CN107123416B - A kind of method and device for identifying the potential pronunciation word of infant - Google Patents
A kind of method and device for identifying the potential pronunciation word of infant Download PDFInfo
- Publication number
- CN107123416B CN107123416B CN201710301049.4A CN201710301049A CN107123416B CN 107123416 B CN107123416 B CN 107123416B CN 201710301049 A CN201710301049 A CN 201710301049A CN 107123416 B CN107123416 B CN 107123416B
- Authority
- CN
- China
- Prior art keywords
- infant
- word
- potential
- pronounce
- pronunciation
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L25/00—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00
- G10L25/48—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use
- G10L25/51—Speech or voice analysis techniques not restricted to a single one of groups G10L15/00 - G10L21/00 specially adapted for particular use for comparison or discrimination
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
- G10L2015/025—Phonemes, fenemes or fenones being the recognition units
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
- G10L2015/027—Syllables being the recognition units
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Machine Translation (AREA)
- Electrically Operated Instructional Devices (AREA)
Abstract
The present invention discloses a kind of method and apparatus for identifying the potential pronunciation word of infant.This method includes：Collect and input the word that can pronounce of infant, its initial consonant and simple or compound vowel of a Chinese syllable are extracted, reconfigured further according to the arrangement mode of Chinese Pinyin syllables table, the potential pronunciation word of different phonetic words, i.e. infant is then obtained, is finally applied to teaching.The device includes：Acoustic information collects input module, identifies acoustic information and carries out Pinyin Input by the input module；Acoustic information system module, the extraction of initial consonant and simple or compound vowel of a Chinese syllable is carried out according to input Pinyin and reconfigured, obtains different phonetic words.The present invention utilizes the formation rule of single syllable Chinese character, with reference to the technological means such as speech recognition and restructuring, only input a variable of " can pronounce word ", you can provide the technical scheme of the module application such as the creation material of suitable individual infant for being related to the types such as nursery rhymes a kind of and the articulation ability Real-Time Evaluation of infant, deaf-mute and China and foreign countries' personage's Chinese.
Description
Technical field
The present invention relates to one kind to identify infant（0-3 year）The method of articulation ability, and in particular to one kind identification infant
The method and device of potential pronunciation word, belongs to computer identification, learning areas.
Technical background
The potentiality of human development depend on children.Children's development of potential, it is the focus of contemporary society's concern, even more education is led
The worldwide focus in domain.Coefficient correlation about studies have shown that language pronouncing ability and the whole intellectual level of baby for 0.7 with
On.That is infancy language ability can draw an about equal sign with intelligence.So exploitation infantile language ability is all
The most important thing of development of potential." language pronouncing ability " is to pass through voice（Syllable）What the quantized datas such as identification were got.Substantially patrol
Collecting is：Syllable is the most natural phonetic unit that auditory perception arrives, and " articulation ability " here refers to that infant can be with（Energy）
The most syllables sent.
At present, it is that raising infant and children's phonetics talk about the software that ability is specially developed, such as language to have on the market a lot
Say the nursery rhymes software of class.But all these nursery rhymes application software are the solutions of " digitlization ", such as prominent domestic youngster
" nursery rhymes are the more " and " shellfish watt nursery rhymes " in APP etc. is sung, its outstanding problem is shown：Technology can not be used in teaching process
Means embody some basic educational thoughts such as incremental, differential treatment in education and good interaction, can not in real time reflect and there emerged a
The articulation ability situation of body oral cavity of baby assessment of growth and development, therefore do not promote to a great extent, or even hinder baby children
The abundant exploitation of potential is sayed in childrenese.
The content of the invention
An object of the present invention is to provide a kind of method for identifying the potential pronunciation word of infant, and another object is to provide one
The device of the potential pronunciation word of kind identification infant, the pronunciation level formed according to infant itself, gradual approaching, personalization
The potential of individual infant's phonetics words is effectively developed, to make up the shortcomings of the prior art.
