KR100405061B1 - Apparatus for training language and Method for analyzing language thereof - Google Patents

Abstract

The present invention stores voice information of native language users and learners, analyzes pronunciation by comparing frequency characteristics, and outputs a result of analyzing stresses and rhythms by comparing envelopes of voice output waveforms. By outputting a mouth shape, the present invention relates to a language learning apparatus and a language analysis method thereof having improved learning effects.

The present invention records the learner's voice information and mouth shape image data in the user data memory using a microphone and a camera, and pronounces the voice information of the native language user recorded in the sample data memory and the learner's voice information recorded in the user data memory. Analysis by dividing into intonation, intonation, rhythm, characterized in that to display the analyzed results in the form of a chart.

Therefore, the present invention checks the objective analysis data on pronunciation, intonation, and rhythm of the words or sentences learned by the learner, and introduces the above-described analysis data into the repetitive learning, so that the pronunciation of the native language user can be learned close to the language. There is an effect that can enhance the learning effect.

Description

Apparatus for training language and Method for analyzing language etc}

The present invention relates to a language learning apparatus and a method for analyzing a language thereof, and more particularly, to store voice information of a native language user and a learner, analyze a pronunciation by comparing frequency characteristics, and encode an envelope of a voice output waveform. The present invention relates to a language learning apparatus and a method for analyzing the language, which outputs a result of analyzing stress and rhythm, and outputs a recorded voice and a recorded mouth shape.

Language learning methods currently commonly used include first, interpersonal learning, second, sentence repetitive learning using audio tape audio only, and third, audiovisual learning using video tape, and fourth, online or offline using books or PC. Learning method using is common.

Despite the emergence of the various learning methods described above, there are vulnerabilities for each method and reference is provided for each method, but a method for accurately analyzing learners' pronunciation, stress, and mouth shape is not provided.

Therefore, the learner could not accurately recognize how different the pronunciation and stress from the original sound and where it was wrong. Therefore, the result of learning is inefficiently progressed, and considerable effort and investment are required to learn the language.

Accordingly, an object of the present invention is to solve such problem (s), and an object of the present invention is to analyze pronunciation of a word or sentence learned by a learner according to frequency characteristics, and stress and rhythm into a voice waveform having an envelope. The present invention provides a language learning apparatus for objectively analyzing pronunciation, stress, rhythm, and mouth shape by analyzing and displaying the result.

Another object of the present invention is to provide a language learning apparatus that enables a learner to selectively reproduce only a part of a word or sentence learned by the learner, thereby doubling the learning effect on a specific word or selection area.

In addition, another object of the present invention is to provide a language learning apparatus that allows the learner to store and repeat the content of the learner to repeat the correction, and to store the learning result to manage the progress of the learner.

Further, another object of the present invention is to provide a language analysis method of a language learning apparatus for achieving each of the above objects.

1 is a schematic block diagram of a language learning apparatus according to the present invention,

2 is an operation flowchart showing a process of performing a language analysis method according to the present invention,

3 is a waveform diagram illustrating a voice waveform analysis method according to the present invention;

4 is a diagram showing a pronunciation analysis chart according to the present invention,

5 is a display state diagram displayed on a display unit according to a waveform analysis result;

6 is an exemplary view of applying the language learning apparatus according to the present invention as an electronic-English-Korean dictionary.

7 is an exemplary view of applying the language learning apparatus according to the present invention as an electronic conversation notebook.

