CN107481582A - A kind of vocality study electron assistant articulatory system - Google Patents
A kind of vocality study electron assistant articulatory system Download PDFInfo
- Publication number
- CN107481582A CN107481582A CN201710768301.2A CN201710768301A CN107481582A CN 107481582 A CN107481582 A CN 107481582A CN 201710768301 A CN201710768301 A CN 201710768301A CN 107481582 A CN107481582 A CN 107481582A
- Authority
- CN
- China
- Prior art keywords
- module
- audio
- frequency
- beat
- data
- 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
Classifications
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B15/00—Teaching music
- G09B15/02—Boards or like means for providing an indication of notes
- G09B15/023—Electrically operated
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Educational Administration (AREA)
- Educational Technology (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Auxiliary Devices For Music (AREA)
Abstract
The invention discloses a kind of vocality study electron assistant articulatory system, it is related to musicology teaching field.The vocality study electron assistant articulatory system includes：Audio collection module, first buffer module, reference audio memory module, second buffer module, first time-frequency convert module, second time-frequency convert module, first beat extraction module, second beat extraction module, audio similarity detection module, tone color contrast module, beat contrast module, comprehensive assessment module, input module, control module, display module and module for reading and writing, the collection of the data in whole pronunciation situation detection process can be completed by the present invention, conversion and assessment, and assess from accuracy in pitch, tone color and the aspect of beat three are assessed, so that student grasps oneself when practising vocal music skill in time, the quality of oneself tone color quality, whether beat aligns, whether accuracy in pitch is accurate, and then student is caused pointedly to carry out vocal music exercise, so as to improve the effect of exercise.
Description
Technical field
The present invention relates to musicology teaching field, more particularly relates to a kind of vocality study electron assistant articulatory system.
Background technology
Vocal music includes：Bel canto, the folk styled singing and popular singing style, it is now Chinese aboriginal music vocals occur again.Vocal music
The musical form performed by one or more singers, using musical accompaniment whether all can, and voice is the emphasis of works.For
For vocal music, it may is that a kind of most ancient musical form, because not needing any musical instrument, only voice.Various
All there is some form of vocal music in music culture.Though language and music purpose are different, all have common pitch and rhythm because
Element.
Vocal music exercise in, breathing be vocal music exercise important component, be vocal music basis and power, musical respiration
The needs of vocals are submitted to, are separated according to the requirement of melody strong, weak, anxious, slow, make breath during performance fuller, more
It is happy, song is had more expressive force.
But in existing vocal music exercise equipment, although can be that practitioner beat beat, appearance etc., nothing are sung in recording, check and correction
Method allows student to understand oneself when practising vocal music skill, and whether the quality of oneself tone color quality, beat align, and whether accuracy in pitch is accurate
Really, and then cause student can not pointedly carry out vocal music exercise, cause the effect of exercise poor.
The content of the invention
The embodiment of the present invention provides a kind of vocality study electron assistant articulatory system, to solve not allowing in the prior art
Student understands oneself when practising vocal music skill, and whether the quality of oneself tone color quality, beat align, and whether accuracy in pitch is accurate, enters
And cause student can not pointedly carry out vocal music exercise, cause exercise effect it is poor the problem of.
The embodiment of the present invention provides a kind of vocality study electron assistant articulatory system, including：Audio collection module, first are delayed
Die block, reference audio memory module, the second buffer module, the first time-frequency convert module, the second time-frequency convert module, first segment
Clap extraction module, the second beat extraction module, audio similarity detection module, tone color contrast module, beat contrast module, synthesis
Evaluation module, input module, control module, display module and module for reading and writing；
The input module, for inputting read write command, acquisition instructions, the first conversion instruction, the first extraction instruction, second
Conversion instruction, the second extraction instruct, tone color contrast instructs, beat contrasts instruction, detect instruction, assessment instructs and idsplay order；
The control module, for sending read write command to the module for reading and writing；Send and adopt to the audio collection module
Collection instruction；The first conversion instruction is sent to the first time-frequency convert module；First is sent to the first beat extraction module
Extraction instruction；The second conversion instruction is sent to the second time-frequency convert module；The is sent to the second beat extraction module
Two extraction instructions；Detection instruction is sent to the audio similarity detection module；Beat pair is sent to the beat contrast module
Than instruction；Tone color contrast instruction is sent to the tone color contrast module；Sent to the comprehensive assessment module and assess instruction；To institute
State display module and send idsplay order；
The audio collection module, after receiving acquisition instructions, gather the sounding audio-frequency signal of practitioner, and by sound
Frequency signal quantifies to be converted into audio time domain data by noise reduction post-sampling, and by the buffer module of audio time domain data storage first；
The reference audio memory module, the reference audio time domain data practised for storing practitioner；
The module for reading and writing, after receiving read write command, when reading the reference audio that current exercise person is practised
Numeric field data, and the reference audio time domain data is stored in the second buffer module；
The first time-frequency convert module, after receiving the first conversion instruction, transfer the sound in the first buffer module
Frequency time domain data, and the audio time domain data are changed into audio frequency domain data by Fast Fourier Transform (FFT)；
The second time-frequency convert module, after receiving the second conversion instruction, transfer the ginseng in the second buffer module
Audio time domain data are examined, and the reference audio time domain data is changed into reference audio frequency domain number by Fast Fourier Transform (FFT)
According to；
The first beat extraction module, after receiving the first extraction instruction, transfer the sound in the first buffer module
Frequency time domain data, and audio time domain data are subjected to Gassian low-pass filter and obtain the envelope data of audio time domain data, and pass through
The peak value of detected envelope data determines beat sequence；
The second beat extraction module, after receiving the second extraction instruction, transfer the ginseng in the second buffer module
Audio time domain data are examined, and reference audio time domain data is carried out to the reference of Gassian low-pass filter acquisition reference audio time domain data
Envelope data, and determine to refer to beat sequence by detecting the peak value of reference envelope data；
The audio similarity detection module, after receiving detection instruction, by audio time domain data and reference audio
Time domain data is normalized respectively, obtains normalizing audio time domain data and normalizes reference audio time domain data, and
Normalization audio time domain data are passed through into MFCC algorithms with normalization reference audio time domain data respectively and obtain normalization characteristic sequence
Row and normalization fixed reference feature sequence, and calculate normalization characteristic sequence and normalize the likelihood ratio of fixed reference feature sequence；
The tone color contrast module, after receiving tone color contrast instruction, by audio frequency domain data and reference audio frequency
Numeric field data is normalized respectively, obtains normalizing audio frequency domain data and normalizes reference audio frequency domain data；For
Will normalization audio frequency domain data carry out respectively low frequency filtering, intermediate frequency filtering and High frequency filter obtain the frequent numeric field data of bass,
Middle frequency domain audio frequency domain data and high-frequency audio frequency domain data；Will normalization reference audio frequency domain data carry out respectively low frequency filtering,
Intermediate frequency filtering and High frequency filter obtain with reference to the frequent numeric field data of bass, the frequency domain audio frequency domain data and with reference to high-frequency audio with reference in
Frequency domain data；The frequent numeric field data of bass, middle frequency domain audio frequency domain data and high-frequency audio frequency domain data are carried out integrating respectively
To low frequency audio power, intermediate frequency audio power and high-frequency audio energy；The frequent numeric field data of bass, the frequency domain audio with reference in will be referred to
Frequency domain data and reference high-frequency audio frequency domain data are integrated to obtain with reference to low frequency audio power, the frequency domain audio energy with reference in respectively
Measure and refer to high-frequency audio energy；For calculating bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value；
The beat contrast module, after receiving beat contrast instruction, calculate beat sequence and with reference to beat sequence
Likelihood ratio；
The comprehensive assessment module, after receiving assessment instruction, referred to according to normalization characteristic sequence and normalization
The likelihood ratio of characteristic sequence determines accuracy in pitch similarity；Determine that beat is accurate according to beat sequence and with reference to the likelihood ratio of beat sequence
Degree；Bass tones quality, middle pitch are determined according to bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value respectively
Tone color quality and high pitch tone color quality.
Preferably, the display module, after receiving control instruction, display accuracy in pitch similarity, the beat degree of accuracy and
Bass tones quality, middle pitch tone color quality and high pitch tone color quality.
Preferably, the comprehensive assessment module is according to normalization characteristic sequence and the likelihood ratio of normalization fixed reference feature sequence
Determine that accuracy in pitch similarity includes：
The comprehensive assessment module is specified in data area from multiple first, obtains normalization characteristic sequence and normalization is joined
First where examining the likelihood ratio of characteristic sequence specifies data area, and first based on acquisition specifies data area, from storage
In corresponding relation between data area and accuracy in pitch similarity, accuracy in pitch similarity corresponding to acquisition.
Preferably, the comprehensive assessment module determines that beat is accurate according to beat sequence and with reference to the likelihood ratio of beat sequence
Degree includes：
The comprehensive assessment module is specified in data area from multiple second, obtains beat sequence and with reference to beat sequence
Second where likelihood ratio specifies data area, and second based on acquisition specifies data area, from the data area and section of storage
Clap in the corresponding relation between the degree of accuracy, the beat degree of accuracy corresponding to acquisition.
Preferably, the comprehensive assessment module determines that bass tones quality includes according to bass frequency difference value：
The comprehensive assessment module is specified in data area from multiple three, obtains the 3rd where bass frequency difference value
Data area is specified, the 3rd based on acquisition specifies data area, between the data area and bass tones quality of storage
In corresponding relation, bass tones quality corresponding to acquisition.
The embodiment of the present invention, there is provided a kind of vocality study electron assistant articulatory system includes：Audio collection module, first
Buffer module, reference audio memory module, the second buffer module, the first time-frequency convert module, the second time-frequency convert module, first
It is beat extraction module, the second beat extraction module, audio similarity detection module, tone color contrast module, beat contrast module, comprehensive
Evaluation module, input module, control module, display module and module for reading and writing are closed, can complete entirely to send out by the system of the present invention
Assessment in sound situation detection process, and assess and assessed in terms of accuracy in pitch, tone color and beat three so that student grasps in time
, when practising vocal music skill, whether the quality of oneself tone color quality, beat align, and whether accuracy in pitch is accurate for oneself, and then to learn
Life can pointedly carry out vocal music exercise, so as to improve the effect of exercise.And invention is by calculating beat sequence and reference node
The likelihood ratio of sequence is clapped, bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value is calculated and calculates and return
The mode of the likelihood ratio of one change characteristic sequence and normalization fixed reference feature sequence carries out the assessment of beat, tone color and accuracy in pitch, assesses
Accuracy rate is high.
Brief description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
There is the required accompanying drawing used in technology description to be briefly described, it should be apparent that, drawings in the following description are only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, can be with
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is a kind of block diagram of vocality study electron assistant articulatory system provided in an embodiment of the present invention.
Description of reference numerals：
1st, audio collection module；2nd, the first buffer module；3rd, reference audio memory module；4th, the second buffer module；5th,
One time-frequency modular converter；6th, the second time-frequency convert module；7th, the first beat extraction module；8th, the second beat extraction module；9th, sound
Frequency similarity detection module；10th, tone color contrast module；11st, beat contrast module；12nd, comprehensive assessment module；13rd, input module；
14th, control module；15th, module for reading and writing；16th, display module.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other under the premise of creative work is not made
Embodiment, belong to the scope of protection of the invention.
The frame for showing a kind of vocality study electron assistant articulatory system provided in an embodiment of the present invention exemplary Fig. 1
Figure, the vocality study electron assistant articulatory system include audio collection module 1, the first buffer module 2, reference audio memory module
3rd, the second buffer module 4, the first time-frequency convert module 5, the second time-frequency convert module 6, the first beat extraction module 7, second section
Clap extraction module 8, audio similarity detection module 9, tone color contrast module 10, beat contrast module 11, comprehensive assessment module 12,
Input module 13, display module 16, control module 14 and module for reading and writing 15.
Specifically, the input module 13, for input read write command, acquisition instructions, the first conversion instruction, first extraction refer to
Make, the second conversion instruction, second extraction instruction, tone color contrast instruction, beat contrast instruction, detection instruction and assess instruction；Should
Control module 14, for sending read write command to the module for reading and writing；Acquisition instructions are sent to the audio collection module；To this first
Time-frequency convert module sends the first conversion instruction；The first extraction instruction is sent to the first beat extraction module；To this second when
Frequency modular converter sends the second conversion instruction；The second extraction instruction is sent to the second beat extraction module；It is similar to the audio
Spend detection module and send detection instruction；Beat contrast instruction is sent to the beat contrast module；Sent to the tone color contrast module
Tone color contrast instruction；Sent to the comprehensive assessment module and assess instruction.
Specifically, the audio collection module 1, after receiving acquisition instructions, the sounding audio-frequency signal of practitioner is gathered,
And audio signal is converted into audio time domain data by the quantization of noise reduction post-sampling, and audio time domain data storage first is buffered
Module 2.
The reference audio memory module 3, the reference audio time domain data practised for storing practitioner.
The module for reading and writing 15, after receiving read write command, when reading the reference audio that current exercise person is practised
Numeric field data, and the reference audio time domain data is stored in the second buffer module 4.
Specifically, the first time-frequency convert module 5, after receiving the first conversion instruction, the first buffer module 2 is transferred
In audio time domain data, and the audio time domain data are changed into audio frequency domain data by Fast Fourier Transform (FFT)；This
Two time-frequency convert modules 6, after receiving the second conversion instruction, transfer the reference audio time domain number in the second buffer module 4
According to, and the reference audio time domain data is changed into reference audio frequency domain data by Fast Fourier Transform (FFT).
Specifically, the first beat extraction module 7, after receiving the first extraction instruction, the first buffer module 2 is transferred
In audio time domain data, and by audio time domain data carry out Gassian low-pass filter obtain audio time domain data envelope data,
And beat sequence is determined by the peak value of detected envelope data.
Specifically, the second beat extraction module 8, after receiving the second extraction instruction, the second buffer module 4 is transferred
In reference audio time domain data, and by reference audio time domain data carry out Gassian low-pass filter obtain reference audio time domain data
Reference envelope data, and by detect reference envelope data peak value determine refer to beat sequence.
Wherein, the first beat extraction module 7 determines that beat sequence is specially by the peak value of detected envelope data：First
Beat extraction module 7 determining that the numerical value in beat sequence is 1, when can't detect peak value, determining beat when detecting peak value
Numerical value in sequence is -1, that is, beat sequence is by a string 1 and -1 sequence formed.
Further, control module 14 sends detection instruction, audio similarity inspection to the audio similarity detection module 9
Module is surveyed, after receiving detection instruction, first transfers audio time domain data from the first buffer module 2, from the second buffering mould
Reference audio time domain data is transferred in block 4, then audio time domain data and reference audio time domain data are normalized respectively
Processing, obtain normalize audio time domain data and normalize reference audio time domain data, and will normalization audio time domain data and
Normalize reference audio time domain data and normalization characteristic sequence and normalization fixed reference feature sequence obtained by MFCC algorithms respectively,
And calculate normalization characteristic sequence and normalize the likelihood ratio of fixed reference feature sequence.
Specifically, tone color contrast module 10, after receiving tone color contrast instruction, by audio frequency domain data and reference
Audio frequency domain data is normalized respectively, obtains normalizing audio frequency domain data and normalizes reference audio frequency domain number
According to；For carrying out low frequency filtering, intermediate frequency filtering and High frequency filter respectively to obtain bass frequent normalization audio frequency domain data
Numeric field data, middle frequency domain audio frequency domain data and high-frequency audio frequency domain data；Normalization reference audio frequency domain data is carried out respectively low
Frequency filtering, intermediate frequency filtering and High frequency filter are obtained with reference to the frequent numeric field data of bass, frequency domain audio frequency domain data and reference with reference in
High-frequency audio frequency domain data；The frequent numeric field data of bass, middle frequency domain audio frequency domain data and high-frequency audio frequency domain data are entered respectively
Row integration obtains low frequency audio power, intermediate frequency audio power and high-frequency audio energy；The frequent numeric field data of bass, reference will be referred to
Middle frequency domain audio frequency domain data and reference high-frequency audio frequency domain data are integrated to obtain with reference to low frequency audio power, with reference in respectively
Frequency domain audio energy and reference high-frequency audio energy；For calculating bass frequency difference value, middle frequency domain audio difference value and high frequency audio
Frequency difference value.
Wherein, tone color contrast module 10 calculates bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value
Including：Bass frequency difference value is determined by low frequency audio power and the ratio with reference to low frequency audio power；Pass through middle frequency domain audio
Frequency domain audio difference value during energy determines with the ratio with reference to intermediate frequency audio power；By high-frequency audio energy and refer to high-frequency audio
The ratio of energy determines high-frequency audio difference value.
In addition, will normalization audio frequency domain data it is used when carrying out low frequency filtering, intermediate frequency filtering and High frequency filter respectively
Wave filter is respectively：Gaussian filter, Butterworth filter and Chebyshev filter.
Furthermore the scope in the frequent domain of bass is 40Hz~160Hz, the scope of middle frequency domain audio frequency domain for 160Hz~
1280Hz；The scope of high-frequency audio frequency domain is 1280Hz~5120Hz.
Specifically, the beat contrast module 11, after receiving beat contrast instruction, beat sequence and reference node are calculated
Clap the likelihood ratio of sequence.
Specifically, comprehensive assessment module 12, after receiving assessment instruction, according to normalization characteristic sequence and normalizing
The likelihood ratio for changing fixed reference feature sequence determines accuracy in pitch similarity；Section is determined according to beat sequence and with reference to the likelihood ratio of beat sequence
Clap the degree of accuracy；Bass tones matter is determined according to bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value respectively
Amount, middle pitch tone color quality and high pitch tone color quality.
Wherein, the comprehensive assessment module 12 is true according to the likelihood ratio of normalization characteristic sequence and normalization fixed reference feature sequence
The quasi- similarity of accordatura includes：
The comprehensive assessment module 12 is specified in data area from multiple first, obtains normalization characteristic sequence and normalization is joined
First where examining the likelihood ratio of characteristic sequence specifies data area, and first based on acquisition specifies data area, from storage
In corresponding relation between data area and accuracy in pitch similarity, accuracy in pitch similarity corresponding to acquisition.
In addition, the plurality of first specifies data area to set in advance, the data model stored in comprehensive assessment module 12
Enclose the corresponding relation between accuracy in pitch similarity to set in advance, and during setting, this can be passed through by user
The control interface of input module 13 is configured, and is commented it is of course also possible to be sent to the synthesis after being configured by user terminal
Estimate module 12, the embodiment of the present invention is not specifically limited to this.
Such as the likelihood of the normalization characteristic sequence that gets of the comprehensive assessment module 12 and normalization fixed reference feature sequence
Than for 0.7, the plurality of first to specify data area be the comprehensive assessment module 12 between the data area and accuracy in pitch similarity
Corresponding relation in acquire, and the corresponding relation between the data area and accuracy in pitch similarity can with as shown in table 1 below,
Be, the plurality of first specify data area can be 0~0.6,0.6~0.8, more than 0.8.Obtained from such as table 1 below corresponding
Accuracy in pitch similarity to be substantially similar.
Table 1
| Accuracy in pitch similarity | It is closely similar | Substantially it is similar | It is dissimilar |
| Data area | More than 0.8 | 0.6~0.8 | 0~0.6 |
It should be noted that in embodiments of the present invention, only with the data area shown in above-mentioned table 1 and accuracy in pitch similarity it
Between corresponding relation exemplified by illustrate, above-mentioned table 1 not to the embodiment of the present invention form limit.
Further, the comprehensive assessment module determines that beat is accurate according to beat sequence and with reference to the likelihood ratio of beat sequence
Degree includes：
The comprehensive assessment module is specified in data area from multiple second, obtains beat sequence and reference beat sequence seemingly
So second where ratio specifies data area, and second based on acquisition specifies data area, from the data area and beat of storage
In corresponding relation between the degree of accuracy, the beat degree of accuracy corresponding to acquisition.
Wherein, the plurality of second data area is specified to set in advance, the data model stored in comprehensive assessment module 12
Enclose the corresponding relation between the beat degree of accuracy to set in advance, and during setting, this can be passed through by user
The control interface of input module 13 is configured, and is commented it is of course also possible to be sent to the synthesis after being configured by user terminal
Estimate module 12, the embodiment of the present invention is not specifically limited to this.
Such as the beat sequence that gets of the comprehensive assessment module 12 and be 0.5 with reference to the likelihood ratio of beat sequence, this is more
Individual second specifies data area as in corresponding relation of the comprehensive assessment module 12 between the data area and the beat degree of accuracy
Acquire, and the corresponding relation between the data area and the beat degree of accuracy can be that is to say with as shown in table 1 below, it is the plurality of
Second specify data area can be 0~0.6,0.6~0.8, more than 0.8.The beat degree of accuracy corresponding to being obtained from such as table 2 below
For inaccuracy.
Table 2
| Accuracy in pitch similarity | It is very accurate | Substantially it is accurate | It is inaccurate |
| Data area | More than 0.8 | 0.6~0.8 | 0~0.6 |
It should be noted that in embodiments of the present invention, only with the data area shown in above-mentioned table 2 and the beat degree of accuracy it
Between corresponding relation exemplified by illustrate, above-mentioned table 1 not to the embodiment of the present invention form limit.
Specifically, the comprehensive assessment module determines that bass tones quality includes according to bass frequency difference value：
The comprehensive assessment module is specified in data area from multiple three, and the 3rd where obtaining bass frequency difference value refers to
Determine data area, the 3rd based on acquisition specifies data area, pair between the data area and bass tones quality of storage
In should being related to, bass tones quality corresponding to acquisition.
Wherein, the plurality of 3rd data area is specified to set in advance, the data model stored in comprehensive assessment module 12
Enclose the corresponding relation between bass tones quality to set in advance, and during setting, can be passed through by user
The control interface of the input module 13 is configured, it is of course also possible to be sent to the synthesis after being configured by user terminal
Evaluation module 12, the embodiment of the present invention is not specifically limited to this.
For example the bass frequency difference value that the comprehensive assessment module 12 is got is 1, the plurality of 3rd specifies data area
To be acquired in corresponding relation of the comprehensive assessment module 12 between the data area and bass tones quality, and the data
Corresponding relation between scope and bass tones quality can be that is to say, the plurality of 3rd specifies data area with as shown in table 3 below
Can be 0~0.8,0.8~1.5, more than 1.5.Bass tones quality corresponding to being obtained from such as table 3 below is that bass is balanced.
Table 3
It should be noted that in embodiments of the present invention, only with the data area shown in above-mentioned table 3 and bass tones quality
Between corresponding relation exemplified by illustrate, above-mentioned table 1 not to the embodiment of the present invention form limit.
In addition, comprehensive assessment module frequency domain audio difference value in determines middle pitch tone color quality, according to high frequency audio frequency difference
Different value determines that the method for high pitch tone color quality determines that the method for bass tones quality is similar with according to bass frequency difference value, this hair
This is no longer going to repeat them for bright embodiment.
Specifically, display module 16, after the control instruction for receiving control module transmission, display accuracy in pitch similarity,
The beat degree of accuracy and bass tones quality, middle pitch tone color quality and high pitch tone color quality.
The embodiment of the present invention, there is provided a kind of vocality study electron assistant articulatory system includes：Audio collection module, first
Buffer module, reference audio memory module, the second buffer module, the first time-frequency convert module, the second time-frequency convert module, first
It is beat extraction module, the second beat extraction module, audio similarity detection module, tone color contrast module, beat contrast module, comprehensive
Evaluation module, input module, control module, display module and module for reading and writing are closed, can be completed in the system by the present invention whole
Assessment in pronunciation situation detection process, and assess and assessed in terms of accuracy in pitch, tone color and beat three so that student slaps in time
Oneself is held when practising vocal music skill, whether the quality of oneself tone color quality, beat align, and whether accuracy in pitch is accurate, and then cause
Student can pointedly carry out vocal music exercise, so as to improve the effect of exercise.And invention is by calculating beat sequence and reference
The likelihood ratio of beat sequence, calculate bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value and calculating
The mode of the likelihood ratio of normalization characteristic sequence and normalization fixed reference feature sequence carries out the assessment of beat, tone color and accuracy in pitch, comments
Estimate accuracy rate height.
Although preferred embodiments of the present invention have been described, but those skilled in the art once know basic creation
Property concept, then can make other change and modification to these embodiments.So appended claims be intended to be construed to include it is excellent
Select embodiment and fall into having altered and changing for the scope of the invention.
Obviously, those skilled in the art can carry out the essence of various changes and modification without departing from the present invention to the present invention
God and scope.So, if these modifications and variations of the present invention belong to the scope of the claims in the present invention and its equivalent technologies
Within, then the present invention is also intended to comprising including these changes and modification.
Claims (5)
- A kind of 1. vocality study electron assistant articulatory system, it is characterised in that including：Audio collection module, the first buffer module, Reference audio memory module, the second buffer module, the first time-frequency convert module, the second time-frequency convert module, the extraction of the first beat Module, the second beat extraction module, audio similarity detection module, tone color contrast module, beat contrast module, comprehensive assessment mould Block, input module, control module, display module and module for reading and writing；The input module, for inputting read write command, acquisition instructions, the first conversion instruction, the first extraction instruction, the second conversion Instruct, the second extraction instructs, tone color contrast instructs, beat contrasts instruction, detect instruction, assessment instructs and idsplay order；The control module, for sending read write command to the module for reading and writing；Collection is sent to the audio collection module to refer to Order；The first conversion instruction is sent to the first time-frequency convert module；The first extraction is sent to the first beat extraction module Instruction；The second conversion instruction is sent to the second time-frequency convert module；Second is sent to the second beat extraction module to carry Instruction fetch；Detection instruction is sent to the audio similarity detection module；Beat contrast is sent to the beat contrast module to refer to Order；Tone color contrast instruction is sent to the tone color contrast module；Sent to the comprehensive assessment module and assess instruction；Show to described Show that module sends idsplay order；The audio collection module, after receiving acquisition instructions, the sounding audio-frequency signal of practitioner is gathered, and audio is believed Number quantify to be converted into audio time domain data by noise reduction post-sampling, and by the buffer module of audio time domain data storage first；The reference audio memory module, the reference audio time domain data practised for storing practitioner；The module for reading and writing, after receiving read write command, read the reference audio time domain number that current exercise person is practised According to, and the reference audio time domain data is stored in the second buffer module；The first time-frequency convert module, after receiving the first conversion instruction, when transferring the audio in the first buffer module Numeric field data, and the audio time domain data are changed into audio frequency domain data by Fast Fourier Transform (FFT)；The second time-frequency convert module, after receiving the second conversion instruction, transfer the reference sound in the second buffer module Frequency time domain data, and the reference audio time domain data is changed into reference audio frequency domain data by Fast Fourier Transform (FFT)；The first beat extraction module, after receiving the first extraction instruction, when transferring the audio in the first buffer module Numeric field data, and audio time domain data are subjected to Gassian low-pass filter and obtain the envelope data of audio time domain data, and pass through detection The peak value of envelope data determines beat sequence；The second beat extraction module, after receiving the second extraction instruction, transfer the reference sound in the second buffer module Frequency time domain data, and by reference audio time domain data carry out Gassian low-pass filter obtain reference audio time domain data reference envelope Data, and determine to refer to beat sequence by detecting the peak value of reference envelope data；The audio similarity detection module, after receiving detection instruction, by audio time domain data and reference audio time domain Data are normalized respectively, obtain normalizing audio time domain data and normalize reference audio time domain data, and will return One change audio time domain data and normalization reference audio time domain data respectively by MFCC algorithms obtain normalization characteristic sequence and Fixed reference feature sequence is normalized, and calculates normalization characteristic sequence and normalizes the likelihood ratio of fixed reference feature sequence；The tone color contrast module, after receiving tone color contrast instruction, by audio frequency domain data and reference audio frequency domain number According to being normalized respectively, obtain normalizing audio frequency domain data and normalize reference audio frequency domain data；For that will return One change audio frequency domain data carries out low frequency filtering, intermediate frequency filtering and High frequency filter and obtains the frequent numeric field data of bass, intermediate frequency respectively Audio frequency domain data and high-frequency audio frequency domain data；Normalization reference audio frequency domain data is subjected to low frequency filtering, intermediate frequency respectively Filtering and High frequency filter obtain with reference to the frequent numeric field data of bass, with reference in frequency domain audio frequency domain data and refer to high-frequency audio frequency domain Data；The frequent numeric field data of bass, middle frequency domain audio frequency domain data and high-frequency audio frequency domain data are integrated to obtain respectively low Frequency domain audio energy, intermediate frequency audio power and high-frequency audio energy；The frequent numeric field data of bass, the frequency domain audio frequency domain with reference in will be referred to Data and integrated to obtain respectively with reference to low frequency audio power with reference to high-frequency audio frequency domain data, with reference to intermediate frequency audio power and With reference to high-frequency audio energy；For calculating bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value；The beat contrast module, after receiving beat contrast instruction, calculate beat sequence and reference beat sequence seemingly So ratio；The comprehensive assessment module, after receiving assessment instruction, according to normalization characteristic sequence and normalization fixed reference feature The likelihood ratio of sequence determines accuracy in pitch similarity；The beat degree of accuracy is determined according to beat sequence and with reference to the likelihood ratio of beat sequence； Bass tones quality, middle pitch sound are determined according to bass frequency difference value, middle frequency domain audio difference value and high-frequency audio difference value respectively Chromaticness amount and high pitch tone color quality.
- 2. vocality study electron assistant articulatory system as claimed in claim 1, it is characterised in that the comprehensive assessment module root Determine that accuracy in pitch similarity includes according to the likelihood ratio of normalization characteristic sequence and normalization fixed reference feature sequence：The comprehensive assessment module is specified in data area from multiple first, obtains normalization characteristic sequence and normalization with reference to special First where levying the likelihood ratio of sequence specifies data area, and first based on acquisition specifies data area, from the data of storage In corresponding relation between scope and accuracy in pitch similarity, accuracy in pitch similarity corresponding to acquisition.
- 3. vocality study electron assistant articulatory system as claimed in claim 1, it is characterised in that the comprehensive assessment module root Determine that the beat degree of accuracy includes according to beat sequence and with reference to the likelihood ratio of beat sequence：The comprehensive assessment module is specified in data area from multiple second, obtains beat sequence and the likelihood with reference to beat sequence Second than place specifies data area, and second based on acquisition specifies data area, accurate from data area and the beat of storage In corresponding relation between exactness, the beat degree of accuracy corresponding to acquisition.
- 4. vocality study electron assistant articulatory system as claimed in claim 1, it is characterised in that the comprehensive assessment module root Determine that bass tones quality includes according to bass frequency difference value：The comprehensive assessment module is specified in data area from multiple three, and the 3rd where obtaining bass frequency difference value specifies Data area, the based on acquisition the 3rd specifies data area, from corresponding between the data area and bass tones quality of storage In relation, bass tones quality corresponding to acquisition.
- 5. vocality study electron assistant articulatory system as claimed in claim 1, it is characterised in that the display module, be used for After receiving control instruction, display accuracy in pitch similarity, the beat degree of accuracy and bass tones quality, middle pitch tone color quality and high pitch sound Chromaticness amount.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710768301.2A CN107481582A (en) | 2017-08-31 | 2017-08-31 | A kind of vocality study electron assistant articulatory system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710768301.2A CN107481582A (en) | 2017-08-31 | 2017-08-31 | A kind of vocality study electron assistant articulatory system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107481582A true CN107481582A (en) | 2017-12-15 |
Family
ID=60604140
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710768301.2A Pending CN107481582A (en) | 2017-08-31 | 2017-08-31 | A kind of vocality study electron assistant articulatory system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107481582A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108428458A (en) * | 2018-03-15 | 2018-08-21 | 河南科技学院 | A kind of vocality study electron assistant articulatory system |
| CN109686376A (en) * | 2019-01-08 | 2019-04-26 | 北京雷石天地电子技术有限公司 | A kind of singing songs evaluation method and system |
| CN111710202A (en) * | 2020-07-07 | 2020-09-25 | 淮南师范学院 | Real-time interaction device for vocal music training |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106097828A (en) * | 2016-08-12 | 2016-11-09 | 淮阴师范学院 | Technical ability detecting system is sung in a kind of musicology teaching |
| CN106228996A (en) * | 2016-07-15 | 2016-12-14 | 黄河科技学院 | Vocality study electron assistant articulatory system |
| CN106448701A (en) * | 2016-08-30 | 2017-02-22 | 苏娜 | Vocal integrated training system |
| CN106898363A (en) * | 2017-02-27 | 2017-06-27 | 河南职业技术学院 | A kind of vocality study electron assistant articulatory system |
-
2017
- 2017-08-31 CN CN201710768301.2A patent/CN107481582A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228996A (en) * | 2016-07-15 | 2016-12-14 | 黄河科技学院 | Vocality study electron assistant articulatory system |
| CN106097828A (en) * | 2016-08-12 | 2016-11-09 | 淮阴师范学院 | Technical ability detecting system is sung in a kind of musicology teaching |
| CN106448701A (en) * | 2016-08-30 | 2017-02-22 | 苏娜 | Vocal integrated training system |
| CN106898363A (en) * | 2017-02-27 | 2017-06-27 | 河南职业技术学院 | A kind of vocality study electron assistant articulatory system |
Non-Patent Citations (3)
| Title |
|---|
| 刘建: "特征参数模式比对在语音评分中的应用研究", 《中南大学硕士学位论文》 * |
| 王佳迪: "鲁棒的音乐评分方法研究", 《电子科技大学硕士论文》 * |
| 王磊，黄申等: "基于动态规划的清唱评分与纠错系统设计与实现", 《第九届全国人机语音通讯学术会议论文集 》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108428458A (en) * | 2018-03-15 | 2018-08-21 | 河南科技学院 | A kind of vocality study electron assistant articulatory system |
| CN109686376A (en) * | 2019-01-08 | 2019-04-26 | 北京雷石天地电子技术有限公司 | A kind of singing songs evaluation method and system |
| CN109686376B (en) * | 2019-01-08 | 2020-06-30 | 北京雷石天地电子技术有限公司 | Song singing evaluation method and system |
| CN111710202A (en) * | 2020-07-07 | 2020-09-25 | 淮南师范学院 | Real-time interaction device for vocal music training |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109448754B (en) | Multidimensional singing scoring system | |
| CN104272382B (en) | Personalized singing synthetic method based on template and system | |
| CN110880329B (en) | Audio identification method and equipment and storage medium | |
| CN109215632A (en) | A kind of speech evaluating method, device, equipment and readable storage medium storing program for executing | |
| CN106997765B (en) | Quantitative characterization method for human voice timbre | |
| CN107316638A (en) | A kind of poem recites evaluating method and system, a kind of terminal and storage medium | |
| CN106935248A (en) | A kind of voice similarity detection method and device | |
| CN102723079B (en) | Music and chord automatic identification method based on sparse representation | |
| CN101923855A (en) | Test-irrelevant voice print identifying system | |
| CN102426834B (en) | Method for testing rhythm level of spoken English | |
| CN104766603A (en) | Method and device for building personalized singing style spectrum synthesis model | |
| CN101023469A (en) | Digital filtering method, digital filtering equipment | |
| CN108206027A (en) | A kind of audio quality evaluation method and system | |
| CN105825852A (en) | Oral English reading test scoring method | |
| CN107481582A (en) | A kind of vocality study electron assistant articulatory system | |
| CN110599987A (en) | Piano note recognition algorithm based on convolutional neural network | |
| CN109979488A (en) | Voice based on stress analysis turns music notation system | |
| CN111433847A (en) | Speech conversion method and training method, intelligent device and storage medium | |
| WO2020140390A1 (en) | Vibrato modeling method, device, computer apparatus and storage medium | |
| CN106097829B (en) | A kind of vocal music exercising auxiliary device | |
| CN107978322A (en) | A kind of K songs marking algorithm | |
| TWI742486B (en) | Singing assisting system, singing assisting method, and non-transitory computer-readable medium comprising instructions for executing the same | |
| CN105845149A (en) | Predominant pitch acquisition method in acoustical signal and system thereof | |
| CN109452932A (en) | A kind of Constitution Identification method and apparatus based on sound | |
| US20100169085A1 (en) | Model based real time pitch tracking system and singer evaluation method |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171215 |
|
| RJ01 | Rejection of invention patent application after publication |