To reach above-mentioned purpose, the concrete technical scheme that the present invention takes is：
A kind of method for identifying the potential pronunciation word of infant, this method using the Chinese characters form of single syllable feature into rule, i.e.,
Each Chinese character is combined by both initial consonant and simple or compound vowel of a Chinese syllable, or is individually composed by simple or compound vowel of a Chinese syllable or initial consonant.
The above method includes collecting infant first can pronounce the phonetic of word, then be inputted, according to the spelling of input
The initial consonant and simple or compound vowel of a Chinese syllable that the phonetic of sound word is formed extract respectively, further according to the permutation and combination method of Chinese Pinyin syllables table
Reconfigured, be then able to obtain the potential pronunciation word of new different phonetic words, i.e. infant, finally carry out defeated
Go out.
Further, the step of above method also includes infant's voice is identified.
Further, the speech recognition steps are：Obtain voice signal, extraction audio fluctuation characteristic signal, with having established
Audio repository matched, the audio signal matched, further according to audio database corresponding with sound, obtain corresponding sound
Message ceases, i.e., the phonetic of the word that can pronounce of above-mentioned infant.
Further, due to infant's burring or the phenomenon of appearance ambiguity, characteristic signal is fluctuated in extraction audio,
And when being matched, the standard of matching is reduced, it can reach matching generally according to the registration for reaching more than 50%.
Further, an increasingly perfect database and correspondingly is formed by infant's voice messaging for being constantly collected into
Audio database, can will match the obtained audio signal priority match database, obtain more accurate acoustic information.
Data of the above method also including can pronounce to described infant word and potential pronunciation word quantify, and wrap
Include：Confirm the number of syllables of pronunciation word and identify the number of its available phonetically similar word.
The above method also includes carrying out the described number of syllables of infant's pronunciation word and the number of its available phonetically similar word
Sort out and integrate, the classification of both syllable and its phonetically similar word number is：Can pronounce word and potential pronunciation word, and single syllable
The classification of number is：Show articulation ability and potential articulation ability.Result after being integrated is：Can pronounce word（Performance hair
Sound ability）+ potential pronunciation word（Potential articulation ability）.
The above method also includes the word that can pronounce to described infant（Show articulation ability）With potential pronunciation word（It is potential
Articulation ability）Result after being sorted out and being integrated is：Obtain infant can be with（Energy）All pronunciation words of pronunciation and pronunciation
Ability.It expresses formula：Can pronounce word（Show articulation ability）+ potential pronunciation word（Potential articulation ability）=all hairs
Sound word（Articulation ability）.
The above method is also included to all pronunciation words of described infant, as writing material, applied to for personalization
The creation of the types such as nursery rhymes, children's folk rhymes and story provides data and supported.
The above method is also included to the articulation ability of described infant, as objective parameter, applied to for individual and group
The evaluation of the articulation ability of body infant provides data and supported.
A kind of device for identifying the potential pronunciation word of infant, the device include：
Acoustic information collects input module, identifies acoustic information and carries out Pinyin Input by the input module；
Acoustic information system module, the extraction of initial consonant and simple or compound vowel of a Chinese syllable is carried out according to input Pinyin, and according to Chinese Pinyin syllables
The permutation and combination method of table is reconfigured, and is then able to obtain the potential pronunciation word of different phonetic words, i.e. infant.
Further, said apparatus also includes sound identification module, receives the voice signal of infant, extracts audio-frequency information
Feature；The device also includes speech processing module, according to the audio-frequency information feature extracted and the audio repository established progress
Match somebody with somebody, the audio signal matched, further according to audio database corresponding with sound, obtain corresponding acoustic information, further according to
Acoustic information input module automatically enters Pinyin information.
Further, in above-mentioned speech processing module, due to infant's burring or there is indistinct phenomenon, carrying
When taking audio to fluctuate characteristic signal, and being matched, the standard of matching is reduced, generally according to the registration for reaching more than 50%
Matching can be reached.
