CN109003613A - The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information - Google Patents
The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information Download PDFInfo
- Publication number
- CN109003613A CN109003613A CN201811017603.7A CN201811017603A CN109003613A CN 109003613 A CN109003613 A CN 109003613A CN 201811017603 A CN201811017603 A CN 201811017603A CN 109003613 A CN109003613 A CN 109003613A
- Authority
- CN
- China
- Prior art keywords
- counterfeit
- application
- voiceprint recognition
- payment information
- model
- 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
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
- G10L17/02—Preprocessing operations, e.g. segment selection; Pattern representation or modelling, e.g. based on linear discriminant analysis [LDA] or principal components; Feature selection or extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q20/00—Payment architectures, schemes or protocols
- G06Q20/38—Payment protocols; Details thereof
- G06Q20/40—Authorisation, e.g. identification of payer or payee, verification of customer or shop credentials; Review and approval of payers, e.g. check credit lines or negative lists
- G06Q20/401—Transaction verification
- G06Q20/4014—Identity check for transactions
- G06Q20/40145—Biometric identity checks
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
- G10L17/04—Training, enrolment or model building
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
- G10L17/16—Hidden Markov models [HMM]
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification
- G10L17/22—Interactive procedures; Man-machine interfaces
Abstract
The invention discloses a kind of Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information, the following steps are included: S1: establishing the phonetic feature model for representing speaker's identity, speaker characteristic is extracted from voice data by programming, using text related data, MFCC model is established in matlab environment, it is read in first using wavread function, then with frame length 256, frame moves 80 and carries out framing, then by sentence x=filter ([1-0.9375], 1, x) realize that the exacerbation to high frequency is handled, filter out low-frequency disturbance, the especially Hz noise of 50 to 60 hz, the high frequency section more useful to speech recognition carries out frequency spectrum promotion, then, in order to keep the short-term stationarity of voice signal.The present invention is able to carry out multiple-authentication, improves payment safety, and verify text using stochastic and dynamic, it is read by artificial speech, then the vocal print comparison that speech verification obtains vocal print and system saves improves safety to avoid the problem that the easily stolen duplication of single vocal print.
Description
Technical field
The present invention relates to sound groove recognition technology in e field more particularly to a kind of Application on Voiceprint Recognition payment informations of combining space information
Method for anti-counterfeit.
Background technique
So-called vocal print is the sound wave spectrum for the carrying verbal information that electricity consumption acoustic instrument is shown.The generation of human language is people
A complicated physiology physical process between body speech center and vocal organs, the phonatory organ that people uses in speech -- tongue,
Tooth, larynx, lung, nasal cavity everyone widely different in terms of size and form, so the voiceprint map of any two people is all
It is variant.The existing relative stability of everyone speech acoustics feature, and have variability, it is not absolute, unalterable.This
Kind variation may be from physiology, pathology, psychology, simulation, camouflage, also related with environmental disturbances.Nevertheless, due to everyone hair
Sound organ is all not quite similar, therefore under normal circumstances, and people remain to distinguish the sound of different people or judge whether it is same
The sound of people.
Vocal print is also the unique individual character biological characteristic of human body, is difficult to find two duplicate people of vocal print, existing branch
The mode of paying is mostly static password validation of payment, due to password invariance, so that the problem of safety is low, and password is easy leakage, existing
Face recognition technology is used, but verification the verifying results are undesirable when light is bad, and to the more demanding of equipment, from forming
This height, and the existing means of payment is single, single verifying is more, so that safety is also low, it is therefore desirable to which a kind of convenience can be with
Multiple-authentication, and solve the above problems to the not high verification method of equipment requirement.
Summary of the invention
It is mostly that static password validation of payment safety is low the purpose of the present invention is to solve the existing means of payment, and people
The shortcomings that face verifying the high requirements on the equipment, and a kind of Application on Voiceprint Recognition payment information method for anti-counterfeit of the combining space information proposed.
To achieve the goals above, present invention employs following technical solutions:
The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information, comprising the following steps:
S1: establishing the phonetic feature model for representing speaker's identity, and speaker spy is extracted from voice data by programming
Sign, using text related data, establishes MFCC model in matlab environment, is read in first using wavread function, then with
Frame length 256, frame move 80 and carry out framing, are then realized to the exacerbation of high frequency by sentence x=filter ([1-0.9375], 1, x)
Reason, filters out the Hz noise of low-frequency disturbance, especially 50 to 60 hz, and the high frequency section more useful to speech recognition carries out frequency
Spectrum is promoted, and then, in order to keep the short-term stationarity of voice signal, is reduced as caused by truncation using Hamming window function
Gibbs effect finally extracts the sampled value in voice signal, frequency and sampling resolution；
S2: speaker model or Application on Voiceprint Recognition model are established using speaker characteristic；
S3: end-point detection vad is carried out；
S4: MFCC parameter algorithm mfcc is carried out；
S5: HMM parameter initialization inithmm is carried out；
S6: viterbit recognizer is carried out；
S7: it is trained process train；
S8: identification main program is carried out.
Preferably, in the S1, phonetic entry is that MFCC model is established in matlab environment, uses wavread first
Function is read in, and framing can be reduced calculation amount but the variation of adjacent interframe less, be easily lost signal characteristic, take frame length 20ms, and frame moves
It is the 1/3~1/2 of frame length, for filtering out low-frequency disturbance, adding window is analyzed the short-time energy of voice, retouched for high-frequency emphasis processing
This changing features situation of predicate sound defines short-time energy are as follows:
Wherein N is that window is long, finally extracts the sampled value in voice signal, frequency and sampling resolution.
Preferably, in the S3, using the algorithm of double threshold, inputting as sampled speech data x, output X1, X2 is starting
The frame number of endpoint and end caps, and the voice data of xi to x2 frame is deposited in sample.wave structural array.
Preferably, it in the S4, inputs as sampled speech data x, exports as mfcc parameter, take 2 frame of x1-2 to x2-
Mfcc parameter is into sample.data structural array.
Preferably, in the S5, inputting as the array M, N of samples and NX*1 is status number, and M is that each state includes
Gaussian Mixture number.
Preferably, in the S6, model and mfcc parameter are deleted in input, recall optimum state path, return output probability and
State path.
Preferably, in the S7, n times iteration is implemented to an iteration function baum.m, the HMM model after exporting as training
Model parameter is stored in hmm { i } by parameter and total output probability.
Preferably, in the S8, end-point detection is carried out with function vad to voice to be identified is inputted, calculates MFCC ginseng
After number, transfers to recognition function viterbi.m that the output probability of its logarithmic form is calculated, finally show recognition result.
Preferably, to the sound bite uploading system of acquisition, and record acoustic wave character is analyzed, for transferring sound in verifying
Wave is compared.
The beneficial effects of the present invention are:
1, by the method for proposition, it is able to carry out multiple-authentication, improves payment safety, and verify using stochastic and dynamic
Text is read by artificial speech, the vocal print comparison that then speech verification obtains vocal print and system saves, to avoid single vocal print
Easily stolen the problem of using duplication, improve safety；
2, by the method for proposition, accuracy can be improved when extracting verifying, guarantees that the verifying sound obtained is more smart
Really, to improve verifying speed, guarantee verification information accuracy and when validity.
Specific embodiment
The following is a clear and complete description of the technical scheme in the embodiments of the invention, it is clear that described embodiment
Only a part of the embodiment of the present invention, instead of all the embodiments.
The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information, comprising the following steps:
S1: establishing the phonetic feature model for representing speaker's identity, and speaker spy is extracted from voice data by programming
Sign, using text related data, establishes MFCC model in matlab environment, is read in first using wavread function, then with
Frame length 256, frame move 80 and carry out framing, are then realized to the exacerbation of high frequency by sentence x=filter ([1-0.9375], 1, x)
Reason, filters out the Hz noise of low-frequency disturbance, especially 50 to 60 hz, and the high frequency section more useful to speech recognition carries out frequency
Spectrum is promoted, and then, in order to keep the short-term stationarity of voice signal, is reduced as caused by truncation using Hamming window function
Gibbs effect finally extracts the sampled value in voice signal, frequency and sampling resolution；
S2: speaker model or Application on Voiceprint Recognition model are established using speaker characteristic；
S3: end-point detection vad is carried out；
S4: MFCC parameter algorithm mfcc is carried out；
S5: HMM parameter initialization inithmm is carried out；
S6: viterbit recognizer is carried out；
S7: it is trained process train；
S8: identification main program is carried out.
In the present embodiment, firstly, user reads one section of random sentence to sound pick-up outfit, equipment carries out voice input, voice
It reads in establish MFCC model in matlab environment, is read in first using wavread function, framing can be reduced calculation amount but phase
Adjacent interframe variation less, is easily lost signal characteristic, takes frame length 20ms, it is the 1/3~1/2 of frame length that frame, which moves, high-frequency emphasis processing
For filtering out low-frequency disturbance, adding window analyzes the short-time energy of voice, describes this changing features situation of voice, definition
Short-time energy are as follows:Wherein N is that window is long, finally extracts and speaks
Sampled value in sound signal, frequency and sampling resolution are inputted as sampled speech data x using the algorithm of double threshold, are exported X1,
X2 is the frame number of starting endpoint and end caps, and the voice data of xi to x2 frame is deposited in sample.wave structural array, defeated
Enter for sampled speech data x, exports as mfcc parameter, take the mfcc parameter of 2 frame of x1-2 to x2-to sample.data structure
In array, inputting as the array M, N of samples and NX*1 is status number, and M is the Gaussian Mixture number that each state includes, defeated
Enter to delete model and mfcc parameter, recall optimum state path, output probability and state path is returned to, to an iteration function
Baum.m implements n times iteration, and model parameter is stored in hmm { i } by HMM model parameter and total output probability after exporting as training,
In the S8, end-point detection is carried out with function vad to voice to be identified is inputted, after calculating MFCC parameter, transfers to identify
The output probability of its logarithmic form is calculated in function viterbi.m, finally shows recognition result, to the sound bite of acquisition
Uploading system, and record acoustic wave character is analyzed, it is compared for transferring sound wave in verifying；
Pop up the verifying sentence of 30 words at random when being verified, in system, user reads verifying language against wheat
Sentence, system carries out Speech processing after receiving voice, vocal print feature is extracted, vocal print modeling, vocal print comparison, differentiates decision,
Finally the sound bite vocal print integrated is compared with the vocal print gene section in library, thus verify whether as my vocal print,
Verifying meets, and passes through, and verifying is not met, and records illegal voice messaging, for being stored in system, gives and transfers verification in person,
To understand illegal user.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Anyone skilled in the art in the technical scope disclosed by the present invention, according to the technique and scheme of the present invention and its
Inventive concept is subject to equivalent substitution or change, should be covered by the protection scope of the present invention.
Claims (9)
1. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information, which comprises the following steps:
S1: establishing the phonetic feature model for representing speaker's identity, extracts speaker characteristic from voice data by programming, adopts
With text related data, MFCC model is established in matlab environment, is read in first using wavread function, then with frame length
256, frame moves 80 and carries out framing, then realizes that the exacerbation to high frequency is handled by sentence x=filter ([1-0.9375], 1, x), filter
Except the Hz noise of low-frequency disturbance, especially 50 to 60 hz, the high frequency section more useful to speech recognition carries out frequency spectrum and mentions
It rises, then, in order to keep the short-term stationarity of voice signal, is reduced as caused by truncation using Hamming window function
Gibbs effect finally extracts the sampled value in voice signal, frequency and sampling resolution；
S2: speaker model or Application on Voiceprint Recognition model are established using speaker characteristic；
S3: end-point detection vad is carried out；
S4: MFCC parameter algorithm mfcc is carried out；
S5: HMM parameter initialization inithmm is carried out；
S6: viterbit recognizer is carried out；
S7: it is trained process train；
S8: identification main program is carried out.
2. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information according to claim 1, which is characterized in that institute
It states in S1, phonetic entry is that MFCC model is established in matlab environment, is read in first using wavread function, framing can subtract
Calculation amount but adjacent interframe change little less, are easily lost signal characteristic, take frame length 20ms, and it is the 1/3~1/2 of frame length that frame, which moves,
For filtering out low-frequency disturbance, adding window analyzes the short-time energy of voice, describes this feature of voice for high-frequency emphasis processing
Situation of change defines short-time energy are as follows:Wherein N is window
It is long, finally extract the sampled value in voice signal, frequency and sampling resolution.
3. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information according to claim 1, which is characterized in that institute
It states in S3, using the algorithm of double threshold, inputs as sampled speech data x, export X1, X2 is the frame of starting endpoint and end caps
Number, and the voice data of xi to x2 frame is deposited in sample.wave structural array.
4. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information according to claim 1, which is characterized in that institute
It states in S4, inputs as sampled speech data x, export as mfcc parameter, the mfcc parameter of 2 frame of x1-2 to x2-is taken to arrive
In sample.data structural array.
5. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information according to claim 1, which is characterized in that institute
It states in S5, inputting as the array M, N of samples and NX*1 is status number, and M is the Gaussian Mixture number that each state includes.
6. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information according to claim 1, which is characterized in that institute
It states in S6, model and mfcc parameter are deleted in input, recall optimum state path, return to output probability and state path.
7. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information according to claim 1, which is characterized in that institute
It states in S7, to an iteration function baum.m implementation n times iteration, HMM model parameter and total output probability after exporting as training,
Model parameter is stored in hmm { i }.
8. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information according to claim 1, which is characterized in that institute
It states in S8, carries out end-point detection with function vad to voice to be identified is inputted, after calculating MFCC parameter, transfer to identification letter
The output probability of its logarithmic form is calculated in number viterbi.m, finally shows recognition result.
9. the Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information described in any one according to claim 1~8,
It is characterized in that, to the sound bite uploading system of acquisition, and analyzes record acoustic wave character, carried out for transferring sound wave in verifying
It compares.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811017603.7A CN109003613A (en) | 2018-09-02 | 2018-09-02 | The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811017603.7A CN109003613A (en) | 2018-09-02 | 2018-09-02 | The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109003613A true CN109003613A (en) | 2018-12-14 |
Family
ID=64590222
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811017603.7A Pending CN109003613A (en) | 2018-09-02 | 2018-09-02 | The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109003613A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110364168A (en) * | 2019-07-22 | 2019-10-22 | 南京拓灵智能科技有限公司 | A kind of method for recognizing sound-groove and system based on environment sensing |
| CN110390948A (en) * | 2019-07-24 | 2019-10-29 | 厦门快商通科技股份有限公司 | A kind of method and system of Rapid Speech identification |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020010646A1 (en) * | 2000-06-30 | 2002-01-24 | Nec Corporation | Voice signature transaction system and method |
| CN104392353A (en) * | 2014-10-08 | 2015-03-04 | 无锡指网生物识别科技有限公司 | Payment method and system of voice recognition terminal |
| CN105894283A (en) * | 2015-01-26 | 2016-08-24 | 中兴通讯股份有限公司 | Mobile payment method and device based on voice control |
| CN106340298A (en) * | 2015-07-06 | 2017-01-18 | 南京理工大学 | Voiceprint unlocking method integrating content recognition and speaker recognition |
| CN107919961A (en) * | 2017-12-07 | 2018-04-17 | 广州势必可赢网络科技有限公司 | A kind of identity authentication protocol and server updated based on dynamic code and dynamic vocal print |
-
2018
- 2018-09-02 CN CN201811017603.7A patent/CN109003613A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020010646A1 (en) * | 2000-06-30 | 2002-01-24 | Nec Corporation | Voice signature transaction system and method |
| CN104392353A (en) * | 2014-10-08 | 2015-03-04 | 无锡指网生物识别科技有限公司 | Payment method and system of voice recognition terminal |
| CN105894283A (en) * | 2015-01-26 | 2016-08-24 | 中兴通讯股份有限公司 | Mobile payment method and device based on voice control |
| CN106340298A (en) * | 2015-07-06 | 2017-01-18 | 南京理工大学 | Voiceprint unlocking method integrating content recognition and speaker recognition |
| CN107919961A (en) * | 2017-12-07 | 2018-04-17 | 广州势必可赢网络科技有限公司 | A kind of identity authentication protocol and server updated based on dynamic code and dynamic vocal print |
Non-Patent Citations (2)
| Title |
|---|
| 靳双燕: "基于隐马尔可夫模型的语音识别技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 马静: "基于HMM模型的汉语数字语音识别算法的研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110364168A (en) * | 2019-07-22 | 2019-10-22 | 南京拓灵智能科技有限公司 | A kind of method for recognizing sound-groove and system based on environment sensing |
| CN110364168B (en) * | 2019-07-22 | 2021-09-14 | 北京拓灵新声科技有限公司 | Voiceprint recognition method and system based on environment perception |
| CN110390948A (en) * | 2019-07-24 | 2019-10-29 | 厦门快商通科技股份有限公司 | A kind of method and system of Rapid Speech identification |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102509547B (en) | Method and system for voiceprint recognition based on vector quantization based | |
| CN103928023B (en) | A kind of speech assessment method and system | |
| CN108922541B (en) | Multi-dimensional characteristic parameter voiceprint recognition method based on DTW and GMM models | |
| CN109215665A (en) | A kind of method for recognizing sound-groove based on 3D convolutional neural networks | |
| CN107731233A (en) | A kind of method for recognizing sound-groove based on RNN | |
| Vyas | A Gaussian mixture model based speech recognition system using Matlab | |
| CN1403953A (en) | Palm acoustic-print verifying system | |
| CN110767239A (en) | Voiceprint recognition method, device and equipment based on deep learning | |
| Zheng et al. | When automatic voice disguise meets automatic speaker verification | |
| Zhang et al. | Voice biometric identity authentication system based on android smart phone | |
| CN111489763A (en) | Adaptive method for speaker recognition in complex environment based on GMM model | |
| Sun et al. | A novel convolutional neural network voiceprint recognition method based on improved pooling method and dropout idea | |
| CN109003613A (en) | The Application on Voiceprint Recognition payment information method for anti-counterfeit of combining space information | |
| Li et al. | A study of voice print recognition technology | |
| CN111524520A (en) | Voiceprint recognition method based on error reverse propagation neural network | |
| Sukor et al. | Speaker identification system using MFCC procedure and noise reduction method | |
| Yuan et al. | Overview of the development of speaker recognition | |
| Wang et al. | Robust Text-independent Speaker Identification in a Time-varying Noisy Environment. | |
| Zailan et al. | Comparative analysis of LPC and MFCC for male speaker recognition in text-independent context | |
| Islam et al. | Improvement of text dependent speaker identification system using neuro-genetic hybrid algorithm in office environmental conditions | |
| Islam et al. | Noise robust speaker identification using PCA based genetic algorithm | |
| Balpande et al. | Speaker recognition based on mel-frequency cepstral coefficients and vector quantization | |
| Singh et al. | Features and techniques for speaker recognition | |
| Dennis et al. | Generalized Hough transform for speech pattern classification | |
| Chaudhary | Short-term spectral feature extraction and their fusion in text independent speaker recognition: A review |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181214 |