrinkle suppression device according to any one of claims 1 to 10. Further comprising a deaf person detection unit for detecting the deaf person based on action information related to the deer's action transmitted from the action notification device carried or attached to the deaf person.
The wrinkle suppression device according to any one of claims 1 to 11. Further comprising a deaf person detection unit for detecting the deaf person by receiving the image of the predetermined region and analyzing the image;
The wrinkle suppression device according to any one of claims 1 to 11. The utterance estimation unit estimates the content of the response utterance by referring to the rule information defining the conversation content.
The wrinkle suppression device according to any one of claims 1 to 13. The utterance estimation unit estimates the content of the response utterance by using a dialogue model generated by a recursive neural network.
The wrinkle suppression device according to any one of claims 1 to 14. A method for suppressing wrinkles in a wrinkle suppression device for suppressing wrinkles by a deaf person,
When the deaf person is detected in a predetermined area inside the building, the utterance to the detected deaf person is output by at least one of audio or video,
Based on the elapsed time from the start of the dialogue with the deaf person, determine whether the foreman has forgotten the initial intention that led to the deception,
When it is determined that the deaf person has forgotten the initial intention, an utterance for guiding the deaf person to a predetermined guidance location is output by at least one of audio and video.
How to suppress wrinkles. A wrinkle suppression program for causing a computer to function as a wrinkle suppression device for suppressing wrinkles by a deaf person,
In the computer,
When the deaf person is detected in a predetermined area inside the building, the utterance to the detected deaf person is output by at least one of audio or video,
Based on the elapsed time from the start of the dialogue with the deaf person, determine whether the foreman has forgotten the initial intention that led to the deception,
When it is determined that the deaf person has forgotten the initial intention, an utterance for guiding the deaf person to a predetermined guidance location is output by at least one of audio and video.
A wrinkle suppression program that executes processing.

Images