JP2002351305A - Robot for language training - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a robot for language training which creates environment for life overseas even at home and enables a trainee to master a foreign language while enjoying it very naturally. SOLUTION: The robot for language training is equipped with a translating function of translating a spoken mother language into a foreign language and voicing it, a foreign language conversation function of replying in the foreign language when spoken to in the mother language or foreign language, a learning function of voicing a question in the mother language or foreign language to a user and deciding whether an answer from the user is correct, a pronunciation evaluating function of voicing a word in the foreign language to the user and evaluating the pronunciation of a word that the user replies, and a free function of moving about while voicing a story or singing a song if there is no user's answer for a specific time; and one of them is selected upon occasion.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] The present invention relates to a home robot.

[0002]

2. Description of the Related Art Recently, a so-called pet-type robot that uses an ultrasonic radar to move using obstacles and obstacles using wheels and feet, and has audio-visual and tactile functions and learning functions using a CCD camera, a microphone, and a touch sensor. Are commercially available and are commercially available. These pet-type robots are equipped with functions such as being able to recognize a language and have a conversational function, and changing the way of reaction when facing, approaching, striking, and so on.

[0003] While such pet-type robots having a strong element of play have become widespread, the direction of the robots that can be useful in the actual home life has not yet been determined, and it is merely a television. It is only a substitute for a remote control.

[0004]

SUMMARY OF THE INVENTION The present invention has been devised in view of the above background, and aims to create an environment of foreign living while staying at home and to learn while enjoying a foreign language naturally. We propose a teacher robot for language training with a clear and clear role.

At present, the functions of the pet-type robot as described above include a visual function using a CCD camera, a tactile / obstacle recognition function using a touch sensor or ultrasonic waves, a conversation function using a microphone and a speaker, LED and head movement, or L
The emotional expression function by the face display using CD has been developed. Furthermore, with the evolution of the computer, the reaction of the robot differs depending on the way of contact with a person, or a specific person is recognized and escapes or approaches by that contact. It is also possible to have a program that shows a complicated reaction such as squatting. Also, with the current technology, a robot that has an Internet connection function and can communicate with a remote place is also possible. In the present invention, these functions are combined with a voice recognition and translation program to provide a language training robot that can enjoy a conversation in a foreign language or become a teacher of a foreign language according to the reaction of a person.

That is, a first feature of the language training robot according to the present invention is that it has a translation function of translating a word spoken in its native language into a foreign language and uttering it. In this robot, for example, first, the robot's computer analyzes the spoken language of the user by a voice recognition function, and the characteristics of the user are stored. After that, when the user speaks in his or her native language, the robot performs translation for each phrase of the word and utters it with synthetic speech or the like. This will allow you to naturally practice foreign language conversations while at home.

As an application of this translation function, if a language spoken in a native language is translated into a foreign language and the translated data can be transmitted to another robot connected to the Internet, You can also enjoy talking with an interpreter. In this case, if the robot has a camera, it can be used like an international videophone.

[0008] A second feature of the language training robot of the present invention is a foreign language conversation function of responding in a foreign language when a spoken language is spoken in a native language or a foreign language. For example, a robot first registers voice data of a short word such as a word spoken by a user irrespective of a native language or a foreign language as a packet, and thereafter, when the user speaks, the word is registered in the registered data as 7 to 7. If it matches 8 percentages, it is judged to be applicable. Then, for the recognized word, the robot reads out the response data for an appropriate answer that has been input in advance, and the robot replies in a foreign language. At that time, several kinds of appropriate response data for one word are prepared, and the situation (date and time, ambient brightness,
If the selection is made according to the loudness of the user's voice or a biorhythm program built into the robot), it is possible to obtain a fun and uninteresting feeling such as strange conversation.

A third feature of the language training robot of the present invention is to test how many foreign words and predicates the user knows, and ask the user questions in his or her native language. This is a study function for determining whether the answer returned from the user is correct. In this case, it is preferable to provide a pet-type function such that the robot gives a cheerful gesture or sings a cheerful song when a predetermined passing score is reached.

A fourth feature of the language training robot of the present invention is a pronunciation evaluation function for uttering a foreign language to a user and imitating the foreign language to evaluate the pronunciation of a word returned by the user. This is a function that evaluates whether the pronunciation of the user can be understood by foreigners, the position of accents, etc. In this case as well, it will give a score according to the evaluation, and if it exceeds the passing score, it will be similar to the above You should take action.

[0011] In the robot having the above functions, if there is no response from the user for a predetermined time or more, the robot is shifted to a free function that moves around while uttering a foreign language story or song. It is preferable because it can reproduce the environment as if you were in a foreign country. This free feature,
If the robot stores multiple story and song data, and selects one of them according to the situation (date and time, surrounding brightness, built-in biorhythm program etc.), the robot's behavior Change is good.

Each of the above functions can be incorporated in one robot so that any one of the functions can be selected according to a user's instruction or the like. In addition, the above language training robots can update the built-in foreign language (and native language) data by inputting the updated data via a personal computer or media, or by downloading via an Internet connection. It is good to be able to input data.

Furthermore, when a native language that the robot cannot recognize during the response with the user comes out, a connection is made to the station where the human teacher waits by using the Internet connection function provided to the robot, and translated into the teacher. It is more preferable that the received voice data be received so as to complement the robot data. By augmenting the data in this way, not only the progress of the robot, but also the level of the user will be improved accordingly. Making the robots evolved in this way contested at competitions and made available for auctions will increase motivation.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows an example of a pet-type language training robot. As shown in the figure, the robot 1 of the present example is of an autonomous walking type in the form of a duck or a penguin as a whole, and its head 2 has a CCD camera serving as an eye and a speaker serving as a mouth (not shown). And a tactile sensor 3 that senses when the user strokes or hits his head, an ear-shaped stereo microphone 4 that listens to the user's voice, and facial expressions (eyes, nose, and mouth). A display screen 5 of an LCD (liquid crystal screen) for displaying is provided.

On the other hand, the body 10 has a hand / neck drive unit 11 for flapping the wings and rotating and moving the neck up and down, a walking drive unit 12 for moving the feet, and an ultrasonic sensor 13 for detecting an obstacle. , A communication modem 14 for connecting to the Internet, and a computer 15 comprising a CPU and a memory (motherboard) which are the heart. The computer 15 includes a camera on the head 2, a microphone 4,
It processes information from the tactile sensor 3 and the ultrasonic sensor 13 on the body 10, and controls the speaker on the head 2, the display screen 5, the hand / neck drive unit 11, the walking drive unit 12, and the communication modem 14 on the body 10.

As an example, such a pet-type robot 1 acts according to a flowchart of language training as shown in FIGS.

First, the robot 1 is in the standby mode 100 in the energy-saving state if no operation is performed for a while. If no action occurs for a considerable time in this state, the robot 1 shifts to the sleep mode 101 in the further energy-saving state. I have. When the user calls in the native language or a foreign language while in the standby mode 100 or the sleep mode 101, the robot 1 proceeds to step 102 of the call response and returns the direction in the called direction. Then, in the next step 103, the robot 1 waits for a predetermined time, for example, three minutes, whether or not there is talk from the user.

When the user speaks during step 103, the process proceeds to step 104, where the robot 1 checks by itself whether or not the translation mode is prioritized.
The priority setting of the translation mode can be appropriately set by a setting at the time of shipment, a setting by a switch (not shown) provided on the robot 1, a setting by connecting the robot 1 to a personal computer, and the like. As a result of step 104, if the translation mode is set, the robot 1 translates the spoken word into a foreign language and utters it, as in step 105. That is, the computer 15 analyzes the voice of the user input from the stereo microphone 4 according to the voice recognition program, and when the words are recognized, translates them into a foreign language according to the translation program and outputs them from the speaker. At this time, the expression on the display screen 5 is also controlled so as to speak.

After the translation is uttered, it is confirmed in a waiting step 106 whether or not the user can speak further for a predetermined time, for example, about 3 minutes. If spoken, the process returns to step 105 to execute the translation. If there is no speech, or if the user leaves the robot 1, the process proceeds to the free mode step 107. In the free mode, the robot 1 roams around
Sing a foreign language song recorded on the internal hard disk, etc., or tell a story in a foreign language. The computer 15 determines which song or story is to be selected from among a large number of recorded songs in accordance with the date and time, the temperature and brightness of the day, or the biorhythm program. The recorded data of songs and stories can be updated by replacing a PC card or ROM or by connecting to a personal computer.

While wandering in the free mode,
The robot 1 constantly checks whether or not it can be done in step 108. If the user is scolded as "noisy" or the like, the process proceeds to step 109 and enters the sleep mode. In other words, "sleep down." Alternatively, it is also possible to perform control to shift to the standby mode after a certain period of time in the free mode. After entering the standby mode and the sleep mode, the state is kept until there is a call (step 102).

On the other hand, if the user does not speak in the predetermined time in step 103, the process proceeds to step 110, in which the robot 1 speaks to the user. For example, the computer 15 selects an appropriate word from the recorded words according to the biorhythm program of the day, and outputs it from a speaker in a foreign language or a native language. In response to this, it is checked in step 111 whether or not there is a response from the user.
Return to If there is a reply, the process proceeds to step 112 to check whether the user's words can be determined. This determination step 112 also proceeds when the translation mode is not the translation mode in the translation mode confirmation step 104. The computer 15 can recognize the user's words if the words input from the microphone 4 and the data of the recorded words match by 70% or more, but otherwise notifies the user that the words cannot be determined.

If the words uttered by the user can be determined, the process proceeds to step 113 to check whether the words require the study mode. This is performed, for example, by confirming whether or not the user has said "Let's study". Alternatively, it can be set by a predetermined switch provided on the robot 1. If the request is not for the study mode, the process proceeds to step 114, where the computer 15 selects an appropriate answer in the foreign language from the recorded words according to the biorhythm program of the day, and responds to the user from the speaker in the foreign language. Time and response from the user.

If the study mode is requested in step 113, the robot 1 proceeds to step 115 in study mode (FIG. 3). In the study mode, the robot 1 first utters the language in its native language as a question in step 116. Then, in response to this, the foreign language translation returned from the user through the microphone 4 is determined, and in step 117, it is determined whether the translation is correct. This is performed by comparing the words answered by the user with the recorded words and determining what percentage or more matches. If correct, step 118
Then, the robot 1 displays a joyful expression on the display screen 5, flaps the wings, etc., and notifies the speaker of the correct answer. If the answer is incorrect, the process proceeds to step 119, where a sad expression is displayed on the display screen 5 to inform the user of the incorrect answer, and an answer from the user is received again at step 120. At step 121, it is determined whether or not the re-answer is a correct answer. If the answer is correct, the process proceeds to step 118 to notify the correct answer. On the other hand, if the answer from the user is incorrect again, or if the user gives up and answers "I do not know", the process proceeds to step 122 and a correct foreign language translation is uttered from the speaker.

Steps 118 and 122
It is also possible to incorporate a step to score points, such as adding points for correct answers and deducting points for incorrect answers, so that when the score exceeds the passing score, the robot can sing a fun song, The control may be such as to make a gesture.

The pronunciation evaluation mode can be executed in the same manner as in the flow of the study function shown in FIG. That is,
At step 115, the mode becomes the pronunciation evaluation mode.
At 6, the robot 1 utters a predetermined foreign language. In response to this, the user pronounces and returns the same word in ohm return, and evaluates in step 117 whether the answered word is correct or not by comparing waveforms. If it is correct, step 118 reports that it is correct; if it is incorrect, step 119 reports that it is incorrect. In the case of inaccuracy, a re-challenge is urged in step 120, and the accuracy is evaluated again in step 121. If correct, step 118
If it is incorrect, the correct pronunciation is output in step 122. At this time, similarly to the above, it is possible to give a score as in the case of karaoke.

Returning to FIG. 1, if the words cannot be determined in the determination of step 112, the process proceeds to the data updating process of FIG. In this process, first, in step 123, the robot 1 notifies that the user's language cannot be understood in a foreign language (or his native language), and in response to this, the user issues an inquiry command such as "try to inquire". It is confirmed in step 124 whether or not the above is true. If there is no inquiry command, the process proceeds to step 125, and it is confirmed whether there is a new data input from the user. If there is an input, the recorded word database is updated in step 126, and the newly input word is uttered to make the user confirm. This input is executed by connecting the robot 1 to a personal computer or via a medium such as a CD-ROM or FD. For this purpose, the robot 1 has a built-in connection interface with a personal computer (not shown) and a drive for media (not shown).

On the other hand, if there is an inquiry command in step 124, in step 127, the robot 1 connects to the Internet by the communication modem 14 and contacts a station where a native language native teacher is waiting. The communication modem 14 for this purpose is suitably of a wireless type. Step 1 after connecting to the station
At 28, the data of the language that was not understood is transmitted and translated by the teacher of the station. When the translated word is returned from the teacher, the recorded word database is updated and input in step 129 as new data, and the input new data is input in step 126.
Speak for user confirmation.

In the course of steps 127 to 129, it is also possible to connect to a Web site via the Internet, search the database thereof, and obtain the relevant data.

Since the robot 1 of the present embodiment has the communication modem 14, it can be connected to another robot having the same function via the Internet to exchange data. In this case, it is possible to project the face of the other party photographed by the camera on each other's display screens 5, and it can be used like a videophone. According to this, for example, by connecting the robot 1 of the Japanese home and the robot 1 of the American home, the robot 1 of the Japanese home becomes Japanese →
While performing English translation, the robot 1 at home in the United States performs English-to-Japanese translation so that it is possible to make an international call without having to speak each other's language.

[0030]

According to the present invention, there is provided a robot having a clear role as a language training teacher, not a vague role such as a pet type. If this language training robot is at home, you can create an environment as if you are living in a foreign country, or an environment where a foreign language teacher is always at hand. But you can naturally practice foreign languages.

[Brief description of the drawings]

FIG. 1 is an external view of a robot according to the present invention.

FIG. 2 is a flowchart showing an operation example of the robot according to the present invention.

FIG. 3 is a flowchart following FIG. 2;

FIG. 4 is a flowchart following FIG. 2;

Continuation of the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat II (reference) G09B 19/00 G09B 19/00 G // A63H 3/33 A63H 3/33 C F term (reference) 2C028 AA03 BA03 BB06 BC01 BC02 BD03 CA06 CA12 CA13 2C150 BA11 CA01 CA02 CA04 DF03 DF06 DF33 ED42 ED47 ED52 EF29 3C007 AS34 BS27 CS08 JS03 KS31 KS39 KT01 MT11 MT15 WB19 WC03 WC07 WC14 5B091 AA01 CB12 CB32 CD15

Claims (13)

[Claims] 1. A language training robot having a translation function of translating a word spoken in a native language into a foreign language and uttering it. 2. The language training robot according to claim 1, wherein the language spoken in the native language is translated into a foreign language, and the translated data is transmitted to another robot connected to the Internet. 3. When spoken in a native language or a foreign language,
A language training robot characterized by having a foreign language conversation function that responds in a foreign language. 4. The language training robot according to claim 3, wherein several types of appropriate response data for one word are prepared, and one of them is selected according to the situation. 5. A language training robot having a study function for asking a user a question in a native language or a foreign language and determining whether an answer returned from the user is correct in response to the question. . 6. A language training robot having a pronunciation evaluation function of uttering a foreign language to a user and evaluating the pronunciation of a word returned by the user by imitating the foreign language. 7. The language training program according to any one of claims 1 to 6, wherein if there is no response from the user for a predetermined time or more, the function shifts to a free function of moving around while uttering a story or song in a foreign language. robot. 8. A plurality of story and song data are stored,
8. The language training robot according to claim 7, wherein one of the robots is selected according to a situation. 9. When there is no response with the user for a predetermined time or more, one of data of a plurality of stored foreign language stories and songs is selected according to the situation, and the selected story or song is uttered. A robot for language training that has a free function that moves around while moving. 10. A translation function for translating a word spoken in a native language into a foreign language and uttering it, a foreign language conversation function for responding in a foreign language when spoken in a native language or a foreign language, A study function that asks a question in the native language or a foreign language and determines whether or not the answer returned from the user in response is correct, and pronounces a foreign language to the user and pronounces the word returned by the user by imitating this. Or a free function that moves around while uttering a foreign language story or song if there is no response from the user for a predetermined time or more, depending on the case. A language training robot characterized in that one of the following is selected. 11. The language training robot according to claim 1, wherein built-in data can be updated by inputting update data via a personal computer or a medium. 12. The language training robot according to claim 1, wherein the built-in data can be updated by inputting update data by downloading via an Internet connection. 13. When an unrecognizable native language comes out during a response with the user, voice data obtained by connecting to a station where a human teacher is on standby via the Internet and translating the unrecognizable native language to a teacher of the station. The language training robot according to any one of claims 1 to 12, wherein the robot is capable of complementing data by receiving the data.