<Description of the code | symbol about the principal part of drawing>

101: microphone 102: analog / digital conversion unit

103: camera 104: video signal input unit

105: interface unit 106: operation unit

107: user data memory 108: sample data memory

109: digital / analog converter 110: speaker

111: analysis unit 112: video output unit

113: display unit 114: control unit

A feature of the present invention for achieving the above object (s) is a voice input means for converting the analog voice signal of the learner input from the microphone into digital voice information; A user data memory for storing voice information of the learner provided from the voice input means; A sample data memory for storing sample data including voice information of a native language user; Voice output means for reproducing voice information of the native language user detected from the sample data memory, converting the voice information into an analog signal, and outputting the same to a speaker; An analysis unit for analyzing the voice information of the native language user reproduced by the voice output means and the frequency characteristics of the pronunciation of the voice information of the learner provided from the voice input means, and generating a pronunciation analysis chart according to the analyzed result; A display unit which displays the pronunciation analysis chart generated by the analysis unit in a predetermined area; Voice information of the native language user stored in the sample data memory in the playback mode is output to the voice output means, voice information of the learner inputted from the voice input means in the recording mode is stored in the user data memory, and stored in the sample data memory in the analysis mode. And control means for outputting voice information of the native language user and voice information of the learner stored in the user data memory to the analysis unit.

The apparatus may further include an interface unit that receives online sample data including voice information of a native language user. The control unit preferably stores the sample data received in the interface unit in the sample data memory.

The analyzer may detect the first envelope of the learner's voice information waveform; Detect a second envelope of a voice information waveform of a native language user; It is preferable to obtain the difference between the first envelope and the second envelope and output the first analysis chart for stress.

In addition, it is preferable that the analysis unit compares the start point of the word of the first envelope with the start point of the word of the second envelope, obtains the time difference, and outputs the time difference as a second analysis chart for the rhythm.

In addition, recording means for adjusting the resolution of the mouth-shaped image of the learner recorded by the camera and performing compression; A video output means for converting the learner's mouth image corresponding to the learner's voice information and the native language user's mouth image corresponding to the voice information of the native language user into a display signal and outputting the signal to the display unit; The controller may store the learner's mouth image output from the recording unit in the recording mode in the user data memory, and output the learner's mouth image and the native language user's mouth image to the video output unit in the analysis mode.

The apparatus may further include an operation unit having a moving key for searching for a sentence to be analyzed among sample data stored in the sample data memory, and a selection key for selecting a corresponding sentence for reproducing and analyzing in the retrieved sentence unit; The controller may be configured to reproduce and analyze a sentence selected by the learner among sample data stored in the sample data memory in response to an operation signal from the operation unit.

The selection key in the operation unit may further include a function of selecting a corresponding word to reproduce and analyze at least one word selected by the learner among a plurality of words included in the searched sentence; The controller may be configured to reproduce and analyze a word in a sentence selected by the learner among sample data stored in the sample data memory in response to an operation signal from the operation unit.

In addition, the selection key in the operation unit further includes a function of selecting an arbitrary region for the learner to select and reproduce and analyze an arbitrary region among the searched sentences; The controller may be configured to reproduce and analyze a region selected by the learner among sample data stored in the sample data memory in response to an operation signal from the operation unit.

On the other hand, another feature of the present invention, the voice information of the learner stored in the user data memory is detected, and the frequency characteristics of the voice information of the learner so as to represent the waveform of the detected voice information of the learner in the frequency domain in the time domain and; Detecting voice information of the native language user stored in the sample data memory, and analyzing frequency characteristics of the voice information of the native language user such that the waveform of the detected native language user's voice information is represented in the frequency domain in the time domain; The frequency of the voice information of the learner and the voice information of the native language user is generated by generating a pronunciation analysis chart showing the output waveform for each band.

Wherein a first envelope of an output waveform for voice information of the learner is detected; Detect a second envelope of an output waveform with respect to voice information of the native language user; Detect a difference between the first envelope and the second envelope; A stress analysis chart showing the difference between the detected first envelope and the second envelope may be generated and output.

Further detecting a first starting point of the next word following the first word in the output waveform for the voice information of the learner; Detecting a second starting point of the next word following the first word in the output waveform for the voice information of the native language user; Detecting a time difference between the first starting point and the second starting point; A rhythm analysis chart indicating a time difference between the detected first start point and the second start point may be generated and output.

Detecting a first non-signal interval between the next word following the first word in the output waveform of the learner's voice information; Detecting a second non-signal interval between the next word following the first word in the output waveform for the voice information of the native language user; Detecting a time difference between the first no-signal section and the second no-signal section; A rhythm analysis chart indicating a time difference between the detected first no-signal section and the second no-signal section may be generated and output.

Hereinafter, exemplary embodiment (s) of the present invention will be described in detail with reference to the accompanying drawings. First, in adding reference numerals to the elements of each drawing, it should be noted that the same elements are denoted by the same reference numerals as much as possible even if they are displayed on different drawings. In addition, in the following description there are shown a number of specific details, such as components of the specific circuit, which are provided only to help a more general understanding of the present invention that the present invention may be practiced without these specific details. It is self-evident to those of ordinary knowledge in Esau. In describing the present invention, when it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

Figure 1 is a schematic block diagram of a language learning apparatus according to the present invention, Figure 2 is an operational flow diagram showing the performance of the language analysis method according to the present invention, Figure 3 voice waveforms according to the present invention A waveform diagram illustrating an analysis method is shown, FIG. 4 is a diagram illustrating a pronunciation analysis chart according to the present invention, and FIG. 5 is a display state diagram displayed on a display unit according to a waveform analysis result. 6 illustrates an example of applying the language learning apparatus according to the present invention as an electronic-English-English dictionary, and FIG. 7 illustrates an example of applying the language learning apparatus according to the present invention as an electronic conversation notebook.

Referring to Figure 1, the configuration of the present invention will be described.

The microphone 101 collects the learner's voice and converts it into an electrical signal in an analog form.

The analog-digital converter 102 converts the analog voice signal input from the microphone 101 into a digital form and outputs the digital signal.

The camera 103 converts image information obtained by photographing a learner as a subject into a video signal. In particular, in the present invention, the camera 103 photographs the learner's mouth shape so that the learner can visually check the shape of his or her mouth.

The video signal input unit 104 adjusts the resolution and data compression of image information about the shape of the learner's mouth input from the camera 103.

The interface unit 105 is a connection unit for connecting to a computer or the Internet, and serves to interface sample data transmitted online.

The operation unit 106 is arranged with a plurality of function keys for controlling the operation of the language learning apparatus. In particular, among the sample data (voice information of the native language user and image data representing the mouth shape) stored in the sample data memory 108 is arranged. A movement key for searching for an arbitrary sentence, and a selection key for selecting a predetermined region of a whole word or sentence included in the sentence or a sentence indicated by the movement key, and includes a sentence, a word, and a predetermined region selected by the selection key. A playback key for reproducing any one of them, a recording key for recording image information representing the voice information and mouth shape of the learner, and an analysis key for analyzing the sample data and the user data of the learner.

The user data memory 107 stores voice information, image data and learning history data recorded by a learner who is a user.

The sample data memory 108 stores voice information and image data of a native language user that is recorded and input directly by the native language user or downloaded online through a computer or an Internet terminal accessible to the interface unit 105.

The digital / analog converting unit 109 converts the voice information in which the playback function is selected by the learner among the digital voice information stored in the user data memory 107 and the sample data memory 108 into an analog voice signal and converts the voice information into a speaker ( Output to 110).

The analysis unit 111 may perform a frequency conversion algorithm (eg, Fourier transform algorithm) for pronunciation analysis on voice information selected by a learner among voice information stored in the user data memory 107 and the sample data memory 108. ), Frequency characteristics are detected, envelopes for stress analysis and rhythm analysis are detected, and the detected results are converted into a chart in a predetermined format and output.

The video output unit 112 selectively detects and restores (decompresses) image data of which the playback function is selected by the learner among the image data of the mouth shape stored in the user data memory 107 and the sample data memory 108. Then output.

The display unit 113 outputs the character information of the voice information selected by the learner using the operator 106 among the voice information of the learner stored in the user data memory 107, and the analysis result and the video analyzed by the analyzer 111. The image data output from the unit 112 is arranged and displayed in a predetermined area.

The controller 114 generally controls the operation state of the language learning apparatus. In particular, the controller 114 stores the digital voice information converted by the analog / digital converter 102 in the user data memory 107 by specifying an address and stores the digital voice information in the digital form by the video signal input unit 104. The compressed mouth image data is stored in the user data memory 107. At this time, the controller 114 stores the image information of the mouth shape in the user data memory 107 by linking and storing the address of the voice information and the address of the mouth shape image data so that the voice information is selected when the voice information is selected in the reproduction process. The mouth image data can be detected at the same time.

In addition, the controller 114 stores the sample data downloaded through the interface unit 105 in the sample data memory 108 by specifying an address, and selects a word, sentence, or region selected by the learner according to an operation signal of the operation unit 106. Voice information and image data stored in the user data memory 107 and the sample data memory 108 are selectively detected, and the digital / analog converter 109, the analyzer 111, and the video output 112 are respectively detected. Output

The language learning apparatus having the configuration as described above may be manufactured by an individual terminal such as an electronic organizer or an electronic conversation dictionary, and may be applied to a computer and implemented on the Internet or in an individual desktop terminal.

An operation of the present invention having such a configuration will be described below with reference to FIGS. 2 to 7.

The operation of the language analyzing apparatus according to the present invention is divided into a selection mode, a reproduction mode, a recording mode, and an analysis mode, and the analysis mode is classified into a stress analysis mode, a rhythm analysis mode, and a pronunciation analysis mode.

First, sample data (voice information and mouth-shaped image data) of a native language user are downloaded from a computer or an Internet terminal connected through the interface unit 105. For this purpose, the interface unit 105 is connected to a computer through a cable or It is configured to satisfy the protocol when connected to the Internet through an Internet terminal. The controller 114 assigns an address to the sample data memory 108 to store voice information and image data of the native language user downloaded through the interface unit 105 and stores the address in the designated area of the sample data memory 108. Therefore, the sample data downloaded through the interface unit 105 is stored in the sample data memory 108.

On the other hand, when the selection mode and the reproduction mode are described, the learner presses a specific key in the operation unit 106 to select the sentence display function, and the control unit 114 responds to the operation signal of the operation unit 106 by the sample data memory 108. Character information corresponding to the plurality of sample data stored in the display unit 113 is displayed. Accordingly, when the learner selects a word or any region included in any sentence or sentence among the sample data (voice information of the native language user and mouth shape image data) displayed on the display unit 113 using the moving key and the selection key, The controller 114 detects address information corresponding to a sentence, word, or region selected by the learner in response to an operation signal from the operation unit 106, and voice information corresponding to the sentence, word, or region from the sample data memory 108. Detect.

Subsequently, when the learner manipulates the play key of the operation unit 106, the control unit 114 corresponds to the sentence, word or region detected from the sample data memory 108 in response to the play key operation signal of the operation unit 106. After outputting the digital voice information to the digital-to-analog converter 109, the digital-to-analog converter 109 converts the digital voice information transmitted from the controller 114 into an analog voice signal, Output through the speaker 110. Therefore, the learner can visually confirm the sample data corresponding to the sentence, word or region to be learned.

Next, when the recording mode is described, the learner hears a voice about a sentence, a word or an area output to the speaker 110 by the playback mode while pressing the record key, and pronounces it. It is detected by the microphone 101 and the camera 103. That is, when a learner selects a recording key among various function keys of the operation unit 106 and pronounces a voice for a sentence, a word, or an area output to the speaker 110 by the playback mode, the learner's pronunciation is the microphone 101. Is detected as an analog voice signal, and the analog voice signal is converted into digital voice information by the analog / digital converter 102 and then transmitted to the controller 114. The controller 114 designates an address of the user data memory 107 to store voice information output from the analog / digital converter 102 and stores the address in the designated area of the user data memory 107.

In addition, the shape of the learner's mouth is photographed by the camera 103, and the captured image is transmitted to the controller 114 after the resolution is adjusted by the video signal input unit 104 and undergoes a compression process. The control unit 114 designates an address of the user data memory 107 to store the image data input from the video signal input unit 104, and stores the address in the designated area of the user data memory 107. At this time, the mouth-shaped image data is assigned an address associated with the address of the voice information is stored. Therefore, the user data of the learner (image data representing voice information and mouth shape) is stored in the user data memory 107.

Next, when the analysis mode is described, when the learner selects the analysis key, the controller 114 displays the sample data corresponding to the sentence or word selected by the learner in the selection mode or the user data recorded by the learner. The sample data memory 108 and the genetic data memory 107 are respectively detected, and the detected sample data and the user data are output to the analyzing unit 111 and the video output unit 112.

Here, the analysis unit 111 analyzes stress, rhythm, and pronunciation, respectively, which will be described in detail with reference to FIGS. 3 to 5.

First, the analysis unit 111 detects voice information of a native language user from sample data and voice information of a learner from user data (S101), analyzes waveforms of detected voice information (S102), and analyzes the result. In operation S103, the envelope of the waveform of each voice information is detected.

Here, the waveform shown at the top of FIG. 3 is the voice information waveform of the native language user, and the waveform shown at the bottom of FIG. 3 is the voice information waveform of the learner. At this time, the waveform of each voice information is represented by the correlation of the magnitude of the output versus the time.

That is, the analysis unit 111 analyzes the voice information waveform P1 of the native language user among the sample data provided from the control unit 114, calculates a peak value, estimates the envelope E1, and obtains the starting point S1 of the first word. In addition, the analysis unit 111 analyzes the voice output waveform P2 of the learner among the user data provided from the control unit 114, calculates peak values, estimates the envelope E2, and obtains the starting point S2 of the first word.

Thereafter, the analyzer 111 adjusts the parallax Sd existing between the first word start point S1 of the mother tongue user and the first word start point S2 of the learner to '0', and then obtains the difference between the envelope E1 of the native language user and the learner envelope E2. (S104).

As a result, the difference of the above-described envelope becomes the difference between the stresses between the native language user and the learner, and the analyzing unit 111 implements the difference of the envelope into a graph of the stressed analysis chart and displays it at a predetermined position of the display unit 113 (S105). . Accordingly, the stress analysis result is displayed on the display 113 as a chart (see 'A' in FIG. 5), and the learner may visually analyze the difference in stress through the analysis chart.

Next, the analysis unit 111 analyzes the rhythm by obtaining the starting point S3 of the next word of the native language user and the starting point S4 of the next word of the learner, or analyzing the rhythm by obtaining the time difference between the no word and the word (S106). In operation S107, the rhythm analysis chart is generated according to the analyzed result and displayed on the display 113.

In other words, the method of analyzing the rhythm as a starting point is S3 when the native speaker pronounces the first word and starts the next word, and S4 starts the next word after the learner pronounces the first word. As long as the learner can get results with rhythm fast or slow. As a result, the analysis unit 111 implements the difference between the starting points of the following words into a graphed rhythm analysis chart and displays the result at a predetermined position of the display unit 113. Therefore, the display unit 113 displays the result of the rhythm analysis as a chart. It is displayed (see 'B' of FIG. 5), the learner can visually analyze the difference between the rhythm through the analysis chart.

In addition, when explaining the method of analyzing the rhythm in the no-signal interval, after the first word of the native language user is finished, the non-signal period is the d1 before the start of the next word, and after the end of the first word of the learner, The signal interval is d2. Therefore, by the difference between them, the learner can get the result with the rhythm quickly or slowly. As a result, the analysis unit 111 implements the time difference of the no-signal section as a graphed rhythm analysis chart and displays it at a predetermined position of the display unit 113. Accordingly, the display unit 113 displays the rhythm analysis result as a chart. (See 'B' of FIG. 5), the learner may visually analyze the difference in rhythm through the analysis chart.

Finally, referring to FIG. 4, the pronunciation analysis method is described above. The upper part of FIG. 4 is a pronunciation analysis chart of a native language user, and the lower part is a pronunciation analysis chart of a learner.

That is, when the analysis unit 111 performs frequency analysis so that the voice output waveforms P1 and P2 of FIG. 3 can be represented in the frequency domain in the time domain by an algorithm such as a Fourier transform, the output waveform for each frequency band as shown in FIG. 4. It can be obtained (S108). As a result, the analysis unit 111 implements the output waveform for each frequency band obtained by the step S108 as a graphed pronunciation analysis chart and displays it at a predetermined position of the display unit 113 (S109), thus, the display unit ( In 113), the pronunciation analysis results are displayed in a chart, and the learner compares the pronunciation of the native language user and the learner's pronunciation shown in the pronunciation analysis chart, and is generated due to the different frequency domain size corresponding to each voiced consonant or voiced vowel. You can visually check the difference in pronunciation.