Further, in above-mentioned speech processing module, a day is formed by the infant's voice messaging being constantly collected into
Beneficial perfect database and corresponding audio database, the obtained audio signal priority match database can will be matched, obtained
To more accurate acoustic information.
The advantages of the present invention are：
The present invention utilizes the formation rule of single syllable Chinese character, can be real-time with reference to the technological means such as speech recognition and restructuring
Detect all pronunciation words for meeting individual infant's Physiological and psychological requirements（Including the word that can pronounce, it is potential pronunciation word and its
Phonetically similar word）, and it conveniently can apply and use, and be carried out artificially by all pronunciation words of output as application program
Active layout, or internal processes layout, nursery rhymes suitable for reading or children's folk rhymes bright etc. are formed, the language being consequently formed is accurate, follows
The progressive articulation ability and cognitive ability that infant can be effectively lifted with the teaching pattern of differential treatment in education of sequence.Also, pass through
The big data analysis of all pronunciation words of colony infant and academic research achievement etc., can form objective evaluation standard, and then
Parents can understand the pronunciation situation of oneself child in real time by the evaluation, obtain the objective basis of a judgement.
The present invention only inputs a variable of " can pronounce word ", it is possible to provides one kind and is related to nursery rhymes, children's folk rhymes and story
It is real-time Deng the creation material of the suitable individual infant of type and the articulation ability of infant, deaf-mute and China and foreign countries' personage's Chinese
The technical scheme of the module applications such as evaluation.
Brief description of the drawings
Fig. 1 is the process chart of the present invention；
Fig. 2 is the quantized data process chart of the present invention；
Fig. 3 is the articulation ability evaluation rubric figure of the present invention；
Fig. 4 is the APP functional schematics of the articulation ability evaluation of the present invention.
Embodiment
The present invention is explained further and illustrated below by way of specific embodiment.
Embodiment 1：
Fig. 1 is shown according to a particular embodiment of the invention, for the place of the recognition methods of the potential pronunciation word of infant
Manage flow chart；Collect including speech recognition, the female extraction of phonetic sound, sound mother combined treatment, sound mother, in addition to specifically should
With：S101 to S105.
In step S101, by obtaining voice signal, extraction audio fluctuation characteristic signal, enter with the audio repository established
Row matching, the audio signal matched, further according to audio database corresponding with sound, obtains corresponding acoustic information：
Can be pronounced word phonetic.In addition, step S101 can also be completed in another way：It is manually entered the phonetic of infant's voice signal
Word, i.e., can pronounce word phonetic.
In step s 102, according to the word phonetic that can pronounce of input, decomposition and the extraction process of initial consonant and simple or compound vowel of a Chinese syllable are carried out,
Obtain corresponding initial consonant and simple or compound vowel of a Chinese syllable.
In step s 103, by the initial consonant and simple or compound vowel of a Chinese syllable of extraction, carried out according to the permutation and combination method of Chinese Pinyin syllables table
Processing is reconfigured, obtains new different phonetic words, i.e., potential pronunciation word.
In step S104, the potential pronunciation word that will be obtained in can pronounce word and the step 103 that are obtained in step S101
Collected, the word that pronounces for obtaining the oral cavity of individual infant and physiologically allowing, i.e., all pronunciation words, i.e. articulation ability.
In step S105, resulting all pronunciation words and articulation ability are respectively applied to write material, realized individual
Property teaching and oral cavity development evaluation platform, and realize Real-Time Evaluation.
It is mentioned above, is to employ voice to know by way of the acoustic information of characteristic signal acquisition is fluctuated extracting audio
The technological means such as not, advantage are can to ensure the timely and complete of data with round-the-clock collected sound signal, but due to baby children
Not yet the factor such as maturation and environmental disturbances, accuracy may be not good enough for the oral cavity development of youngster.So in extraction audio fluctuation characteristic
When signal is matched, matching standard is reduced, may achieve matching generally according to the registration for reaching more than 50%.
Further, it is previously mentioned, using speech recognition technology, it may occur however that accuracy or error of speech recognition etc. are asked
Topic.So another scheme is manually entered phonetic word, i.e. parent particularly has the word that can pronounce of the infant heard
The voice messaging of language meaning, it is directly inputted to mobile terminal or computer.It is that parent can be with using the advantages of this mode
Micro-judgment, input compares to press close to language and the phonetic word lived into application program, and then solves accuracy problem.The latter's
Scheme can also be used as supplementary means.
Further, it is previously mentioned, the advantages of being manually entered, but counts the word that can pronounce of infant, be not with a certain
Secondary completion, it is necessary to which one cumulative with before.So the word that can pronounce being manually entered every time has one once to tire out with preceding
The step of word that can pronounce of meter is added.
Embodiment 2：
Now the quantized data handling process in the present invention and articulation ability evaluation rubric are introduced respectively.
Fig. 2 shows quantized data process chart according to an embodiment of the invention.Referring to Fig. 2, the process is at least
Including step S201 to step S204.
In step s 201, by way of speech recognition or being manually entered, the spelling of the word that can pronounce of infant is inputted
Sound, such as：Mother [ma], eat [chi] and fat [pang] etc..
In step S202, it will can pronounce the initial consonant of word and simple or compound vowel of a Chinese syllable is collected and extracted, further according to Chinese phonetic alphabet sound
The permutation and combination method of section table is reconfigured, and can obtain new different phonetic words, i.e., potential pronunciation word.Than as described above
Available potential pronunciation word is in the phonetic word of three words that can pronounce：Rice [mi] and hurry [mang], tea [cha] and sing
[chang], climb [pa] and wind [pi] etc..
In step S203, the potential pronunciation word that can pronounce word and step S202 that step S201 is inputted are obtained is integrated
Together, i.e., can pronounce word+potential pronunciation word.
In step S204, the result after integration be exactly obtain infant can be with（Energy）All words of pronunciation, i.e., it is all
Pronounce word.Quantization principles are：Can pronounce word+potential pronunciation word=all pronunciations word, for example can have been sent out from above-mentioned three
Following all pronunciation words and its phonetically similar word can be obtained in sound word and six potential pronunciation words：
Can pronounce word：Mother, horse, scolding, fiber crops, smear, ant etc.；Eat, chi, pond, tooth, shame, late, wing, spoon, sneer；Fat, other,
Pang, crab, etc.；
Potential pronunciation word：Rice, close, miaow, honey, fan etc.；Busy, blind etc.；Tea, difference, look into, insert, pitching；Sing, grow, intestines, tasting；
Climb, be afraid of, lying prone, pa, rake etc.；Wind, skin, split, beer, drape over one's shoulders, Pi etc..
Thus quantify and integrate obtained all pronunciation words of the baby（Articulation ability）Can be following nursery rhymes creation and
Pronunciation evaluation provides data and supported.
Embodiment 3
As mentioned above, the word that can pronounce of individual infant can be obtained speech recognition and by way of being manually entered
With potential pronunciation word, they are quantified and integrated, all pronunciation words of individual infant can be obtained（Articulation ability）.Thus
Obtain following two formulas：
Formula 1：Can pronounce word+potential pronunciation word=all pronunciations word（See module 1）
Formula 2：Show articulation ability+potential articulation ability=articulation ability（See module 2）
Formula 1 includes phonetically similar word, and material is write applied to providing；Formula 2 does not include phonetically similar word, is commented applied to articulation ability
Valency.
It is described below respectively with the module of two difference in functionalitys below：
Module 1：Using all writing materials for being created as nursery rhymes of pronunciation words, there is provided physiology and psychologically meet individual baby
The words of demand, create the nursery rhymes of personalization.Such as：According to all pronunciation words and its phonetically similar word referred in step S204,
Give play to a subjective initiative, parent can be with self-editing from the following nursery rhymes of wound：
Mother sings（It is self-editing to create certainly）
Mother sings,
Taste,
Eat loquat,
Skin skin is fat.
For another example：According to certain baby（0 12 days 1 month 1 year old）All pronunciation words and phonetically similar word and human-subject test, parent
By referring to the form of other nursery rhymes, children's folk rhymes etc. such as classics《Lamp stand on mouselet》Deng being adapted into and be more suitable for oneself baby's
Nursery rhymes：
Lamp stand on mouselet（It is classical）Flower stand on the tabby cat（Reorganization）
Mouselet, upper lamp stand, the tabby cat, upper flower stand,
Oil is eaten steathily, is unable to come down, grabs butterfly, is unable to come down,
Mew mew mew, cat are coming, tearful Wang, and dog comes,
It is fast and furious to roll down.Bouncing and vivacious run down is come.
By materials such as all pronunciation words, parents create in real time according to the infant of Different Month is adapted to individual raw
The nursery rhymes of reason and psychology, so as to embody some basic educational thoughts such as incremental, differential treatment in education and good interaction, and then
Lift IQ and feeling quotrient.
Module 2：As mentioned above, the word that can pronounce of individual infant is inputted, by the computing of formula 2, can be obtained in real time
The performance articulation ability and potential articulation ability of individual, i.e. individual articulation ability.Meanwhile by countless similar individual infants
The word that can pronounce input, computing, the performance articulation ability of colony and potential articulation ability can be obtained in real time again, and then obtain
With colony's articulation ability of monthly age infant, i.e. average value, it can realize real time contrast using the average value and evaluate individual baby children
The articulation ability of youngster.
Fig. 3 shows articulation ability evaluation rubric figure according to an embodiment of the invention.Referring to Fig. 3, the process is at least
Including step S301 to step S304.
In step S301, by way of speech recognition or being manually entered, for individual infant is collected and inputted
Pronounce word（Show articulation ability）.Meanwhile by way of the speech recognition of countless individual infants or being manually entered, collect and
Input the word that can pronounce of colony（Colony shows articulation ability）.
In step s 302, it will can pronounce the initial consonant of word and simple or compound vowel of a Chinese syllable extracts, further according to Chinese Pinyin syllables table
Permutation and combination method is reconfigured, and can obtain new different phonetic words, i.e., potential pronunciation word（Potential articulation ability）.
Meanwhile after countless individual infants complete above-mentioned extraction and restructuring, it will obtain the potential pronunciation word of a colony（Group
The potential articulation ability of body）.
In step S303, the word that can pronounce that step S301 is inputted（Show articulation ability）Obtained with step S302
Potential pronunciation word（Potential articulation ability）Quantized data combine.Meanwhile countless individual infants are pressed into step S301
The word that can pronounce of input（Show articulation ability）The potential pronunciation word obtained with step S302（Potential articulation ability）Quantization
Data Integration is together.
In step s 304, by the step S301 word that can pronounce（Show articulation ability）With step S302 potential pronunciation
Word（Potential articulation ability）Quantized data combine after, all pronunciation words can be obtained（Articulation ability）.Meanwhile by nothing
The word that can pronounce that several infants are inputted by step S301（Show articulation ability）The potential pronunciation word obtained with step S302
（Potential articulation ability）Quantized data combine after, all pronunciation words of a colony will be obtained（Colony's pronunciation energy
Power）.
Further, can pronounce word, potential pronunciation word and all pronunciation words, including phonetically similar word；And corresponding thereto
Performance articulation ability, potential articulation ability and articulation ability refer to number of syllables, not including phonetically similar word.
Further, above mentioned colony's performance articulation ability, the potential articulation ability of colony and colony's articulation ability are equal
Refer to the average value of colony.
Fig. 4 is shown according to one embodiment of the invention（Fig. 3）The mobile terminal APP of design functional schematic.Referring to
Fig. 4, the figure are " articulation ability " page, point or so two parts.The left side represents the number of syllables that infant can send, that is, shows
Articulation ability；The right represents the number of syllables that infant may send, i.e., potential articulation ability.The right and left collects structure together
Into all pronunciation words, i.e. articulation ability.The figure comprises at least 9 parts, and numbering 1-9 and its content are as follows：
1 ：The number of syllables that infant can send, that is, show articulation ability.
2 ：The number of syllables that infant may send, i.e., potential articulation ability.
3 ：The number of syllables that individual infant can send, i.e. individual performance articulation ability.
4 ：The number of syllables that individual infant may send, the i.e. potential articulation ability of individual.
5 ：The number of syllables that this App colonies infant can send（Average value）, i.e., colony performance articulation ability.
6 ：The number of syllables that this App colonies infant may send（Average value）, i.e. the potential articulation ability of colony.
7 ：The number of syllables that the colony infant of academic research can send（Average value）, i.e., colony performance articulation ability.
8 ：The number of syllables that the colony infant of academic research may send（Average value）, i.e. the potential articulation ability of colony.
9 ：Individual infant the actual monthly age display, may be configured as 0 between -36 months, may be alternatively provided as the date or
It etc..
The above-mentioned numbering 1-2 referred to mark, number size can be shown with scroll-up/down；Numbering 3-8 mark, can be with
Move up and down, display capabilities size.
By speech recognition or artificial mode, the word that can pronounce is inputted in related pages such as mobile terminal App homepages, that
This " articulation ability " page will real-time display go out the information of corresponding articulation ability.
Further, by individual speech recognition or artificial mode, the word that can pronounce is sporadically inputted, then its
Performance articulation ability and potential articulation ability etc. the data of body infant will be constantly real on the left side of the page and the right respectively
When show.
Further, by way of artificial or both the mixing of colony's speech recognition or colony, sporadically input and can
Pronounce word, then the data such as performance articulation ability and potential articulation ability of its colony infant will be constantly respectively in the page
The left side and the right real-time display come out.
Further, the performance articulation ability of related academic research achievement and potential articulation ability, it is public as parents
The standard recognized, the left side and the right of the page are respectively displayed on, strengthen and ensure the consciousness and specification of individual parent's input with this
Property, and then the accuracy that individual and colony parents input is lifted, so as to realize the performance articulation ability of this App colonies infant
With the accuracy of the data such as potential articulation ability.The baby of the final benign cycle for showing one " All for one and one for all "
The appraisement system of child's articulation ability.
The preferred embodiments of the present invention are the foregoing is only, are not used for limiting the practical range of the present invention；Every foundation is originally
The made equivalence changes of invention and modification, are all covered by the scope of the claims of the present invention.
Claims (13)
- A kind of 1. method for identifying the potential pronunciation word of infant, it is characterised in that this method utilizes the Chinese character of single syllable feature Rule is formed, i.e., each Chinese character is combined by both initial consonant and simple or compound vowel of a Chinese syllable, or is individually composed by simple or compound vowel of a Chinese syllable or initial consonant；Above-mentioned side Method includes collecting infant first can pronounce the phonetic of word, then be inputted, according to the phonetic of the phonetic word of input by its The initial consonant and simple or compound vowel of a Chinese syllable of composition extract respectively, are reconfigured further according to the permutation and combination method of Chinese Pinyin syllables table, It is then able to obtain the potential pronunciation word of new different phonetic words, i.e. infant, is finally exported.
- 2. as claimed in claim 1 identification infant it is potential pronunciation word method, it is characterised in that the above method also include pair Described infant can pronounce word and it is potential pronunciation word data quantified, including：Confirm pronunciation word number of syllables and Identify the number of its available phonetically similar word.
- 3. as claimed in claim 1 identification infant it is potential pronunciation word method, it is characterised in that the above method also include pair The described number of syllables of infant's pronunciation word and the number of its available phonetically similar word is sorted out and integrated, syllable and its phonetically similar word The classification of both numbers is：Can pronounce word and potential pronunciation word, and individually the classification of number of syllables is：Show articulation ability With potential articulation ability；Result after being integrated is：Can pronounce word+potential pronunciation word.
- 4. the method for the potential pronunciation word of identification infant as claimed in claim 3, it is characterised in that described infant can Pronunciation word and it is potential pronunciation word sorted out and integrated after result be：Obtain all the pronunciation words and hair that can pronounce of infant Sound ability；It expresses formula：Can pronounce word+potential pronunciation word=all pronunciations word.
- 5. the method for the potential pronunciation word of identification infant as described in any claims of claim 1-4 can be applied to be individual Property nursery rhymes, children's folk rhymes and story type creation provide data support.
- 6. the method for the potential pronunciation word of identification infant as described in any claims of claim 1-4 can be applied to be individual The evaluation of body and the articulation ability of colony infant provides data and supported.
- 7. as claimed in claim 1 identification infant it is potential pronunciation word method, it is characterised in that the above method also include pair The step of infant's voice is identified.
- 8. the method for the potential pronunciation word of identification infant as claimed in claim 7, it is characterised in that the speech recognition steps For：Voice signal is obtained, extraction audio fluctuation characteristic signal, is matched with the audio repository established, the audio matched Signal, further according to audio database corresponding with sound, corresponding acoustic information is obtained, i.e., above-mentioned infant's can pronounce The phonetic of word.
- 9. the method for the potential pronunciation word of identification infant as claimed in claim 8, it is characterised in that due to infant's pronunciation not When phenomenon that is clear or that ambiguity occur, characteristic signal is fluctuated in extraction audio, and being matched, the standard of matching is reduced, is pressed It can reach matching according to the registration for reaching more than 50%.
- 10. a kind of device for identifying the potential pronunciation word of infant, it is characterised in that the device includes：Acoustic information collects input module, identifies acoustic information and carries out Pinyin Input by the input module；Acoustic information system module, the extraction of initial consonant and simple or compound vowel of a Chinese syllable is carried out according to the phonetic of input, and according to Chinese Pinyin syllables table Permutation and combination method reconfigured, be then able to obtain the potential pronunciation word of different phonetic words, i.e. infant.
- 11. the device of the potential pronunciation word of identification infant as claimed in claim 10, it is characterised in that said apparatus also includes Sound identification module, the voice signal of infant is received, extract audio-frequency information feature；The device also includes speech processing module, Audio-frequency information feature according to extracting is matched with the audio repository established, the audio signal matched, further according to sound Frequency database corresponding with sound, obtains corresponding acoustic information, and phonetic letter is automatically entered further according to acoustic information input module Breath.
- 12. the device of the potential pronunciation word of identification infant as claimed in claim 11, it is characterised in that above-mentioned speech processes mould In block, due to infant's burring or there is indistinct phenomenon, in extraction audio fluctuation characteristic signal, and matched When, the standard of matching is reduced, can reach matching according to the registration for reaching more than 50%.
- 13. the device of the potential pronunciation word of identification infant as claimed in claim 12, it is characterised in that above-mentioned speech processes mould In block, an increasingly perfect database and corresponding voice data are formed by the infant's voice messaging being constantly collected into Storehouse, the obtained audio signal priority match database can will be matched, obtain more accurate acoustic information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710301049.4A CN107123416B (en) | 2017-05-02 | 2017-05-02 | A kind of method and device for identifying the potential pronunciation word of infant |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710301049.4A CN107123416B (en) | 2017-05-02 | 2017-05-02 | A kind of method and device for identifying the potential pronunciation word of infant |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107123416A CN107123416A (en) | 2017-09-01 |
| CN107123416B true CN107123416B (en) | 2018-03-23 |
Family
ID=59726557
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710301049.4A Active CN107123416B (en) | 2017-05-02 | 2017-05-02 | A kind of method and device for identifying the potential pronunciation word of infant |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107123416B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102663925A (en) * | 2012-05-09 | 2012-09-12 | 长春大学 | Method and system for tongue training for language training of hearing-impaired children |
| CN103405217A (en) * | 2013-07-08 | 2013-11-27 | 上海昭鸣投资管理有限责任公司 | System and method for multi-dimensional measurement of dysarthria based on real-time articulation modeling technology |
| CN203733335U (en) * | 2014-01-23 | 2014-07-23 | 王亚婷 | Chinese phonetic alphabet teaching aid |
| CN104598758A (en) * | 2015-02-12 | 2015-05-06 | 上海市徐汇区中心医院 | System and method for evaluating hearing-speech rehabilitation training and curative effect of patients with post-stroke dysarthria |
| US9508268B2 (en) * | 2006-05-22 | 2016-11-29 | Koninklijke Philips N.V. | System and method of training a dysarthric speaker |
-
2017
- 2017-05-02 CN CN201710301049.4A patent/CN107123416B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9508268B2 (en) * | 2006-05-22 | 2016-11-29 | Koninklijke Philips N.V. | System and method of training a dysarthric speaker |
| CN102663925A (en) * | 2012-05-09 | 2012-09-12 | 长春大学 | Method and system for tongue training for language training of hearing-impaired children |
| CN103405217A (en) * | 2013-07-08 | 2013-11-27 | 上海昭鸣投资管理有限责任公司 | System and method for multi-dimensional measurement of dysarthria based on real-time articulation modeling technology |
| CN203733335U (en) * | 2014-01-23 | 2014-07-23 | 王亚婷 | Chinese phonetic alphabet teaching aid |
| CN104598758A (en) * | 2015-02-12 | 2015-05-06 | 上海市徐汇区中心医院 | System and method for evaluating hearing-speech rehabilitation training and curative effect of patients with post-stroke dysarthria |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107123416A (en) | 2017-09-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Perdue | Wisdom & creation: The theology of wisdom literature | |
| Whitney | Language, and the study of language | |
| Khaldūn | The muqaddimah: an introduction to history-abridged edition | |
| Bonfiglio | Inventing the native speaker | |
| Bergeson et al. | Mothers' speech to hearing‐impaired infants and children with cochlear implants | |
| MacWhinney | Language development | |
| Soeng | Trust in mind: The rebellion of Chinese Zen | |
| Haryono et al. | Planting Values through Character Education for Early Childhood | |
| Burgess et al. | A tale of three learners: The cultural dimension of classroom language learning | |
| CN109473007B (en) | English natural spelling teaching method and system combining phonemes with sound side | |
| Grzega | Towards Global English Via Basic Global English (BGE) | |
| Yang | On the obstacles and strategies in English listening teaching | |
| CN107123416A (en) | A kind of method and device for recognizing the potential pronunciation word of infant | |
| Patel | Reading acquisition in India: Models of learning and dyslexia | |
| Napitupulu et al. | Analysis of Language Acquisition in Children Aged 0-5 Years in the Care of Career Mother: Behaviorism Theory-BF Skinner | |
| Iseli | Deaf Ni-Vanuatu and their signs: A sociolinguistic study | |
| CN202615660U (en) | A learning device for fast memorizing English words | |
| Lendo et al. | An analysis of readability of reading texts on English national examination on junior high school | |
| Berry et al. | Elicited imitation and production of language by severely subnormal children | |
| Schackow | A grammar of Yakkha | |
| Sari | Language acquisition 3-Year-Old Children Through Visual Audio Learning Media During Covid-19 Pandemics | |
| Catalán | The Artisan Poetry of the Romancero1 | |
| Anastasiow | Oral Language Development, From Nonverbal Communication to Speech | |
| Purba et al. | An Analysis Of Language Acquisition On Stunting Children Aged 4-5 Years In Pekan Labuhan Village | |
| Szpotowicz | Second language learning processes in lower primary children. Vocabulary acquisition |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |