CN105245497A - Identity authentication method and device - Google Patents
Identity authentication method and device Download PDFInfo
- Publication number
- CN105245497A CN105245497A CN201510542515.9A CN201510542515A CN105245497A CN 105245497 A CN105245497 A CN 105245497A CN 201510542515 A CN201510542515 A CN 201510542515A CN 105245497 A CN105245497 A CN 105245497A
- Authority
- CN
- China
- Prior art keywords
- voice
- speech
- vector
- formula
- signal
- 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.)
- Granted
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/08—Network architectures or network communication protocols for network security for authentication of entities
- H04L63/0861—Network architectures or network communication protocols for network security for authentication of entities using biometrical features, e.g. fingerprint, retina-scan
-
- 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
- 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]
Abstract
The invention provides an identity authentication method and device, and aims at carrying out identity authentication accurately. The method comprises the steps that 1) a voice authentication database is established; 2) a voice file to be authenticated is received and stored; 3) whether the voice file to be authenticated is true is identified, if yes, a step 4) is moved to, and otherwise, authentication is ended, and a message of authentication failure is directly output; 4) features of voice signals are extracted and reconstructed; and 5) reliable characteristic vector of voice is authenticated, and an identity authentication result is output.
Description
Technical field
The application relates to technical field of biometric identification, particularly relates to a kind of identity identifying method and device.
Background technology
Along with the development of the develop rapidly of information technology, particularly Internet, deepening continuously of data message.Increasing affairs, it handles process all needs to carry out authentication, such as: at the intelligent robot of the deploying to ensure effective monitoring and control of illegal activities for intelligent entrance guard, intelligent video monitoring, public security of public safety field, customs's authentication, actual driving license checking etc.; In civil and economic field, all kinds of bank card, fiscard, credit card, the holder that saves card are carried out to the intelligent robot of authentication.In order to information security, usually need before transacting business by after checking personnel identity.
Biological identification technology is exactly a kind of technology utilizing human body biological characteristics to carry out authentication.Biological characteristic is unique, can measure or the physiological property that can automatically identify and verify or behavior.Biological recognition system samples biological characteristic, extract its unique feature and be converted into digital code, and further by these codes composition feature templates, when people carry out authentication alternately with recognition system, recognition system obtains its feature and contrasts in feature templates, to determine whether coupling.
Biological identification technology is recognition technology the most convenient and safe at present, does not need to remember complicated password, does not also need the thing carrying with key, smart card and so on.The feature of common bio-identification has fingerprint, iris, retina, DNA and voice etc.Language is as one of the natural quality of the mankind, the voice of speaker have respective biological characteristic, everyone vocal organs not only have inborn differences of Physiological, also with posteriori behavior difference, this makes to adopt speech analysis to carry out speaker ' s identity identification in increasing identity identifying technology.
But the voice document of people is also easily replicated or steals, if the voice document of people to be verified is used to handle by fraudulent copying the business agreed to without party, so just bring very large risk.
In addition, differentiated the identity of personnel by voice, when receiving phonetic entry, ambient noise can affect the accuracy differentiated for voice.How solving environmental noise for the impact of speech recognition is also a problem urgently to be resolved hurrily.
Summary of the invention
In view of this, the application provides a kind of for identity identifying method and device, and its problem can avoided fraudulent copying and reduce recognition accuracy realizes carrying out personnel identity discriminating with high-accuracy.
The application provides a kind of identity identifying method, and described method comprises:
Step one, sets up voice authentication database;
Step 2, receives and stores voice document to be certified;
Step 3, identifies the true and false of voice document to be certified, if very, then enters step 4, if there is vacation, then terminates direct authentication output failed message;
Step 4, extracts phonic signal character and the reliable speech feature of rebuilding in voice signal;
Step 5, carries out certification to the reliable characteristic vector of voice and exports identity authentication result.
In the application one specific embodiment, described step one comprises: gather proprietary reliable speech, reliable speech feature extraction and in described voice authentication database, record voice characteristics information.
In the application one specific embodiment, described step 3 is: the first threshold whether exceeding setting in qualification voice document owing to copying the special characteristic brought.
In the application one specific embodiment, copy the special characteristic brought described in calculating and comprise:
If
represent frame number be the voice signal of T, then q (
) frame signal
discrete Fourier transform be:
formula (3-1)
Wherein, N is the status number of equine husband chain in voice signal;
Then the expression formula of average frame is:
formula (3-2)
The difference that original real speech and copying voice exist at frequency-portions is extracted by formula (3-3):
formula (3-3)
Wherein, filter () is filter function arbitrary in prior art.
In the application one specific embodiment, described step 4 comprises:
The speech feature extraction algorithm used when adopting speech recognition library to set up, extracts the voice signal inputted in daily life
feature vector, X;
By the voice signal inputted in daily life
feature vector, X be separated, be divided into speech vector
and noise vector
;
Feature vector, X is divided into speech vector according to the average of its Gaussian function and variance
and noise vector
, and calculate the individual given speech vector of v
prior probability
:
(formula 4-2)
By the speech vector of voice signal
as the reliable characteristic vector of input speech signal.
Disclosed herein as well is a kind of identification authentication system, described device comprises:
Voice authentication database 1, for storing the voice characteristics information of all personnel;
Acquisition module 2, gathers the voice messaging of personnel to be certified;
Authenticity module 3, for the identification of the true and false of voice document to be certified;
Characteristic extracting module 4, extracts phonic signal character and the feature of rebuilding in voice signal;
Authentication module 5, carries out certification to the reliable characteristic vector of voice and exports identity authentication result.
In the application one specific embodiment, described voice authentication database 1 for gather proprietary reliable speech, reliable speech feature extraction and in described voice authentication database, record voice characteristics information.
In the application one specific embodiment, whether described authenticity module 3 is specifically for exceeding the first threshold of setting owing to copying the special characteristic brought in qualification voice document.
In the application one specific embodiment, copy the special characteristic brought described in calculating and comprise:
If
represent frame number be the voice signal of T, then q (
) frame signal
discrete Fourier transform be:
formula (3-1)
Wherein, N is the status number of equine husband chain in voice signal;
Then the expression formula of average frame is:
formula (3-2)
The difference that original real speech and copying voice exist at frequency-portions is extracted by formula (3-3):
formula (3-3)
Wherein, filter () is filter function arbitrary in prior art.
In the application one specific embodiment, described authentication module 5 comprises:
The speech feature extraction algorithm used when adopting speech recognition library to set up, extracts the voice signal inputted in daily life
feature vector, X;
By the voice signal inputted in daily life
feature vector, X be separated, be divided into speech vector
and noise vector
;
Feature vector, X is divided into speech vector according to the average of its Gaussian function and variance
and noise vector
, and calculate the individual given speech vector of v
prior probability
:
(formula 4-2)
By the speech vector of voice signal
as the reliable characteristic vector of input speech signal.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present application or technical scheme of the prior art, be briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, the accompanying drawing that the following describes is only some embodiments recorded in the application, for those of ordinary skill in the art, other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is the flow chart setting up voice authentication database in the application;
In Fig. 2 the application, identity identifying method flow chart is;
Fig. 3 is identification authentication system structure chart in the application.
Embodiment
Technical scheme in the application is understood better in order to make those skilled in the art, below in conjunction with the accompanying drawing in the embodiment of the present application, technical scheme in the embodiment of the present application is clearly and completely described, obviously, described embodiment is only some embodiments of the present application, instead of whole embodiments.Based on the embodiment in the application, the every other embodiment that those of ordinary skill in the art obtain, all should belong to the scope of the application's protection.
To carry out in authentication, due to the problem that fraudulent copying and environmental noise make speech recognition accuracy rate decline, this application discloses a kind of identity identifying method and device to solve the use voice existed in prior art.
The application's specific implementation is further illustrated below in conjunction with illustrations.
As shown in Figure 1, first need to set up voice authentication database before carrying out voice authentication, the process of establishing of speech database can utilize any existing speech feature extraction technology, preferably, can use following process of establishing:
As required, the voice messaging that all n needs certification personnel is gathered.In order to ensure the accuracy rate of speech recognition, all personnel's voice messaging in varied situations constantly can be increased, to improve discrimination.
The corresponding audio signal x (i) of voice messaging for the personnel i collected is stored to the raw tone memory block in speech recognition database, i=1 ..., n (i is positive integer).
The characteristic extraction procedure of voice comprises the following steps:
Following process is carried out for each audio signal x (i) stored in raw tone memory block:
(1) audio signal x (i) is divided into a series of continuous print frame, Fourier transform is done to every frame signal.
(2) filter is used to process audio signal, to reduce the mutual leakage of spectrum energy between nearby frequency bands; The filter function used in filter is:
) formula (1)
Wherein:
Parameter θ is the initial phase of filter, and n is the exponent number of filter;
As t<0, u (t)=0, as t>0, u (t)=1;
B=1.019*ERB (
), ERB (
) be the Equivalent Rectangular Bandwidth of filter, its same filter centre frequency
pass be:
ERB (
)=24.7+0.108
formula (2).
(3) the middle deviation of audio signal is removed.
After audio signal framing, the frame of some is formed a segmentation, preferably 7 frames are formed a segmentation in the present invention, this can be arranged according to the disposal ability of system.
The frame length that most of speech recognition system uses is 20ms-30ms, preferably use 26.5ms as Hamming window in the present invention, overlapping frame length is 10ms, the intermediate quantity Q (i of every frame, j) obtained by the mean value of frame energy P (i, j) in compute segment:
formula (3)
In formula (3) due to the present invention preferably 7 frames form segmentation, thus a M=3.I is channel number, and j is the sequence of required frame, and j ' is the sequence of frame each in required segmentation.
In noise energy removal process, use the ratio (AM/GM) of arithmetic mean and geometrical mean can represent the degree that voice signal is corroded, obtain after logarithm is asked to above-mentioned ratio:
formula (4)
In formula (4), z is floor coefficient, in order to avoid negative infinitesimal valuation to ensure that the deviation of result of calculation is in allowed band; J is the sequence sum of frame.
Suppose that B (i) is the deviation caused by background noise, i represents channel sequence, is obtained by that thing of conditional probability, removes the intermediate quantity Q ' after deviation (i, j|B (i)) to be:
formula (5)
Can obtain:
formula (6)
For formula (6), when AM/GM value closest to acoustic signal of the ratio of AM/GM under noise situations, can be in the hope of the estimated value of B (i):
formula (7)
Wherein,
represent G (i) respective value in acoustic signal, obtain after each channel computing formula (7), for each time-frequency BIN signal (i, j), the ratio of noise removal is:
formula (8)
In order to smoothing computation, average to the noise removal ratio of channel i-N to i+N, after adjustment, final function is:
formula (10)
Use formula (10) to process audio signals all in filter, remove the output as filter after middle deviation.
(4) audio signal data exported all filters does non-linear power-function arithmetic, and the power function used is:
formula (11).
(5) speech characteristic parameter is obtained after discrete cosine transform being done further to the output of (4) step.
Because discrete cosine transform (DCT) is the known processing mode in speech processes field, do not repeat them here.
The phonetic feature calculated is stored in a database.
As shown in Figure 2, this application discloses a kind of identity identifying method, it comprises the following steps:
Step1: set up voice authentication database.
Arbitrary speech feature extraction technology in prior art can be used to set up voice authentication database, also can to set up voice authentication database by above-described optimal way.
Step2: receive and store voice document to be certified.
Information of voice prompt can be set in voice capture device, point out the personnel of identity to be identified to input voice document.Such as, by the voice of microphone collector.Also other voice capture device can be adopted.
Step3: the true and false identifying voice document to be certified, if very, then enters Step4, if there is vacation, then terminates direct authentication output failed message.
Usually, the approach that false voice document is normally agreed to without party is obtained by the method copied, but the copying voice file of the mode by multiple copies, will inevitably change the characteristic information in voice document, and this change normally exists along with the signal in whole voice document equably.Thus, be tested and appraised in voice document and whether exceed the first threshold of setting to carry out authenticity owing to copying the special characteristic brought.
First, if
represent frame number be the voice signal of T, then q (
) frame signal
the discrete Fourier transform of (N is the status number of equine husband chain in voice signal) is:
formula (3-1)
Then the expression formula of average frame is:
formula (3-2)
Usually, original real speech and copying voice there are differences at frequency-portions, can pass through formula (3-3) and extract this species diversity:
formula (3-3)
Wherein, filter () is filter function arbitrary in prior art, such as, filter function in formula (1).
In this application, through overtesting, preferred first threshold=0.53.
Step4: extract phonic signal character and the feature of rebuilding in voice signal.
Usually the voice signal used when building voice authentication database normally carries out specialty collection under quiet environment, and in actual authentication process, when phonetic entry normally in daily living environment, various noise may be there is, if directly carry out on the voice signal inputted under noise conditions the impact that feature extraction can be subject to noise speech information, and then affect the accuracy rate of authentication.
The speech feature extraction algorithm used when adopting speech recognition library to set up, extracts the voice signal inputted in daily life
feature vector, X.Thus, the voice signal that can will input in daily life
feature vector, X be separated, be divided into speech vector
and noise vector
.
By prior probability p (X) Modling model of voice signal, then obtained by merging, training data:
(formula 4-1)
Wherein, V is the quantity of mixed cell, and v is sequence number, and p (v) is the prior probability of a mixed cell
represent v Gaussian Profile, those skilled in the art's its Mean Matrix known is
, diagonal covariance matrix is
.If the feature of a given voice signal, be just divided into speech vector according to the average of its Gaussian function and variance
and noise vector
, and then calculate the individual given speech vector of v
prior probability
:
(formula 4-2)
In process of reconstruction, the speech vector of voice signal
be retained the reliable characteristic vector being used as input speech signal.
Step5: the reliable characteristic vector of voice is carried out to certification and exports identity authentication result.
The reliable characteristic of voice signal vector input voice authentication database is compared, if find the voice signal conformed at voice authentication database, by checking, if do not find the voice signal conformed at voice authentication database, not by checking.
As shown in Figure 3, present invention also provides a kind of identification authentication system, it comprises:
Voice authentication database 1, for storing the voice characteristics information of all personnel.
Arbitrary speech feature extraction technology in prior art can be used to set up voice authentication database, also can to set up voice authentication database by optimal way described below.
As shown in Figure 1, first need to set up voice authentication database before carrying out voice authentication, the process of establishing of speech database can utilize any existing speech feature extraction technology, preferably, can use following process of establishing:
As required, the voice messaging that all n needs certification personnel is gathered.In order to ensure the accuracy rate of speech recognition, all personnel's voice messaging in varied situations constantly can be increased, to improve discrimination.
The corresponding audio signal x (i) of voice messaging for the personnel i collected is stored to the raw tone memory block in speech recognition database, i=1 ..., n (i is positive integer).
The characteristic extraction procedure of voice comprises the following steps:
Following process is carried out for each audio signal x (i) stored in raw tone memory block:
(1) audio signal x (i) is divided into a series of continuous print frame, Fourier transform is done to every frame signal.
(2) filter is used to process audio signal, to reduce the mutual leakage of spectrum energy between nearby frequency bands; The filter function used in filter is:
) formula (1)
Wherein:
Parameter θ is the initial phase of filter, and n is the exponent number of filter;
As t<0, u (t)=0, as t>0, u (t)=1;
B=1.019*ERB (
), ERB (
) be the Equivalent Rectangular Bandwidth of filter, its same filter centre frequency
pass be:
ERB (
)=24.7+0.108
formula (2).
(3) the middle deviation of audio signal is removed.
After audio signal framing, the frame of some is formed a segmentation, preferably 7 frames are formed a segmentation in the present invention, this can be arranged according to the disposal ability of system.
The frame length that most of speech recognition system uses is 20ms-30ms, preferably use 26.5ms as Hamming window in the present invention, overlapping frame length is 10ms, the intermediate quantity Q (i of every frame, j) obtained by the mean value of frame energy P (i, j) in compute segment:
formula (3)
In formula (3) due to the present invention preferably 7 frames form segmentation, thus a M=3.I is channel number, and j is the sequence of required frame, and j ' is the sequence of frame each in required segmentation.
In noise energy removal process, use the ratio (AM/GM) of arithmetic mean and geometrical mean can represent the degree that voice signal is corroded, obtain after logarithm is asked to above-mentioned ratio:
formula (4)
In formula (4), z is floor coefficient, in order to avoid negative infinitesimal valuation to ensure that the deviation of result of calculation is in allowed band; J is the sequence sum of frame.
Suppose that B (i) is the deviation caused by background noise, i represents channel sequence, is obtained by that thing of conditional probability, removes the intermediate quantity Q ' after deviation (i, j|B (i)) to be:
formula (5)
Can obtain:
formula (6)
For formula (6), when AM/GM value closest to acoustic signal of the ratio of AM/GM under noise situations, can be in the hope of the estimated value of B (i):
formula (7)
Wherein,
represent G (i) respective value in acoustic signal, obtain after each channel computing formula (7), for each time-frequency BIN signal (i, j), the ratio of noise removal is:
formula (8)
In order to smoothing computation, average to the noise removal ratio of channel i-N to i+N, after adjustment, final function is:
formula (10)
Use formula (10) to process audio signals all in filter, remove the output as filter after middle deviation.
(4) audio signal data exported all filters does non-linear power-function arithmetic, and the power function used is:
formula (11).
(5) speech characteristic parameter is obtained after discrete cosine transform being done further to the output of (4) step.
Because discrete cosine transform (DCT) is the known processing mode in speech processes field, do not repeat them here.
The phonetic feature calculated is stored in a database.
Acquisition module 2, gathers the voice messaging of personnel to be certified.
Information of voice prompt can be set in voice capture device, point out the personnel of identity to be identified to input voice document.Such as, by the voice of microphone collector.Also other voice capture device can be adopted.
Authenticity module 3, for the identification of the true and false of voice document to be certified.
Usually, the approach that false voice document is normally agreed to without party is obtained by the method copied, but the copying voice file of the mode by multiple copies, will inevitably change the characteristic information in voice document, and this change normally exists along with the signal in whole voice document equably.Thus, be tested and appraised in voice document and whether exceed the first threshold of setting to carry out authenticity owing to copying the special characteristic brought.
First, if
represent frame number be the voice signal of T, then q (
) frame signal
the discrete Fourier transform of (N is the status number of equine husband chain in voice signal) is:
formula (3-1)
Then the expression formula of average frame is:
formula (3-2)
Usually, original real speech and copying voice there are differences at frequency-portions, can pass through formula (3-3) and extract this species diversity:
formula (3-3)
Wherein, filter () is filter function arbitrary in prior art, such as, filter function in formula (1).
In this application, through overtesting, preferred first threshold=0.53.
Characteristic extracting module 4, extracts phonic signal character and the feature of rebuilding in voice signal.
Usually the voice signal used when building voice authentication database normally carries out specialty collection under quiet environment, and in actual authentication process, when phonetic entry normally in daily living environment, various noise may be there is, if directly carry out on the voice signal inputted under noise conditions the impact that feature extraction can be subject to noise speech information, and then affect the accuracy rate of authentication.
The speech feature extraction algorithm used when adopting speech recognition library to set up, extracts the voice signal inputted in daily life
feature vector, X.Thus, the voice signal that can will input in daily life
feature vector, X be separated, be divided into speech vector
and noise vector
.
By prior probability p (X) Modling model of voice signal, then obtained by merging, training data:
(formula 4-1)
Wherein, V is the quantity of mixed cell, and v is sequence number, and p (v) is the prior probability of a mixed cell
represent v Gaussian Profile, those skilled in the art's its Mean Matrix known is
, diagonal covariance matrix is
.If the feature of a given voice signal, be just divided into speech vector according to the average of its Gaussian function and variance
and noise vector
, and then calculate the individual given speech vector of v
prior probability
:
(formula 4-2)
In process of reconstruction, the speech vector of voice signal
be retained the reliable characteristic vector being used as input speech signal.
Authentication module 5, carries out certification to the reliable characteristic vector of voice and exports identity authentication result.
The reliable characteristic of voice signal vector input voice authentication database is compared, if find the voice signal conformed at voice authentication database, by checking, if do not find the voice signal conformed at voice authentication database, not by checking.
Certainly, the arbitrary technical scheme implementing the application must not necessarily need to reach above all advantages simultaneously.
It will be understood by those skilled in the art that the embodiment of the application can be provided as method, device (equipment) or computer program.Therefore, the application can adopt the form of complete hardware embodiment, completely software implementation or the embodiment in conjunction with software and hardware aspect.And the application can adopt in one or more form wherein including the upper computer program implemented of computer-usable storage medium (including but not limited to magnetic disc store, CD-ROM, optical memory etc.) of computer usable program code.
The application describes with reference to according to the flow chart of the method for the embodiment of the present application, device (equipment) and computer program and/or block diagram.Should understand can by the combination of the flow process in each flow process in computer program instructions realization flow figure and/or block diagram and/or square frame and flow chart and/or block diagram and/or square frame.These computer program instructions can being provided to the processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing device to produce a machine, making the instruction performed by the processor of computer or other programmable data processing device produce device for realizing the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.
These computer program instructions also can be stored in can in the computer-readable memory that works in a specific way of vectoring computer or other programmable data processing device, the instruction making to be stored in this computer-readable memory produces the manufacture comprising command device, and this command device realizes the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.
These computer program instructions also can be loaded in computer or other programmable data processing device, make on computer or other programmable devices, to perform sequence of operations step to produce computer implemented process, thus the instruction performed on computer or other programmable devices is provided for the step realizing the function of specifying in flow chart flow process or multiple flow process and/or block diagram square frame or multiple square frame.
Although described the preferred embodiment of the application, those skilled in the art once obtain the basic creative concept of cicada, then can make other change and amendment to these embodiments.So claims are intended to be interpreted as comprising preferred embodiment and falling into all changes and the amendment of the application's scope.Obviously, those skilled in the art can carry out various change and modification to the application and not depart from the spirit and scope of the application.Like this, if these amendments of the application and modification belong within the scope of the application's claim and equivalent technologies thereof, then the application is also intended to comprise these change and modification.
Claims (10)
1. an identity identifying method, described method comprises:
Step one, sets up voice authentication database;
Step 2, receives and stores voice document to be certified;
Step 3, identifies the true and false of voice document to be certified, if very, then enters step 4, if there is vacation, then terminates direct authentication output failed message;
Step 4, extracts phonic signal character and the reliable speech feature of rebuilding in voice signal;
Step 5, carries out certification to the reliable characteristic vector of voice and exports identity authentication result.
2. method according to claim 1, is characterized in that, described step one comprises: gather proprietary reliable speech, reliable speech feature extraction and in described voice authentication database, record voice characteristics information.
3. method according to claim 1, is characterized in that, described step 3 is: the first threshold whether exceeding setting in qualification voice document owing to copying the special characteristic brought.
4. method according to claim 3, is characterized in that, copies the special characteristic brought and comprise described in calculating:
If
represent frame number be the voice signal of T, then q (
) frame signal
discrete Fourier transform be:
formula (3-1)
Wherein, N is the status number of equine husband chain in voice signal;
Then the expression formula of average frame is:
formula (3-2)
The difference that original real speech and copying voice exist at frequency-portions is extracted by formula (3-3):
formula (3-3)
Wherein, filter () is filter function arbitrary in prior art.
5. method according to claim 1, is characterized in that, described step 4 comprises:
The speech feature extraction algorithm used when adopting speech recognition library to set up, extracts the voice signal inputted in daily life
feature vector, X;
By the voice signal inputted in daily life
feature vector, X be separated, be divided into speech vector
and noise vector
;
Feature vector, X is divided into speech vector according to the average of its Gaussian function and variance
and noise vector
, and calculate the individual given speech vector of v
prior probability
:
(formula 4-2)
By the speech vector of voice signal
as the reliable characteristic vector of input speech signal.
6. an identification authentication system, described device comprises:
Voice authentication database 1, for storing the voice characteristics information of all personnel;
Acquisition module 2, gathers the voice messaging of personnel to be certified;
Authenticity module 3, for the identification of the true and false of voice document to be certified;
Characteristic extracting module 4, extracts phonic signal character and the feature of rebuilding in voice signal;
Authentication module 5, carries out certification to the reliable characteristic vector of voice and exports identity authentication result.
7. device according to claim 6, is characterized in that, described voice authentication database 1 for gather proprietary reliable speech, reliable speech feature extraction and in described voice authentication database, record voice characteristics information.
8. device according to claim 6, is characterized in that, whether described authenticity module 3 is specifically for exceeding the first threshold of setting owing to copying the special characteristic brought in qualification voice document.
9. device according to claim 8, is characterized in that, copies the special characteristic brought and comprise described in calculating:
If
represent frame number be the voice signal of T, then q (
) frame signal
discrete Fourier transform be:
formula (3-1)
Wherein, N is the status number of equine husband chain in voice signal;
Then the expression formula of average frame is:
formula (3-2)
The difference that original real speech and copying voice exist at frequency-portions is extracted by formula (3-3):
formula (3-3)
Wherein, filter () is filter function arbitrary in prior art.
10. device according to claim 6, is characterized in that, described authentication module 5 comprises:
The speech feature extraction algorithm used when adopting speech recognition library to set up, extracts the voice signal inputted in daily life
feature vector, X;
By the voice signal inputted in daily life
feature vector, X be separated, be divided into speech vector
and noise vector
;
Feature vector, X is divided into speech vector according to the average of its Gaussian function and variance
and noise vector
, and calculate the individual given speech vector of v
prior probability
:
(formula 4-2)
By the speech vector of voice signal
as the reliable characteristic vector of input speech signal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510542515.9A CN105245497B (en) | 2015-08-31 | 2015-08-31 | A kind of identity identifying method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510542515.9A CN105245497B (en) | 2015-08-31 | 2015-08-31 | A kind of identity identifying method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105245497A true CN105245497A (en) | 2016-01-13 |
| CN105245497B CN105245497B (en) | 2019-01-04 |
Family
ID=55042997
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510542515.9A Active CN105245497B (en) | 2015-08-31 | 2015-08-31 | A kind of identity identifying method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105245497B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107886956A (en) * | 2017-11-13 | 2018-04-06 | 广州酷狗计算机科技有限公司 | Audio identification methods, device and computer-readable storage medium |
| CN109215643A (en) * | 2017-07-05 | 2019-01-15 | 阿里巴巴集团控股有限公司 | A kind of exchange method, electronic equipment and server |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5054083A (en) * | 1989-05-09 | 1991-10-01 | Texas Instruments Incorporated | Voice verification circuit for validating the identity of an unknown person |
| CN1397929A (en) * | 2002-07-12 | 2003-02-19 | 清华大学 | Speech intensifying-characteristic weighing-logrithmic spectrum addition method for anti-noise speech recognization |
| CN1627365A (en) * | 2003-12-09 | 2005-06-15 | 摩托罗拉公司 | Method and system adaptive to speaker without vetating to speech identifying database |
| CN102324232A (en) * | 2011-09-12 | 2012-01-18 | 辽宁工业大学 | Method for recognizing sound-groove and system based on gauss hybrid models |
| CN103236260A (en) * | 2013-03-29 | 2013-08-07 | 京东方科技集团股份有限公司 | Voice recognition system |
| US20140079248A1 (en) * | 2012-05-04 | 2014-03-20 | Kaonyx Labs LLC | Systems and Methods for Source Signal Separation |
| CN103730114A (en) * | 2013-12-31 | 2014-04-16 | 上海交通大学无锡研究院 | Mobile equipment voiceprint recognition method based on joint factor analysis model |
-
2015
- 2015-08-31 CN CN201510542515.9A patent/CN105245497B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5054083A (en) * | 1989-05-09 | 1991-10-01 | Texas Instruments Incorporated | Voice verification circuit for validating the identity of an unknown person |
| CN1397929A (en) * | 2002-07-12 | 2003-02-19 | 清华大学 | Speech intensifying-characteristic weighing-logrithmic spectrum addition method for anti-noise speech recognization |
| CN1627365A (en) * | 2003-12-09 | 2005-06-15 | 摩托罗拉公司 | Method and system adaptive to speaker without vetating to speech identifying database |
| CN102324232A (en) * | 2011-09-12 | 2012-01-18 | 辽宁工业大学 | Method for recognizing sound-groove and system based on gauss hybrid models |
| US20140079248A1 (en) * | 2012-05-04 | 2014-03-20 | Kaonyx Labs LLC | Systems and Methods for Source Signal Separation |
| CN103236260A (en) * | 2013-03-29 | 2013-08-07 | 京东方科技集团股份有限公司 | Voice recognition system |
| CN103730114A (en) * | 2013-12-31 | 2014-04-16 | 上海交通大学无锡研究院 | Mobile equipment voiceprint recognition method based on joint factor analysis model |
Non-Patent Citations (2)
| Title |
|---|
| 吴迪等: "《基于自适应高斯混合模型与静动态听觉特征融合的说话人识别》", 《光学精密工程》 * |
| 魏峰等: "《基于HMM的语音识别技术研究》", 《2006通信理论与技术新进展-第十一届全国青年通信学术会议论文集》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109215643A (en) * | 2017-07-05 | 2019-01-15 | 阿里巴巴集团控股有限公司 | A kind of exchange method, electronic equipment and server |
| CN109215643B (en) * | 2017-07-05 | 2023-10-24 | 阿里巴巴集团控股有限公司 | Interaction method, electronic equipment and server |
| CN107886956A (en) * | 2017-11-13 | 2018-04-06 | 广州酷狗计算机科技有限公司 | Audio identification methods, device and computer-readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105245497B (en) | 2019-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104700018B (en) | A kind of recognition methods for intelligent robot | |
| Ali et al. | Edge-centric multimodal authentication system using encrypted biometric templates | |
| CN104700094B (en) | A kind of face identification method for intelligent robot and system | |
| CN108429619A (en) | Identity identifying method and system | |
| CN104751845A (en) | Voice recognition method and system used for intelligent robot | |
| CN106919921B (en) | Gait recognition method and system combining subspace learning and tensor neural network | |
| CN106709402A (en) | Living person identity authentication method based on voice pattern and image features | |
| CN102629320B (en) | Ordinal measurement statistical description face recognition method based on feature level | |
| CN110458662B (en) | Anti-fraud wind control method and device | |
| CN104751143A (en) | Person and credential verification system and method based on deep learning | |
| CN105138886B (en) | Robot biometric identification system | |
| CN110767239A (en) | Voiceprint recognition method, device and equipment based on deep learning | |
| CN110827832A (en) | Video identity recognition equipment and method | |
| CN103279744A (en) | Multi-scale tri-mode texture feature-based method and system for detecting counterfeit fingerprints | |
| CN110598522A (en) | Identity comparison method based on face and palm print palm vein recognition | |
| CN105304091A (en) | Voice tamper recovery method based on DCT | |
| CN113343198B (en) | Video-based random gesture authentication method and system | |
| CN111081223A (en) | Voice recognition method, device, equipment and storage medium | |
| CN105245497A (en) | Identity authentication method and device | |
| Revathi et al. | Person authentication using speech as a biometric against play back attacks | |
| Bansal et al. | Multimodal biometrics by fusion for security using genetic algorithm | |
| CN104091104A (en) | Feature extraction and authentication method for multi-format audio perceptual Hashing authentication | |
| Shenai et al. | Fast biometric authentication system based on audio-visual fusion | |
| Vishwakarma et al. | Biometric Iris Recognition using Sobel Edge Detection for Secured Authentication | |
| CN113838469A (en) | Identity recognition method, system and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |