JP2007213315A - Live assisting device, life assisting method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To alleviate a user's psychological burden by giving a trigger for reminding a user of the content of habitual daily activity even if the user forgets the activity. <P>SOLUTION: This life assisting device comprises an activity pattern input part for inputting an activity pattern; a storage part for storing a plurality of inputted patterns paired with input times thereof; and a voice input part for inputting a keyword related to either of the activity pattern and the input time by a voice. The life assisting device further comprises a voice recognition part for analyzing the voice inputted from the voice input part and recognizing the keyword; a voice synthesis part for synthesizing the activity pattern in the storage part as voice; a voice output part for outputting the synthesized voice; and a control part for selecting the activity pattern corresponding to the recognized keyword from a plurality of activity patterns in the storage part and controlling the voice synthesis part based on the selected activity pattern to synthesize the pattern as a voice. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a life assistance device, a life assistance method, and a program.

In recent years, when there is an abnormality in a user's daily life, an abnormality determination system for a living behavior pattern that detects the abnormality and notifies a monitoring person has been developed (for example, see Patent Document 1). As a result, even if an elderly person living alone falls due to some accident, the abnormality is notified to the supervisor, so that it is possible to cope with it quickly.
JP 2005-173668 A

  However, the above system is to notify the supervisor when an abnormality occurs, but if the user himself forgets the behavior that is habitual due to forgetting things, It was difficult.

  An object of the present invention is to reduce the mental burden of a user by giving the user a chance to remind the user of the behavioral content even if the user forgets about a routine daily behavior. That is.

The life assistance device in the invention according to claim 1 is:
An action pattern input unit for inputting an action pattern;
A storage unit that stores a plurality of sets of the action pattern and the input time of the action pattern that are input to the action pattern input unit;
A voice input unit in which a keyword related to any one of the behavior pattern and the input time is input by voice;
A voice recognition unit that analyzes voice input from the voice input unit and recognizes the keyword;
A speech synthesizer that synthesizes the behavior pattern in the storage unit as speech;
A voice output unit for outputting the voice synthesized by the voice synthesis unit;
The behavior pattern corresponding to the keyword recognized by the speech recognition unit is selected from the plurality of behavior patterns in the storage unit, the speech synthesis unit is controlled based on the selected behavior pattern, and the speech And a control unit for synthesizing as a feature.

The invention according to claim 2 is the life assistance device according to claim 1,
When the input time of the action pattern is reached, the control unit controls the voice synthesis unit based on the action pattern in the storage unit that is paired with the input time, and synthesizes the voice as a voice.

The life assistance method in the invention of claim 3 is:
An action pattern input process in which an action pattern is input;
A storage step of storing a plurality of the behavior patterns input in the behavior pattern input step;
A voice input step in which keywords related to the behavior pattern are input by voice;
Analyzing the voice input in the voice input step and recognizing the keyword;
A selection step of selecting the behavior pattern corresponding to the keyword recognized in the voice recognition step from the plurality of behavior patterns stored in the storage step;
A speech synthesis step of synthesizing the behavior pattern selected in the selection step as speech;
A voice output step of outputting the voice synthesized in the voice synthesis step.

The program according to the invention of claim 4 is:
An action pattern input step in which an action pattern is input;
A storage step for storing a plurality of the behavior patterns input in the behavior pattern input step;
A voice input step in which keywords related to the behavior pattern are input by voice;
A speech recognition step of analyzing the speech input in the speech input step and recognizing the keyword;
A selection step of selecting the behavior pattern corresponding to the keyword recognized in the voice recognition step from the plurality of behavior patterns stored in the storage step;
A speech synthesis step of synthesizing the behavior pattern selected in the selection step as speech;
A voice output step of outputting the voice synthesized in the voice synthesis step.

  According to the present invention, when a keyword related to any one of the action pattern and the input time is input to the voice input unit by voice, the voice recognition unit analyzes the voice and recognizes the keyword. Then, the control unit selects an action pattern corresponding to the keyword recognized by the voice recognition unit from a plurality of action patterns in the storage unit, and controls the voice synthesis unit based on the selected action pattern to To synthesize. As a result, the behavior pattern corresponding to the keyword is output from the voice output unit, so that even if the user forgets the daily behavior, the user can be reminded of the content of the behavior, and the user's spirit Burden can be reduced.

  Also, when the action pattern input time comes, the speech synthesizer synthesizes as speech based on the action pattern in the storage unit paired with the input time, so when the action pattern is input, the action pattern is Will be output. Thereby, even if there is no keyword input from a user, a user can be notified of an action pattern.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing the overall configuration of the life assistance device 1 of the present embodiment. As shown in FIG. 1, the life assistance device 1 includes a voice input unit 2, a voice recognition unit 3, a changeover switch 4, a storage unit 5, a voice synthesis unit 6, a voice output unit 7, and a control. Part 8 is provided.

The voice input unit 2 is composed of, for example, a microphone or the like, and is adapted to input the content of a message from the user.
The voice recognizing unit 3 is for recognizing the content of a message input to the voice input unit 2. Specifically, the speech recognition unit 3 recognizes the content of the speech by acquiring the speech waveform of the content of the speech from the speech input unit 2. The speech recognition unit 3 is provided with a database 31 that stores words, speech waveforms of the words, and uses in association with each other. Then, the speech recognition unit 3 compares the acquired speech waveform (hereinafter, acquired waveform) with the speech waveform of each word in the database 31 and extracts a word having a waveform approximate to the acquired waveform. At this time, one or more words having a waveform approximate to the acquired waveform may be extracted. In this case, it is determined that the word having the closest waveform matches the acquired waveform.

  The change-over switch 4 is used for switching whether the content of the statement from the user indicates a question or an action pattern. For example, when the user inputs a question, the user himself / herself operates the changeover switch 4 to switch to the question input mode, and when the user inputs a behavior pattern, the user himself / herself operates the changeover switch 4 to change the behavior pattern input mode. Switch to. Since the switching instruction by the selector switch 4 is input to the control unit 8, the control unit determines whether the utterance content recognized by the voice recognition unit 3 indicates a question or an action pattern. It can be judged on the 8th side.

  The storage unit 5 stores, as an action pattern, a word, a voice waveform, and a use that are determined by the control unit 8 that the utterance content recognized by the voice recognition unit 3 is input in the action pattern input mode. At this time, since the input time of the content of the message is input from the control unit 8 in the storage unit 5, the action pattern and the input time are stored as a set. When a plurality of action patterns are input, the storage unit 5 stores a plurality of sets of action patterns and input times of the action patterns.

Specifically, the stored contents will be described with reference to FIGS. 2 and 3 are explanatory diagrams illustrating an example of a list of behavior patterns stored in the storage unit 5.
In the list H1 shown in FIG. 2, an action “medicine after breakfast” is set every day from 7 o'clock, and an action “medication after dinner” is set from 18:00. In addition, an action “walk” is set at 10:00 on Monday, Tuesday, Thursday, and Friday, and an action “go to hospital” is set at 10:00 on Wednesday.
If a new action “walking dog” is newly added to the list H1 of FIG. 2 from 16:00 every day, a list H2 as shown in FIG. 3 is formed and stored in the storage unit 5. Will be.

The voice synthesizer 6 reads out a voice waveform as an action pattern stored in the storage unit 5 and synthesizes it to create a voice waveform of the entire response content.
The voice output unit 7 is composed of, for example, a speaker, and outputs the response content as a voice based on the voice waveform of the entire response content created by the voice synthesis unit 6.

  When the changeover switch 4 is instructed to switch the action pattern input mode, the control unit 8 stores, in the storage unit 5, words, voice waveforms, and uses based on the content of the speech recognized by the voice recognition unit 3 as action patterns. Remember. On the other hand, when the changeover switch 4 instructs to switch the question input mode, the control unit 8 recognizes the content of the speech recognized by the voice recognition unit 3 as a keyword. Thereafter, the control unit 8 selects an action pattern corresponding to the keyword from a plurality of action patterns in the storage unit 5, and controls the speech synthesis unit 6 based on the voice waveform of the selected action pattern, It is designed to synthesize as speech.

  Here, when selecting an action pattern corresponding to a keyword, it has a use related to the use among the word, the speech waveform, and the use of the keyword input in the question input mode and recognized by the speech recognition unit 3. An action pattern is selected from a plurality of action patterns in the storage unit 5.

Hereinafter, behavior patterns having uses related to the use of keywords will be described.
For example, if the question “Where is 17:00?” Is input, the usage of “17:00” is “Hourly”, and the usage of “Where?” Is “Question about location”. The control unit 8 extracts the action pattern set near the time for a word for which “time signal” is used, by the control unit 8. After that, the control unit 8 determines that the action pattern corresponds to the keyword if the use of the extracted action pattern is related to “place” such as “destination” and “place”, for example. Then select.

  In addition to this, when there is no keyword having the use of “time signal” in the question content, the control unit 8 responds to the keyword based on the time when the question is input and the use of the question content. Select a behavior pattern. More specifically, for example, “Where are you going?” Is entered as a question without a keyword that has a “time signal” application, and it is related to “destination”, which is the application of “where to go”. An action pattern having a use is selected by the control unit 8.

  Next, a program executed by the life assistance device 1 of this embodiment will be described. By executing this program, the interaction method of the present embodiment is realized by the life assistance device 1.

  FIG. 4 is a flowchart of a program executed by the life assistance device 1. As shown in FIG. 4, first, in step S1, the control unit 8 determines whether the instruction from the changeover switch 4 is a switching instruction to the behavior pattern input mode or a switching instruction to the question input mode. If it is an instruction to switch to the action pattern input mode, the process proceeds to step S2, and if it is an instruction to switch to the question input mode, the process proceeds to step S10.

  In step S2, the control unit 8 determines whether or not the instruction from the changeover switch 4 has been switched to the instruction to switch to the question input mode, and if so, the process proceeds to step S10 and has not been switched. If so, the process proceeds to step S3.

In step S3, the control unit 8 determines whether or not the action pattern is input to the voice input unit 2 as a voice. If input, the process proceeds to step S4. The process proceeds to S2.
In step S4, the voice input from the voice input unit 2 is recognized as a behavior pattern by the voice recognition unit 3 (behavior pattern input step, behavior pattern input step).
In step S5, the control unit 8 stores the action pattern recognized by the voice recognition unit 3 and the input time of the action pattern in the storage unit 5 as a set, and proceeds to step S2 (storage process, storage). Step).

  In step S10, the control unit 8 determines whether or not the instruction from the changeover switch 4 has been switched to an instruction to switch to the action pattern input mode. If the instruction has been switched, the process proceeds to step S2 and has been switched. If not, the process proceeds to step S11.

  In step S11, the control unit 8 determines whether or not a keyword related to the action pattern has been input as a voice to the voice input unit 2, and if it has been input (speech input step, voice input step), step S12. If it is not input, the process proceeds to step S10.

In step S12, the control unit 8 recognizes the voice input from the voice input unit 2 as a keyword by the voice recognition unit 3 (voice recognition step, voice recognition step).
In step S13, the control unit 8 determines whether or not an action pattern corresponding to the keyword recognized by the voice recognition unit 3 exists in the plurality of action patterns in the storage unit 5. If not, the process moves to step S20.

In step S14, the control unit 8 selects an action pattern corresponding to the keyword in the storage unit 5 (selection process, selection step).
In step S15, the control unit 8 controls the speech synthesizer 6 based on the selected action pattern to synthesize it as speech (speech synthesis step, speech synthesis step), and proceeds to step S16.

  In step S20, the control unit 8 synthesizes that there is no action pattern corresponding to the keyword as a voice by controlling the voice synthesis unit 6, and proceeds to step S16.

  In step S16, the control unit 8 outputs the synthesized voice from the voice output unit 7 (voice output step, voice output step), and proceeds to step S1.

  As described above, according to the present embodiment, when a keyword related to any one of the action pattern and the input time is input to the voice input unit 2 by voice, the voice recognition unit 3 analyzes the voice, and the keyword Recognize Then, the control unit 8 selects an action pattern corresponding to the keyword recognized by the voice recognition unit 3 from a plurality of action patterns in the storage unit 5, and uses the voice synthesis unit 6 based on the selected action pattern. Control and synthesize as speech. Thereby, since the action pattern corresponding to the keyword is output from the voice output unit 7, even if the user forgets the daily action, it is possible to give an opportunity to remind the user of the action content. Mental burden can be reduced.

Of course, the present invention is not limited to the above-described embodiment and can be modified as appropriate.
For example, in the present embodiment, the case where the voice input unit 2 and the voice recognition unit 3 also serve as the behavior pattern input unit of the present invention has been described. However, the voice input unit and the voice recognition unit dedicated to the behavior pattern input unit are provided. May be. In addition, as long as an action pattern is input, an input format other than voice input may be used. For example, depending on the type of action pattern, an object may be operated in association with action execution. It is also possible to arrange a sensor for the object to be operated, detect the presence or absence of the operation by the sensor, and cause the control unit 8 to output the recognition result as an action pattern. Specifically, in the case of inputting a behavior pattern called medication, a sensor is arranged in the medication box, the opening and closing of the medication box is detected by this sensor, and the detection result is output to the control unit 8, which is called medication. The behavior pattern can be recognized.

  In addition to this, for example, depending on the type of action pattern, the user may move somewhere during action execution. In this case, it is possible to cause the control unit 8 to recognize the movement destination by allowing the user to carry the GPS terminal and allowing the GPS terminal and the control unit 8 to communicate with each other. Specifically, in the case of inputting a behavior pattern called hospital visit, the control unit 8 can be made to recognize the behavior pattern called hospital visit from the position information of the hospital inputted from the GPS terminal.

Further, in the present embodiment, when a keyword is input, the case where an action pattern corresponding to the keyword is output as a sound under the control of the control unit 8 has been described as an example. In this case, the speech synthesis unit 6 may be controlled based on the action pattern in the storage unit 5 that is paired with the input time, and synthesized as speech. As described above, when the action pattern input time is reached, the speech synthesizer 6 synthesizes the speech based on the action pattern in the storage unit 5 paired with the input time. Therefore, the action pattern is input when the action pattern is input. Is output from the audio output unit 7. Therefore, even if there is no keyword input from the user, the behavior pattern can be notified to the user.
In the present embodiment, it can be used for schedule management, TV / radio program viewing management, and the like in addition to the daily activities that have become a habit as described above.

It is a block diagram showing schematic structure of the life assistance apparatus which concerns on this embodiment. It is explanatory drawing showing an example of the list of the action patterns memorize | stored in the memory | storage part of the life assistance apparatus of FIG. It is explanatory drawing showing an example of the list of the action patterns memorize | stored in the memory | storage part of the life assistance apparatus of FIG. It is a flowchart of the program performed with the life assistance apparatus of FIG.

Explanation of symbols

1 Life assistance device 2 Voice input unit (behavior pattern input unit)
3 Voice recognition unit (behavior pattern input unit)
4 switch 5 storage unit 6 voice synthesis unit 7 voice output unit 8 control unit 31 database

Claims (4)

An action pattern input unit for inputting an action pattern;
A storage unit that stores a plurality of sets of the action pattern and the input time of the action pattern that are input to the action pattern input unit;
A voice input unit in which a keyword related to any one of the behavior pattern and the input time is input by voice;
A voice recognition unit that analyzes voice input from the voice input unit and recognizes the keyword;
A speech synthesizer that synthesizes the behavior pattern in the storage unit as speech;
A voice output unit for outputting the voice synthesized by the voice synthesis unit;
The behavior pattern corresponding to the keyword recognized by the speech recognition unit is selected from the plurality of behavior patterns in the storage unit, the speech synthesis unit is controlled based on the selected behavior pattern, and the speech A life assistance device comprising: a control unit that synthesizes as: The life assistance device according to claim 1,
The control unit synthesizes the voice as a voice by controlling the voice synthesis unit based on the behavior pattern in the storage unit paired with the input time at the input time of the behavior pattern. Auxiliary device. An action pattern input process in which an action pattern is input;
A storage step of storing a plurality of the behavior patterns input in the behavior pattern input step;
A voice input step in which keywords related to the behavior pattern are input by voice;
Analyzing the voice input in the voice input step and recognizing the keyword;
A selection step of selecting the behavior pattern corresponding to the keyword recognized in the voice recognition step from the plurality of behavior patterns stored in the storage step;
A speech synthesis step of synthesizing the behavior pattern selected in the selection step as speech;
And a voice output step of outputting the voice synthesized in the voice synthesis step. An action pattern input step in which an action pattern is input;
A storage step for storing a plurality of the behavior patterns input in the behavior pattern input step;
A voice input step in which keywords related to the behavior pattern are input by voice;
A speech recognition step of analyzing the speech input in the speech input step and recognizing the keyword;
A selection step of selecting the behavior pattern corresponding to the keyword recognized in the voice recognition step from the plurality of behavior patterns stored in the storage step;
A speech synthesis step of synthesizing the behavior pattern selected in the selection step as speech;
A voice output step of outputting the voice synthesized in the voice synthesis step.

Images