In addition, the video output unit 112 detects image data indicating a mouth shape of a native language user among sample data provided from the controller 114 and image data indicating a mouth shape of a learner among user data (S110). After restoring (uncompressing) the data (S111), the restored image is output to the display unit 113, and the display unit 113 outputs the image provided from the video output unit 112 to the various units provided from the analysis unit 111. Displayed in combination with the chart information (S112).

On the other hand, the stress, rhythm, pronunciation analysis chart and the image showing the shape of the mouth as described above may be combined in the display unit 113, as shown in FIG.

Referring to FIG. 5, a voice waveform of a native language user is output on the right side of an area marked 'original sound', and a voice waveform of the learner is output on the right side of an area labeled 'my voice'. The voice waveform has a shape in which the voice output waveform and the envelope waveform overlap each other and are displayed on the time axis.

In addition, a character corresponding to a voice of 'Can you speak English?' Is output on the upper part of the waveform of the native language user, and the leader may be configured to move according to the output state of the voice. In this case, the leader line may be formed over the voice waveform of the native language user and the voice waveform of the learner.

The stress analysis chart A and the rhythm analysis chart B may be output at the lower part of the learner voice waveform, and an image representing a mouth shape corresponding to the right side of the voice waveform of the native language user and the right side of the voice waveform of the learner may be output. Can be.

Further, by configuring the pronunciation analysis chart button on the operation unit 106, when the learner selects the pronunciation analysis chart button, the pronunciation analysis chart as shown in FIG. 4 for the pronunciation may be output on a single screen.

On the other hand, the control unit 114 stores the analyzed results (emphasis analysis chart, rhythm analysis chart and pronunciation analysis chart) in the user data memory 107 (S113). Therefore, the analysis result stored in the user data memory 107 can be utilized for the learner's history management, and the learner can thereby confirm the improvement of pronunciation of the same sentence or word or a predetermined region, thereby enjoying a more effective learning effect. can do.

On the other hand, if you understand the technical idea of the present invention, it is possible to implement a variety of language learning apparatus using the technical idea of the present invention, as an example of the implementation of the electronic notebook and the language learning dictionary as shown in FIG. Can be presented.

In the case of the electronic notebook of FIG. 6, a word, a pronunciation code, an interpretation, and the like may be displayed in letters in a display area, and an analysis chart may be displayed below. And, the microphone may be configured to be connected for input of the pronunciation.

In addition, in the case of the language learning dictionary of FIG. 7, the original text and the translated text may be displayed in the display area, and an analysis chart may be displayed at the bottom thereof. In addition, a microphone for input of a pronunciation may be configured, and a connection terminal for connection with a computer may be configured.

As described above, although the specific embodiment (s) have been described in the detailed description of the present invention, various modifications are possible without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiment (s), but should be defined by the claims below and equivalents thereof.

As a result, according to the language learning apparatus and its language analysis method according to the present invention, the following advantage (s) occurs.

That is, by confirming the objective analysis data on pronunciation, intonation and rhythm of the words or sentences learned by the learner, and introducing the above analysis data into the repetitive learning, the language learning effect can be learned close to the pronunciation of the native language user. Promoted.

In addition, by allowing the learner to selectively reproduce only a part of the word or sentence learned by the learner, the learning effect is doubled for a specific word or selection area.

In addition, since the learner's learning results are history-managed, it is possible to improve the language learning effect.

Claims (12)

Voice input means for converting the learner's analog voice signal input from the microphone into digital voice information and outputting the voice information; A user data memory storing voice information of the learner provided from the voice input means; A sample data memory for storing sample data including voice information of a native language user; Voice output means for reproducing the voice information of the native language user detected from the sample data memory, converting the voice information into an analog signal, and outputting the same to a speaker; An analyzer for analyzing the voice information of the native language user reproduced by the voice output means and the frequency characteristics of the pronunciation of the learner's voice information provided from the voice input means, and generating a pronunciation analysis chart according to the analyzed result ; A display unit configured to display the pronunciation analysis chart generated by the analysis unit in a predetermined area; And Outputting voice information of the native language user stored in the sample data memory to the voice output means in a reproduction mode, storing voice information of the learner input from the voice input means in the user data memory in an analysis mode, and analyzing mode And control means for outputting voice information of the native language user stored in the sample data memory and voice information of the learner stored in the user data memory to the analyzing unit. The method of claim 1, An interface unit for receiving online the sample data including voice information of the native language user; And the control unit stores the sample data received in the interface unit in the sample data memory. The method of claim 1, wherein the analysis unit, Detect a first envelope of the learner's voice information waveform; Detect a second envelope of a voice information waveform of the native language user; Obtaining the difference between the first envelope and the second envelope language learning apparatus, characterized in that for outputting the first analysis chart for stress. The method of claim 3, wherein the analysis unit, And a start point of a word of the first envelope and a start point of a word of the second envelope, and calculate a time difference to output the second analysis chart of the rhythm. The method of claim 1, Recording means for adjusting the resolution of the mouth-shaped image recorded by the camera and performing compression; A video output means for converting the learner's mouth shape image corresponding to the learner's voice information and the native language user's mouth image image corresponding to the voice information of the native language user into a display signal and outputting the signal to the display unit; Includes; The controller stores the learner's mouth shape image output from the recording unit in the recording mode in the user data memory, and outputs the video of the learner's mouth image and the native language user's mouth image in the analysis mode. Language learning apparatus characterized in that the output to the negative. The method of claim 1, An operation unit having a moving key for searching for a sentence to be analyzed among the sample data stored in the sample data memory and a selection key for selecting a corresponding sentence for reproducing and analyzing in the retrieved sentence unit; And the controller is configured to reproduce and analyze a sentence selected by the learner among the sample data stored in the sample data memory in response to an operation signal from the operation unit. The method of claim 6, The selection key in the operation unit further includes a function of selecting a corresponding word to reproduce and analyze at least one word selected by a learner among a plurality of words included in the searched sentence; And the controller is configured to reproduce and analyze a word in a sentence selected by the learner among the sample data stored in the sample data memory in response to an operation signal from the operation unit. The method of claim 6, The selection key in the operation unit further includes a function of selecting an arbitrary region for the learner to select, reproduce and analyze an arbitrary region among the searched sentences; And the controller is configured to reproduce and analyze the region selected by the learner among the sample data stored in the sample data memory in response to an operation signal from the operation unit. Detecting voice information of a learner stored in a user data memory, and analyzing frequency characteristics of the voice information of the learner such that the detected waveform of the voice information of the learner is represented in a frequency domain in a time domain; Detecting voice information of a native language user stored in a sample data memory, and analyzing frequency characteristics of voice information of the native language user such that the detected waveform of the voice information of the native language user is represented in the frequency domain in the time domain; ; And And generating and outputting a pronunciation analysis chart indicating a frequency-specific output waveform of the frequency of the voice information of the learner and the voice information of the native language user. The method of claim 9, Detecting a first envelope of an output waveform with respect to voice information of the learner; Detecting a second envelope of an output waveform of voice information of the native language user; Detecting a difference between the first envelope and the second envelope; And And generating and outputting an accent analysis chart indicating a difference between the detected first envelope and the second envelope. The method according to claim 9 or 10, Detecting a first starting point of a next word following the first word in an output waveform for the voice information of the learner; Detecting a second starting point of a next word following the first word in an output waveform of voice information of the native language user; Detecting a time difference between the first starting point and the second starting point; And And generating and outputting a rhythm analysis chart indicating a time difference between the detected first starting point and the second starting point. The method according to claim 9 or 10, Detecting a first non-signal interval between the next word following the first word in the output waveform for the voice information of the learner; Detecting a second non-signal interval between the next word following the first word in the output waveform of the voice information of the native language user; Detecting a time difference between the first no-signal section and the second no-signal section; And And generating and outputting a rhythm analysis chart indicating a time difference between the detected first non-signal section and the second no-signal section.