CN1581130A - Interactive language-learning method with speech-sound indentification function - Google Patents
Interactive language-learning method with speech-sound indentification function Download PDFInfo
- Publication number
- CN1581130A CN1581130A CNA031535364A CN03153536A CN1581130A CN 1581130 A CN1581130 A CN 1581130A CN A031535364 A CNA031535364 A CN A031535364A CN 03153536 A CN03153536 A CN 03153536A CN 1581130 A CN1581130 A CN 1581130A
- Authority
- CN
- China
- Prior art keywords
- speech
- speech recognition
- voice
- data
- language learning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Electrically Operated Instructional Devices (AREA)
Abstract
The present invention relates to an interactive language learning method with speech identification, in particular, it is a method using speech identification technique as interactive type language learning measure to analyze that the language exercised by comparison exerciser is correct or net. Said invention includes a speak-following mode or an interactive mode, and said method includes the following steps: picking and back playing any language sound data; waiting a time value; letting exerciser to input an exercised sound signal, making speech identification and generating said speech identification data; comparing said speech identification data with said language sound data to produce an approximation value, finally comparing said approximation value and said preset regulation value, storing correct or error information so as to attain the goal of said invention.
Description
Technical field
Whether correct the present invention is a kind of interacting language learning method of tool speech recognition, particularly about a kind of interactively interactive learning methods, utilize the speech recognition technical Analysis comparison language that the practitioner practised mode.
Background technology
Current, English is the widest language of a kind of world pop, march toward international today in order to face, and the world that adds behind the W.T.O. is impacted, want to survive in just strengthening promoting English ability in the world, therefore how encouraging the automatic and spontaneous English learning of people, with Promoting International Competition Ability, is considerable.But the most important key that learns a language is exactly a vocabulary, unless yet have the language teacher to assist dialogue on the side, and the pronunciation of correcting the practitioner, otherwise most people only can learn to listen from books, audiotape or computer software, reading and writing, and can't practise.
Present language teaching medium among the people are various, numerous, single with regard to English teaching material and many rapid-result form of teachings, mostly focus on the listening of language, reading and writing memory exercise, and the exercise that can't focus on, chief reason is be that the practitioner can't judge whether voluntarily correct, does not have that relevant software and hardware auxiliary exercise person judges whether yet correctly.
Aspect patent documentation, as Taiwan patent announcement number No. 470904, be a kind of interactive instructional system and method, it is to disclose a kind of e-learning system and interactively computer learning method that utilizes computing machine, can be online by plurality of subscriber and a server, see through the learning system database of this server, carry out the language learning on the network.
And for example the Taiwan patent announcement is number No. 472222, be a kind of computer aided language learning method and system, a kind of computing machine that utilizes of same exposure assists the user to carry out language exercises such as vocabulary, the syntax, sentence pattern, wherein more include speech database, can send correct voice for user's exercise.
Yet two above-mentioned patents similarly have the shortcoming of can't auxiliary exercise person judging that its language of saying is whether correct, therefore the above-mentioned existing language teaching auxiliary media of the artificial solution of this case invention can't be assisted the defective of novels, anecdotes, etc. branch, be specially to concentrate on studies and cooperate the utilization of scientific principle, a kind of interacting language learning method of tool speech recognition is proposed, be to utilize very popular at present speech recognition technology, be combined in language learning assistant software or the hardware, can be by the part of speech recognition assisting language learning person practice theory, be a kind of reasonable in design and effectively improve the invention of above-mentioned defective.
Summary of the invention
The present invention provides the purpose that following technological means is reached interacting language learning:
Major technique feature of the present invention is the interacting language learning method that is to provide a kind of tool speech recognition, whether the language with the analyses and comparison practitioner is correct, the present invention includes one with saying pattern or an interactive mode, its method at first captures and plays arbitrary these speech data, wait for a time value, after allowing the practitioner import an exercise voice signal, carry out speech recognition and produce this speech recognition data, then compare these speech recognition data and this speech data and produce a degree of approximation value, relatively this degree of approximation value is preset adjusted value with this at last, store the correct or error message that this practitioner practises these speech data, so that add up the correct or error message record of these all exercises of practitioner, reach the effect of interacting language learning.
Description of drawings
Fig. 1 is the configuration diagram that the present invention is used in an one-of-a-kind system.
Fig. 2 is the configuration diagram that the present invention is used in a network system.
Fig. 3 is the schematic flow sheet that first embodiment of the invention is followed pattern.
Fig. 4 is the schematic flow sheet of second embodiment of the invention interactive mode.
Symbol description
1 one-of-a-kind system
2 computer installations
3 language learning main frames
10 central processing units
11 speech recognition devices
12 language Storage Medias
13 voice playing device
14 sound capture devices
15 displays
Embodiment
Reach technology, means and the effect that predetermined purpose is taked in order to make your auditor can further understand the present invention, see also following about detailed description of the present invention and accompanying drawing, believe purpose of the present invention, feature and characteristics, go deep into and concrete understanding when obtaining one thus, yet appended graphic reference and the explanation usefulness of only providing not is to be used for the present invention is limited.
Seeing also shown in Figure 1ly, is the configuration diagram that the present invention is used in an one-of-a-kind system, and Fig. 2 is the configuration diagram that the present invention is used in a network system.The interacting language learning method of tool speech recognition of the present invention can be used with on the one-of-a-kind system 1, as a personal computer (PC) or a carry-on language learner, a practitioner can be learned a language by this one-of-a-kind system 1.The present invention also can use in the network system of a master-slave architecture, utilizes a computer installation 2 to be online on the amoyese speech study main frame 3, so that allow this practitioner of plural number carry out language learning.
If the present invention is when being used in one-of-a-kind system 1, at least comprise a central processing unit 10, speech recognition device 11, language Storage Media 12, a voice playing device 13 and a sound capture device 14 in this language learner, the present invention is if use when network system, at least comprise a central processing unit 10, speech recognition device 11, a language Storage Media 12 in this language learner 3, and this remote computer device 2 comprises a voice playing device 13 and a sound capture device 14 at least.
Wherein this language Storage Media 12 is to be a language database or language archives, literal and speech datas such as individual character, phrase, statement or question answer dialog that plural language learning is used have wherein been stored, and this voice playing device 13 is in order to play the speech data in this language Storage Media 12, can be a sound card or loudspeaker, the output terminal of this sound card can connect these loudspeaker, and this sound capture device 14 is in order to capture this practitioner's exercise sound, can be a sound card or a microphone, the input end of this sound card is connected to this microphone.
Wherein this central processing unit 10 is in order to carry out a language learning program, can be by this program control or write down this practitioner's study schedule or statistical learning achievement etc., and this speech recognition device 11 is the exercise sound of being imported in order to this practitioner of identification, so that it is compare, whether correct to determine the exercise sound that this practitioner imported with the speech data that is stored in this language Storage Media 12.
The language learning program that the present invention is performed, mainly include two kinds of modes of learning, first for following the pattern of saying, second is interactive mode, and each pattern can include the bilingual kenel, for example with Chinese study English with saying or talking with kenel, perhaps with English learning English with saying or talking with kenel, be the schematic flow sheet that first embodiment of the invention is followed pattern as shown in Figure 3, therefore before the present invention carried out this language learning program, needing to set earlier this language learning pattern was with saying pattern or interactive mode 100.
In this embodiment, at first acquisition is stored in arbitrary these speech data in this language Storage Media 12, as English-word or statement, and play out this speech data 101 by these loudspeaker, and can capture these speech data of required study one by one according to the course progress of study, with Chinese study English is example, just may include English voice and a Chinese speech in these speech data, and this Chinese speech is the translated speech corresponding to these English voice, when playing these speech data, can play this Chinese speech earlier, play this English voice again, allow the practitioner by an exercise of this microphone input voice signal, also promptly with saying this English voice then.
Then the present invention waits for a time value 102, five seconds for example, if this practitioner is with saying this English voice in this five seconds, also promptly in this five seconds for not importing this exercise voice signal, perhaps represent that this practitioner does not catch as yet, then repeat to play again these speech data once, make this practitioner repeat to listen to.After this practitioner imported this exercise voice signal 103 by this microphone, the present invention promptly carried out this exercise voice signal of speech recognition, produces speech recognition data 104.
At the speech recognition technical elements, have greatly improved at present, most typical speech recognition includes: should connect distinctiveness ratio relative method, LPC characteristic parameter acquisition method and voice sound bag analytical comparison ... or the like, there can be any more thousands of pieces relevant paper and numerous scholar experts to work out technology up to 90% discrimination power, because the present invention is not an application speech recognition technology, but in detail use this speech recognition technology, so its technology contents is described in detail no longer.The present invention is an example with LPC characteristic parameter acquisition method, exercise voice signal with this practitioner is converted to a speech waveform earlier, then this speech waveform is distinguished into a series of sound frame, obtain one group of linear predictive coefficient for each sound frame then, capture the characteristic ginseng value of its alt wave energy at last, to produce this speech recognition data.
After the present invention obtains these speech recognition data, then compare these speech recognition data and this speech data and produce a degree of approximation value 105, determine this practitioner to practise the correctness of these speech data by this degree of approximation value.And the comparison method also the method with speech recognition is identical, should practise voice signal and this speech data all are converted to speech waveform, by at least one characteristic ginseng value of acquisition in this speech waveform, whether close, and produce this degree of approximation value if comparing this characteristic ginseng value more one by one.
Relatively this degree of approximation value and a default adjusted value 106 at last, if this degree of approximation value is higher than this default adjusted value, represent that promptly practitioner the exercise voice signal of following and these speech sound data of being play are approximate, finish the language learning of this individual character or statement, but if this degree of approximation value is lower than this default adjusted value, then can send the voice of an error messages, require the practitioner again again with saying once, and should default adjusted value can adjust the compared proportions of itself and this degree of approximation value in advance, the present invention is distinguished into high/medium/low three kinds of comparison accuracy with it, the beginner is with the default adjusted value of low accuracy, and advance the rank person can use in/the high default adjusted value of accuracy.
No matter the present invention's correctness after finishing the statement exercise each time all can store the correct or error message 107 that this practitioner practises these speech data, and write down numbering, exercise number of times or the practice periods of these speech data of being practised.If after finishing course or learning phase, can add up the correct of these all exercises of practitioner or error message record 108, and the back of mark is with a display 15 these appraisal result of demonstration.And the numbering of these speech data that write down, exercise number of times or practice periods can be used as and repeat the reference data practised backward, and serve as preferential acquisition and play reference that also can practice periods isolating this speech data number more of a specified duration be preferential acquisition and broadcast reference with mistake this speech data number more repeatedly.
See also shown in Figure 4, it is the schematic flow sheet of second embodiment of the invention interactive mode, the flow process of interactive mode of the present invention roughly with saying that pattern is identical, its difference is in and includes question sentence voice and one answer a voice in these speech data, and these question sentence voice are the usefulness as broadcast, and this answers the usefulness that a voice is exercise voice signals of this practitioner as a comparison.
In this embodiment, acquisition earlier similarly is stored in arbitrary these speech data in this language Storage Media 12, and play out this speech data 201 by these loudspeaker, with Chinese study English is example, include English question sentence voice, Chinese question sentence voice and an English in these speech data and answer a voice, and play these Chinese question sentence voice earlier, play this English question sentence voice again, allow the practitioner answer out this English then and answer a voice by this microphone input.
Then the present invention waits for a time value 202, after this practitioner imports this exercise voice signal 203 by this microphone, the present invention promptly carries out this exercise voice signal of speech recognition, produce this speech recognition data 204, then compare these speech recognition data and produce a degree of approximation value 205 with English these speech data of answering sentence, relatively this degree of approximation value is preset adjusted value 206 with this at last, store the correct or error message 207 that this practitioner practises these speech data, so that add up the correct of these all exercises of practitioner or error message record 208.
So the present invention really can borrow above-mentioned disclosed technology, provides a kind of far different in known person's design, may be able to improve whole use value, do not see publication or public use before its application again, really met the requirement of patent of invention, so propose the application of patent of invention in accordance with the law.
More than disclosed graphic, explanation, only be embodiments of the invention, allly be skillful in this skill person when can doing other all improvement according to above-mentioned explanation, and these change still belong to invention spirit of the present invention and below in the claim that defined.
Claims (24)
1. the interacting language learning method of a tool speech recognition is characterized in that this method comprises the following steps: at least
Acquisition is also play speech data;
A practitioner's of input exercise voice signal;
Speech recognition should be practised voice signal, produced speech recognition data; And
Compare these speech recognition data and this speech data, produce a degree of approximation value, determine this practitioner to practise the correctness of these speech data by this degree of approximation value.
2. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, more comprises before wherein capturing the step of these speech data:
Set a language learning pattern and be one with saying pattern or an interactive mode.
3. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, wherein capturing these speech data is by arbitrary these speech data of acquisition in the data storage medium.
4. the interacting language learning method of tool speech recognition as claimed in claim 3 is characterized in that, wherein capturing these speech data is by capturing wherein a certain these speech data one by one according to the course progress in this data storage medium.
5. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, wherein this speech packet contains first voice and second voice, and these second voice are the translated speech corresponding to these first voice.
6. the interacting language learning method of tool speech recognition as claimed in claim 4 is characterized in that wherein this first language is English voice, and this second language is a Chinese speech.
7. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, wherein playing these speech data is to play this speech data by these loudspeaker.
8. the interacting language learning method of tool speech recognition as claimed in claim 1, it is characterized in that, wherein play in the step of these speech data, if in these speech data when comprising one first voice and one second voice, play these second voice earlier, play this first voice again.
9. the interacting language learning method of tool speech recognition as claimed in claim 8 is characterized in that, wherein these first voice are English voice, and these second voice are Chinese speeches.
10. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, more comprises the following steps: before wherein importing this practitioner's the step of this exercise voice signal
Wait for a time value; And
In this time value, do not import this exercise voice signal as if this practitioner, then these speech data of repeat playing.
11. the interacting language learning method of tool speech recognition as claimed in claim 10 is characterized in that, wherein this time value can be five seconds.
12. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, wherein imports this practitioner and should practise voice signal and be to use a microphone.
13. the interacting language learning method of tool speech recognition as claimed in claim 1, it is characterized in that, wherein these speech data can be answered a voice for question sentence voice and one, and these question sentence voice are as the usefulness of broadcast, and this answers the usefulness that a voice is exercise voice signals of this practitioner as a comparison.
14. the interacting language learning method of tool speech recognition as claimed in claim 13 is characterized in that, wherein these question sentence voice are to can be used as an English voice question sentence or a Chinese speech question sentence.
15. the interacting language learning method of tool speech recognition as claimed in claim 13 is characterized in that, wherein this answer a voice be can be used as English voice answer the sentence or a Chinese speech answer sentence.
16. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, wherein more comprises the following steps: in the step of this exercise voice signal of speech recognition
Changing this exercise voice signal is a speech waveform; And
Capture at least one characteristic ginseng value in this speech waveform, produce this speech recognition data.
17. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, wherein compares in the step of these speech recognition data and these speech data, this comparison mode more comprises the following steps:
Should practise voice signal and this speech data all are converted to speech waveform;
By at least one characteristic ginseng value of acquisition in this speech waveform, whether close, and produce this degree of approximation value if comparing this characteristic ginseng value more one by one.
18. the interacting language learning method of tool speech recognition as claimed in claim 1 is characterized in that, more comprises the following steps: after wherein comparing the step of these speech recognition data and these speech data
Relatively this degree of approximation and a default adjusted value;
If this degree of approximation value is higher than this default adjusted value, then finish this language learning; And
If this degree of approximation value is lower than this default adjusted value, then send a garbled voice, require to re-enter this exercise voice signal.
19. the interacting language learning method of tool speech recognition as claimed in claim 18 is characterized in that, wherein should can adjust the relatively ratio of this degree of approximation value in advance by default adjusted value, and is distinguished into high/medium/low three kinds of comparison accuracy.
20. the interacting language learning method of tool speech recognition as claimed in claim 1, it is characterized in that, more comprise after wherein comparing the step of these speech recognition data and these speech data: store the correct or error message that this practitioner practises these speech data, and write down numbering, exercise number of times or the practice periods of these speech data.
21. the interacting language learning method of tool speech recognition as claimed in claim 20, it is characterized in that, it more comprises after storing the step of comparing record: all practise the correct or error message record of these speech data to add up this practitioner, after being marked, show this appraisal result with a display.
22. the interacting language learning method of tool speech recognition as claimed in claim 21 is characterized in that, wherein writes down the numbering of these speech data, the step of exercise number of times or practice periods, can be used as and repeats the reference data practised backward.
23. the interacting language learning method of tool speech recognition as claimed in claim 22 is characterized in that, wherein this reference data that repeats to practise is to serve as preferential acquisition and play with mistake this speech data number more repeatedly.
24. the interacting language learning method of tool speech recognition as claimed in claim 22 is characterized in that, wherein this reference data that repeats to practise is to serve as preferential acquisition and play with practice periods isolation this speech data number more of a specified duration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA031535364A CN1581130A (en) | 2003-08-15 | 2003-08-15 | Interactive language-learning method with speech-sound indentification function |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA031535364A CN1581130A (en) | 2003-08-15 | 2003-08-15 | Interactive language-learning method with speech-sound indentification function |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1581130A true CN1581130A (en) | 2005-02-16 |
Family
ID=34580100
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA031535364A Pending CN1581130A (en) | 2003-08-15 | 2003-08-15 | Interactive language-learning method with speech-sound indentification function |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1581130A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006136061A1 (en) * | 2005-06-24 | 2006-12-28 | Intel Corporation | Measurement and presentation of spoken language fluency |
| US8234114B2 (en) | 2009-02-27 | 2012-07-31 | Industrial Technology Research Institute | Speech interactive system and method |
| CN103514768A (en) * | 2013-10-24 | 2014-01-15 | 苏州市思玛特电力科技有限公司 | Auxiliary teaching system |
| CN103941868B (en) * | 2014-04-14 | 2017-08-18 | 美的集团股份有限公司 | Voice command accuracy rate method of adjustment and system |
| CN108847068A (en) * | 2018-07-11 | 2018-11-20 | 北京美高森教育科技有限公司 | View-based access control model is associated instantaneously with saying speech training method |
| CN109147404A (en) * | 2018-07-11 | 2019-01-04 | 北京美高森教育科技有限公司 | A kind of detection method and device of the phonetic symbol by incorrect pronunciations |
| CN109147419A (en) * | 2018-07-11 | 2019-01-04 | 北京美高森教育科技有限公司 | Language learner system based on incorrect pronunciations detection |
| CN109255988A (en) * | 2018-07-11 | 2019-01-22 | 北京美高森教育科技有限公司 | Interactive learning methods based on incorrect pronunciations detection |
-
2003
- 2003-08-15 CN CNA031535364A patent/CN1581130A/en active Pending
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2006136061A1 (en) * | 2005-06-24 | 2006-12-28 | Intel Corporation | Measurement and presentation of spoken language fluency |
| US8234114B2 (en) | 2009-02-27 | 2012-07-31 | Industrial Technology Research Institute | Speech interactive system and method |
| CN103514768A (en) * | 2013-10-24 | 2014-01-15 | 苏州市思玛特电力科技有限公司 | Auxiliary teaching system |
| CN103941868B (en) * | 2014-04-14 | 2017-08-18 | 美的集团股份有限公司 | Voice command accuracy rate method of adjustment and system |
| CN108847068A (en) * | 2018-07-11 | 2018-11-20 | 北京美高森教育科技有限公司 | View-based access control model is associated instantaneously with saying speech training method |
| CN109147404A (en) * | 2018-07-11 | 2019-01-04 | 北京美高森教育科技有限公司 | A kind of detection method and device of the phonetic symbol by incorrect pronunciations |
| CN109147419A (en) * | 2018-07-11 | 2019-01-04 | 北京美高森教育科技有限公司 | Language learner system based on incorrect pronunciations detection |
| CN109255988A (en) * | 2018-07-11 | 2019-01-22 | 北京美高森教育科技有限公司 | Interactive learning methods based on incorrect pronunciations detection |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Chen | Developing and evaluating an oral skills training website supported by automatic speech recognition technology | |
| Bernstein et al. | Automatic evaluation and training in English pronunciation. | |
| CN1804934A (en) | Computer-aided Chinese language phonation learning method | |
| CN1510590A (en) | Language learning system and method with visual prompting to pronunciaton | |
| CN101551952A (en) | Device and method for evaluating pronunciation | |
| JP2003150041A (en) | Story interactive grammar teaching system and method | |
| CN1581130A (en) | Interactive language-learning method with speech-sound indentification function | |
| CN1424665A (en) | Device and operation for dictation test and automatic | |
| Ting et al. | Mobile application and traditional approach for Chinese stroke order instruction in foreign language classroom/Ting Hie-Ling, Ch’ng Looi-Chin and Norseha Unin | |
| CN111383495A (en) | In-class explanation system, method, device and medium for spoken language teaching | |
| US20050144010A1 (en) | Interactive language learning method capable of speech recognition | |
| CN1095580C (en) | Method for deaf-dumb voice learning dialogue and pronunciation synchronous feedback device | |
| CN109448464A (en) | A kind of English- word spelling exercising method | |
| CN111507581B (en) | Course matching method, system, equipment and storage medium based on speech speed | |
| CN1521657A (en) | Computer aided language teaching method and apparatus | |
| CN1448897A (en) | System and method for training foreign language listening and speaking ability by random question asking and sentence making | |
| Pellegrini et al. | Overview of Computer-assisted Language Learning for European Portuguese at L2f. | |
| CN1521703A (en) | Sentence construction and conversation teaching system and method having situation role selection function | |
| CN109545014A (en) | A kind of foreign language word exercising method based on interactive voice | |
| Watanabe et al. | Investigating the Effect of Chinese Pronunciation Teaching Materials Using Speech Recognition and Synthesis Functions. | |
| CN1607519A (en) | English learning and testing system | |
| TW581988B (en) | System of listening for spelling and the method thereof | |
| CN1490749A (en) | Computer aided foreign language listening and speaking teaching system and method with situated reading and testing | |
| TWI220970B (en) | Linear listening, speaking and follow-reading language learning system and method | |
| Li et al. | Research on Oral English Learning Platform Based on Intelligent Recognition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |