CN105933272A - Voiceprint recognition method capable of preventing recording attack, server, terminal, and system - Google Patents
Voiceprint recognition method capable of preventing recording attack, server, terminal, and system Download PDFInfo
- Publication number
- CN105933272A CN105933272A CN201511020257.4A CN201511020257A CN105933272A CN 105933272 A CN105933272 A CN 105933272A CN 201511020257 A CN201511020257 A CN 201511020257A CN 105933272 A CN105933272 A CN 105933272A
- Authority
- CN
- China
- Prior art keywords
- character
- user speech
- voiceprint
- voice
- described user
- 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
- 238000000034 method Methods 0.000 title claims abstract description 42
- 230000001755 vocal effect Effects 0.000 claims description 7
- 239000000284 extract Substances 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 5
- 208000031481 Pathologic Constriction Diseases 0.000 claims description 2
- 210000001215 vagina Anatomy 0.000 claims description 2
- 238000012795 verification Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000005070 sampling Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 230000002194 synthesizing effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000013011 mating Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000005236 sound signal Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
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 TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/08—Speech classification or search
- G10L15/18—Speech classification or search using natural language modelling
- G10L15/183—Speech classification or search using natural language modelling using context dependencies, e.g. language models
- G10L15/187—Phonemic context, e.g. pronunciation rules, phonotactical constraints or phoneme n-grams
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L17/00—Speaker identification or verification techniques
-
- 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
- H04L63/14—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic
- H04L63/1441—Countermeasures against malicious traffic
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/32—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols including means for verifying the identity or authority of a user of the system or for message authentication, e.g. authorization, entity authentication, data integrity or data verification, non-repudiation, key authentication or verification of credentials
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Computer Security & Cryptography (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Computational Linguistics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computer Hardware Design (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Telephonic Communication Services (AREA)
Abstract
The invention provides a voiceprint recognition method capable of preventing recording attack, a server, a terminal, and a system. The voiceprint recognition method includes generating character combination and character pronunciation rules based on a user's voiceprint recognition request; sending the character combination and character pronunciation rules to a request terminal; receiving the user voice input by the request terminal based on the character combination and character pronunciation rules; performing voiceprint recognition based on the user voice, and the character combination and character pronunciation rules; and sending the voiceprint recognition result to the request terminal. According to the invention, the recording attack can be effectively prevented.
Description
Technical field
The invention belongs to Application on Voiceprint Recognition field, be prevented from, particularly to a kind of, voiceprint authentication method, the clothes that recording is attacked
Business device, terminal and system.
Background technology
Vocal print is the same with fingerprint, is a kind of very important biological characteristic that can characterize people's identity.Compare traditional password
The means such as certification, the feature such as vocal print high security and convenience.Attack means the most frequently used in voiceprint mainly has record
Sound replay attack, speaker's bogus attack and forgery certification voice are attacked.
Wherein recording replay attack refers to that assailant passes through the sound pick-up outfit of high-fidelity by various means acquisition user's
Speech samples, uses the original recording of user or by synthesizing " the true sound of speaker " after the process of the means such as cutting, splicing,
Then, when Verification System gathers user speech, played back by the power amplifier of high-fidelity, thus attack.Speak
People's bogus attack refers to that some are good at the assailant giving tacit consent to other people voice by imitating the tongue of speaker and sending out
Sound feature is attacked.The technological means such as forgery certification voice attack refers to by synthesizing, changing, splicing forge quilt
The voice of assailant is attacked.
Speaker's bogus attack needs assailant to have good ability to model, forges the attack of certification voice and also tends to needs
Higher professional skill, both attacks itself are attacked the highest, and the most either imitative sound still forges sound, eventually
Studying carefully is not true sound, and existing sound groove recognition technology in e substantially copes with this two class and attacks.
Recording replay attack is the very important problem faced in Application on Voiceprint Recognition, and assailant passes through software after obtaining sound
Synthesis is attacked.Recording attack has two kinds of situations, and one is that user's sound of speaking in other cases is stolen
Attack；Another kind be user when carrying out Application on Voiceprint Recognition, by Malware admission user sound attack.
Attack for recording, in prior art, mainly have a following two kinds solution:
The first scheme is to be recorded by analysis and between raw tone, in channel characteristics pattern, difference is told and is
No is recording substance；First scheme is while the vocal print of checking speaker, also verifies the content of speaking of speaker,
Because recording assailant is not aware that this content of speaking.
But, a pair sound signal quality of scheme, signal to noise ratio, channel quality etc. require the highest, take in actual applications
The effect obtained not is fine.
If the most random big section of word of user writable that allow in scheme two, Consumer's Experience is poor, if reducing user's
Phonetic entry, such as patent (application number: 201310123555.0；Denomination of invention: body based on dynamic password voice
Part confirms system and method), from 26 English alphabets and 10 numerals, select combination, each random group symphysis
After producing dynamic password, user is allowed to be inputted by voice, owing to being not aware that the dynamic password of production every time in advance,
So attack of simply recording can be resisted, it it is a kind of preferably solution.But owing to this patent is only at 26
Totally 36 character random combines in English alphabet and 10 numerals, if assailant is by the way of recording separates, point
Being separated out this 36 characters, then no matter obtain which kind of random string, assailant has only to simply by 36 words
Symbol carries out splicing attack.
Summary of the invention
The present invention provide a kind of have prevent recording from attacking the voiceprint authentication method of function, server and terminal, be used for solving
Certainly prior art prevent recording attack method from there is leak, it is impossible to effectively to prevent the defect that recording is attacked.
In order to solve above-mentioned technical problem, the present invention provides a kind of and is prevented from the voiceprint authentication method that recording is attacked,
Voiceprint request according to a user generates character combination and the pronunciation rule of character；
The pronunciation rule of described character combination and character is sent to requesting terminal；
Receive the user speech that described requesting terminal inputs according to the pronunciation rule of described character combination and character；
Pronunciation rule according to described user speech, described character combination and character carries out voiceprint；；
Described voiceprint result is sent to described requesting terminal.
The present invention separately provides a kind of and is prevented from the voiceprint authentication method that recording is attacked,
The voiceprint sending a user is asked to server；
Receive and show character combination and the pronunciation rule of character that described server sends；
Receive the user speech that user inputs according to the pronunciation rule of described character combination and character；
Described user speech is sent to described server；
Receive the voiceprint result that described server sends.
The present invention separately provides a kind of voiceprint server being prevented from recording,
Signal generating unit, generates character combination and the pronunciation rule of character for the request according to a user；
Transmitting element, for being sent to requesting terminal by the pronunciation rule of described character combination and character, by voiceprint
Result sends to described requesting terminal；
Receive unit, for receiving the use that described requesting terminal inputs according to the pronunciation rule of described character combination and character
Family voice；
Sound detection unit, for carrying out sound according to the pronunciation rule of described user speech, described character combination and character
Stricture of vagina certification；
The present invention provides again a kind of and is prevented from the voiceprint terminal that recording is attacked,
Request unit, asks to server for sending the voiceprint of a user；
Receive unit, for receiving and show character combination and the pronunciation rule of character that described server sends, receive
The voiceprint result that described server sends；
Typing unit, for receiving the user speech that user inputs according to the pronunciation rule of described character combination and character；
Transmitting element, for sending described user speech to described server.
The present invention reoffers a kind of voiceprint authentication system being prevented from recording attack, and this system includes server and request
Terminal, wherein, described server asks to generate character combination and the pronunciation of character for the voiceprint according to a user
Rule；The pronunciation rule of described character combination and character is sent to requesting terminal；Receive described requesting terminal according to institute
State the user speech of the pronunciation rule input of character combination and character；According to described user speech, described character combination and
The pronunciation rule of character carries out voiceprint；Described voiceprint result is sent to described requesting terminal；
Described requesting terminal is asked to server for the voiceprint sending a user；Receive and show described server
The character combination sent and the pronunciation rule of character；Receive user defeated according to the pronunciation rule of described character combination and character
The user speech entered；Described user speech is sent to described server；Receive the voiceprint that described server sends
Result.
What the present invention proposed is prevented from voiceprint authentication method, server, terminal and the system that recording is attacked, by testing
Character combination that character in card user speech and articulation type and server generate and the pronunciation rule of character whether one
Cause, it is possible to effectively preventing recording from attacking, assailant is i.e. enabled the user speech got by other channels and meets language
Sound content, also cannot meet the requirement of articulation type.Further, for the user speech preventing user from repeatedly inputting
Attacked by recording, it is judged that character combination that character in user speech and articulation type generate with server and character
Pronunciation rule consistent after, also judge voice the most to be verified and the voice of this user in history sound bank whether
Cause, if consistent, illustrate that there is recording attacks.The present invention can effectively prevent the recording in voiceprint from attacking.
Accompanying drawing explanation
In order to be illustrated more clearly that the technical scheme of the embodiment of the present invention, in embodiment being described below required for make
Accompanying drawing be briefly described, it should be apparent that, below describe in accompanying drawing be only some embodiments of the present invention,
For those of ordinary skill in the art, on the premise of not paying creative work, it is also possible to according to these accompanying drawings
Obtain other accompanying drawing.
Fig. 1 is the voiceprint authentication method flow chart being prevented from recording attack of one embodiment of the invention；
Fig. 2 is the voiceprint process flow diagram flow chart being prevented from recording attack of one embodiment of the invention；
Fig. 3 is the voiceprint process flow diagram flow chart being prevented from recording attack of one embodiment of the invention；
Fig. 4 is the oscillogram of the pronunciation correspondence of the digital " 0 " of one embodiment of the invention；
Fig. 5 is the voiceprint authentication method flow chart being prevented from recording attack of one embodiment of the invention；
Fig. 6 is the voiceprint server being prevented from recording attack of one embodiment of the invention；
Fig. 7 is the voiceprint terminal being prevented from recording attack of one embodiment of the invention；
Fig. 8 is the voiceprint authentication system being prevented from recording attack of one embodiment of the invention；
Fig. 9 be one embodiment of the invention have prevent recording attack function voiceprint authentication method flow chart.
Detailed description of the invention
Technical characterstic and effect in order to make the present invention become apparent from, and do technical scheme below in conjunction with the accompanying drawings
Further illustrating, the present invention also can have other different instantiations be illustrated or implement, any art technology
The equivalents that personnel do within the scope of the claims belongs to the protection category of the present invention.
As it is shown in figure 1, the voiceprint authentication method flow chart being prevented from recording attack that Fig. 1 is one embodiment of the invention.
The present embodiment is the voiceprint authentication method described from server side, according to user speech, the server of terminal feedback
The character combination generated and the pronunciation rule of character carry out voiceprint, and the present embodiment can prevent recording to a certain extent
Attack.
Concrete, it is possible to the voiceprint authentication method preventing recording from attacking comprises the steps:
Step 101: ask to generate character combination and the pronunciation rule of character according to the voiceprint of a user；
Character combination includes but not limited to letter, numeral, Chinese character etc., and the pronunciation rule of character includes but not limited to pronunciation
Tone, the length etc. of pronunciation, in an embodiment, the corresponding pronunciation rule of each character in character combination, separately
In one embodiment, the corresponding pronunciation rule of two characters in character combination, the present invention is to character combination and character group
The concrete form of the pronunciation rule of the character in conjunction does not limits.
In the application one embodiment, the pronunciation rule of described character combination and character is randomly generated.
Step 102: the pronunciation rule of character combination and character is sent to requesting terminal；
Terminal of the present invention includes but not limited to mobile phone, PAD, computer and notebook.
Step 103: receive user's language that described requesting terminal inputs according to the pronunciation rule of described character combination and character
Sound；
Step 104: carry out voiceprint according to the pronunciation rule of described user speech, described character combination and character；
Step 105: described voiceprint result is sent to described requesting terminal.
In the present embodiment, even if assailant can obtain phonetic characters information, also cannot obtain the pronunciation rule of character,
By adding the certification of pronunciation rule, it is possible to effectively prevent recording from attacking.
Specifically, step 104 farther includes:
Judge that whether the voice of described user speech and described user's history input is the sound of same people；
Judge that the character in described user speech is the most identical with the character in described character combination；
Judge whether the articulation type of the character in described user speech mates with the pronunciation rule of described character；
The most described user speech is the character in same people, described user speech with the voice of described user's history input
The articulation type of character in and described user speech identical with the character in described character combination and sending out of described character
When sound rule match meets simultaneously, voiceprint just passes through, and other situation voiceprints do not pass through, the most described user
Voice is not the character in same people, and/or described user speech and described word with the voice of described user's history input
Character in symbol combination is different, and/or the pronunciation rule of the articulation type of the character in described user speech and described character
Do not mate, then voiceprint does not passes through.
The present invention is not limiting as the order of above-mentioned judge process, and the combination of any order all can realize sentencing of voiceprint
Disconnected.
Preferably, as in figure 2 it is shown, step 104 farther includes:
Step 201: first judge that whether the voice of described user speech and described user's history input is the sound of same people
Sound；If not being the sound of same people, then voiceprint does not passes through, and if the sound of same people, continues step
202；
When being embodied as, before carrying out step 202, need first to separate, according to character, the user speech sent in client,
Then the character in user speech is extracted.
Step 202: judge that the character in described user speech is the most identical with the character in described character combination；
If the character in described user speech is different from the character in described character combination, then voiceprint is not by i.e.
Voiceprint failure；
If the character in described user speech is identical with the character in described character combination, then continue step 203；
Step 203: judge that whether the pronunciation rule of the articulation type of character in described user speech and described character
Join；
If the articulation type of the character in described user speech does not mates with the pronunciation rule of described character, then vocal print is recognized
Card does not passes through；
If the articulation type of the character in described user speech mates with the pronunciation rule of described character, then voiceprint
Pass through.
Carrying out voiceprint according to the order described in the present embodiment and can accelerate the speed of certification, it is same that prevention recording is attacked
The experience effect of Shi Tigao user.In below embodiment, if do not done specified otherwise, all suitable according to described in the present embodiment
Sequence carries out voiceprint.
Refer to Fig. 2 again, it is judged that the voice that described user speech and described user's history input is same people, described
Character in user speech is identical with the character in described character combination and the pronunciation side of character in described user speech
Formula also includes after mating with the pronunciation rule of described character being stored by user speech to history sound bank, it is simple to follow-up tune
Take the voice messaging of family input.
As it is shown on figure 3, in the application one embodiment, it is judged that the language that described user speech inputs with described user's history
Sound is that the character in same people, described user speech is identical with the character in described character combination and in described user speech
The articulation type of character mate with the pronunciation rule of described character after also include:
Step 204: judge that described user speech is the most consistent with described user voice in history sound bank；
If described user speech is consistent with described user voice in history sound bank, then voiceprint does not passes through；
If described user speech is inconsistent with described user voice in history sound bank, then voiceprint passes through,
Described user speech is stored to history sound bank.
The most consistent with the voice of this user in history sound bank by checking user speech, it is possible to prevent same user
Not homogeneous voice authentication in input same subscriber voice occur recording attack.
In one embodiment of the invention, the step 204 of a upper embodiment farther includes:
Extract the characteristic parameter of described user speech；
Calculate the characteristic parameter of described user speech and the characteristic parameter of described user voice in historical data base
Euclidean distance, when described Euclidean distance is less than predetermined threshold, described user speech and described user are in history
Voice in sound bank is consistent, when described Euclidean distance is more than predetermined threshold, and described user speech and described user
Voice in history sound bank is inconsistent.
Predetermined threshold described in the present embodiment can send the diversity of same sound according to people and determine.
When being embodied as, it is judged that the detailed mistake that user speech is the most consistent with described user voice in history sound bank
Cheng Wei:
1) by character, user speech is divided into multistage voice, every section of voice is carried out pretreatment, including framing, pre-add
Weight, windowing etc. process, and obtain the one section of sound that can calculate further.
2) beginning and end of efficient voice part in every section of voice is found.
As shown in Figure 4, Fig. 4 is the oscillogram of the pronunciation correspondence of digital " 0 ", as seen from Figure 4 at sound
Front and back have a lot without segment or trickle noise segment.If not removing these invalid acoustical signals, assailant
Can carry out processing at the invalid sound end of recording and affect the effect of recording detection.
When being embodied as, the beginning and end of voice live part can be judged by short-time energy and short-time zero-crossing rate.
Wherein short-time energy refers to the intensity sum of a frame voice signal, the short-time energy En of n-th frame voice signal:
Wherein, m is n-th frame m-th sampled point, and N is the size of this frame, xnM () is the sampling of n-th frame m-th
Frequency after some normalization.
Short-time zero-crossing rate refers at a frame voice signal waveform through the number of times of transverse axis, is designated as Zn,
Wherein, m is n-th frame m-th sampled point, and N is the size of this frame, xnM () is the sampling of n-th frame m-th
Frequency after some normalization.
When short-time energy En exceedes threshold values E or short-time zero-crossing rate Zn exceedes threshold values Z, this voice is effective language
The beginning of sound, when short-time energy En is less than threshold values Z less than threshold values E or short-time zero-crossing rate Zn, this voice is
The end of efficient voice.
3) use Mel yardstick cepstrum coefficient (MFCC) that efficient voice is extracted characteristic parameter.The method is current sound
Characteristic parameter extraction way more common in process, here is omitted for the present invention.
Record user this walk pretreatment through first three, voice inactive portion is fallen in segmentation and extracts after characteristic parameter, user
The voice of certain character be expressed as T:
T have N frame vector T (1), T (2) ... T (n) ..., T (N) }, T (n) is the speech characteristic vector of n-th frame.
Carry out same pretreatment for the character sound of this user in history library, voice inactive portion is fallen in segmentation and extracts spy
Levying parameter postscript is R:
R has M frame vector R={R (1), R (2) ... R (m) ..., R (M) }, R (m) is that the voice of m frame is special
Levy vector.
4) calculate user voice and the similarity of the sound of storage in history sound bank, be and calculate the similar of T to R
Property, this similarity judges can be by calculating the Euclidean distance of T and R.
d(T(in),R(im)) represent in T i-thnFrame feature and i in RmEuclidean distance between frame feature, if two
Individual waveform is completely superposed at certain frame, then distance d is 0.In order to compare the similarity between them, can be calculated it
Distance D [T, R] between, the least then similarity of distance is the highest.
If N=M, i.e. two sections voice length are identical, direct simple computation user speech and storage in history sound bank
Euclidean distance D [T, R] of voice=d (1,1)+d (2,2)+...+d (N, N), if two ends voice is just the same, then
D [T, R]=0, may only judge that T with R is the most identical in this way, but recording assailant is in reality
Attack often taked original recording is stretched at portion, shorten or the operation such as deletion, if so
Both simple computation distance can not well defend this type of to attack.
When N and M is differed, consider to align T (n) and R (m).Alignment can use the side of linear expansion
Method, if N < M can by the sequence that T Linear Mapping is a M frame, then calculate it with R (1), R (2) ...,
R (M) } between distance.But whole section of sound will not be processed by assailant, and often only part to sound
Position processes, if taking the method will recognise that, the two assonance degree is the lowest.
Therefore the similarity comparing voice T and R needs to combine time rule and range measurement, by finding
Function im=Φ (in), the time shaft n of T is non-linearly mapped on the time shaft m of R, and makes this T's Yu R
Distance D [T, R] meets:
Wherein:
Φ(in+1)≥Φ(in)
Φ(in+1)-Φ(in)≤1
It is evident that meet the condition of dynamic programming, it is possible to use dynamic programming algorithm solves, Qi Zhongdong
State planning multinomial is:
D (in, im)=d (T (in), R (im))+min{D (in-1, im), D (in-1, im-1), D (in-1, im-2) }
(making D (1,1)=0) search of setting out, recursion repeatedly is so put from (l, 1), until (N, M) can be obtained by optimal path,
And D (N, M) is exactly the matching distance corresponding to best matching path.
Owing to everyone speech is affected by many factors, anyone repeats to send out the sound of identical characters can not on sound wave
Can be the most similar, the property of there are differences certainly, defining this diversity is the reservation threshold judged.If D (N, M)=0,
Then explanation two ends voice T and R is completely the same, may certify that for voice T and R be a sound, it is understood that there may be record
Sound is attacked；If D (N, M) < threshold values, then explanation two ends voice T and R similarity degree are the highest, it is equally possible to exist
Recording is attacked；If D (N, M) >=threshold values, then explanation T and R is not same voice, there is not recording and attacks.
The present invention propose be prevented from recording attack voiceprint authentication method, by checking user speech in character and
Articulation type is the most consistent with the pronunciation rule of the character combination that server generates and character, it is possible to effectively prevent recording
Attacking, assailant is i.e. enabled the user speech got by other channels and meets voice content, also cannot meet pronunciation
The requirement of mode.Further, the user speech in order to prevent user from repeatedly inputting is attacked by recording, it is judged that use
After character in the voice of family is consistent with the pronunciation rule of the character combination that server generates and character with articulation type, also sentence
Disconnected voice the most to be verified is the most consistent with the voice of this user in history sound bank, if consistent, illustrates to there is record
Sound is attacked.The present invention can effectively prevent the recording in voiceprint from attacking.
As it is shown in figure 5, the voiceprint authentication method flow chart being prevented from recording attack that Fig. 5 is one embodiment of the invention.
The method is the description carried out from side, requesting terminal, and concrete, voiceprint authentication method includes:
Step 501: the voiceprint sending a user is asked to server；
Step 502: receive and show character combination and the pronunciation rule of character that described server sends；
Step 503: receive the user speech that user inputs according to the pronunciation rule of described character combination and character；
Step 504: described user speech is sent to described server；
Step 505: receive the voiceprint result that described server sends.
As shown in Figure 6, Fig. 6 is that a kind of of one embodiment of the invention is prevented from the voiceprint server that recording is attacked,
This server 600 includes, signal generating unit 601, for generating sending out of character combination and character according to the request of a user
Sound rule；
Transmitting element 602, for being sent to requesting terminal by the pronunciation rule of described character combination and character, by vocal print
Authentication result sends to described requesting terminal；
Receive unit 603, input according to the pronunciation rule of described character combination and character for receiving described requesting terminal
User speech；
Sound detection unit 604, for entering according to the pronunciation rule of described user speech, described character combination and character
Row voiceprint.
As it is shown in fig. 7, the voiceprint terminal being prevented from recording attack that Fig. 7 is one embodiment of the invention.Specifically
, this certification terminal 700 includes: request unit 701, asks to server for sending the voiceprint of a user；
Receive unit 702, for receiving and show character combination and the pronunciation rule of character that described server sends,
Receive the voiceprint result that described server sends；
Typing unit 703, for receiving user's language that user inputs according to the pronunciation rule of described character combination and character
Sound；
Transmitting element 704, for sending described user speech to described server.
As shown in Figure 8, Fig. 8 is the voiceprint authentication system being prevented from recording attack of one embodiment of the invention.
This voiceprint authentication system includes server 600 and requesting terminal 700, and wherein, described server 600 is for root
Ask to generate character combination and the pronunciation rule of character according to the voiceprint of a user；By described character combination and character
Pronunciation rule is sent to requesting terminal；Receive described requesting terminal defeated according to the pronunciation rule of described character combination and character
The user speech entered；Pronunciation rule according to described user speech, described character combination and character carries out voiceprint；
Described voiceprint result is sent to described requesting terminal；
Described requesting terminal 700 is asked to server for the voiceprint sending a user；Receive and show described clothes
The character combination of business device transmission and the pronunciation rule of character；Receive user to advise according to the pronunciation of described character combination and character
The user speech then inputted；Described user speech is sent to described server；Receive the vocal print that described server sends
Authentication result.
What the present invention proposed is prevented from voiceprint authentication method, server, terminal and the system that recording is attacked, by testing
Character combination that character in card user speech and articulation type and server generate and the pronunciation rule of character whether one
Cause, it is possible to effectively preventing recording from attacking, assailant is i.e. enabled the user speech got by other channels and meets language
Sound content, also cannot meet the requirement of articulation type.Further, for the user speech preventing user from repeatedly inputting
Attacked by recording, it is judged that character combination that character in user speech and articulation type generate with server and character
Pronunciation rule consistent after, also judge voice the most to be verified and the voice of this user in history sound bank whether
Cause, if consistent, illustrate that there is recording attacks.The present invention can effectively prevent the recording in voiceprint from attacking.
For the technical scheme of clearer explanation the application, illustrate with a specific embodiment below, in conjunction with Fig. 9
Shown in, the working-flow preventing recording from attacking is:
Step 901: client sends ID authentication request to server；
Step 902: server receives ID authentication request；
Step 903: server is according to the checking character combination of ID authentication request stochastic generation and the pronunciation side of character
Formula, and send it to client；
Step 904: after client receives the pronunciation rule of character combination to be verified that server issues and character, carry
Show that user reads in character on request；
Step 905: client receives the user speech that user is read in, and user speech user read in sends to clothes
Business device；
Step 906: server carries out voice print verification, it is judged that the language of the user speech of reception and this user prestored
Whether sound is same people, can use the most conventional voice print verification algorithm when being embodied as；
If voice print verification is not same person, the most directly return user authentication failure to client；
If voice print verification is same people, then continue recording detection；
Step 907: the character in the character combination that the character in checking user voice and server generate is the most identical；
If the character in the character combination that the character in user voice and server generate differs, then the word in user voice
Symbol checking is not passed through, and returns user authentication failure to client；If the character in user voice generates with server
Character in character combination is identical, then the character in user voice is verified, and continues step 908；
Step 908: whether the character sound mode that the articulation type of the character in checking user voice and server generate
Identical, if the character sound mode that the articulation type of the character in user voice and server generate differs, then use
Character sound mode in the sound of family is verified and is not passed through, and returns user authentication failure to client；If in user voice
The articulation type of character identical with the character sound mode that server generates, then the character sound mode in user voice
It is verified, continues step 909；
Step 909: whether checking user voice is present in history sound bank, if it is present prove to there is recording
Attack, authentification failure, authentification failure result is sent to client；If it does not exist, then voiceprint passes through, will
User voice is stored in history sound bank, and by result, voiceprint is sent to client.
Whether checking user voice is present in the process in history sound bank has carried out detailed the most in the above-described embodiments
Illustrating, here is omitted.After voiceprint passes through, client continues corresponding operation, and this is not limited by the present invention
System.
What the present invention proposed is prevented from voiceprint authentication method, server, terminal and the system that recording is attacked, by testing
Character combination that character in card user speech and articulation type and server generate and the pronunciation rule of character whether one
Cause, it is possible to effectively preventing recording from attacking, assailant is i.e. enabled the user speech got by other channels and meets language
Sound content, also cannot meet the requirement of articulation type.Further, for the user speech preventing user from repeatedly inputting
Attacked by recording, it is judged that character combination that character in user speech and articulation type generate with server and character
Pronunciation rule consistent after, also judge voice the most to be verified and the voice of this user in history sound bank whether
Cause, if consistent, illustrate that there is recording attacks.The present invention can effectively prevent the recording in voiceprint from attacking.
The above is merely to illustrate technical scheme, and any those of ordinary skill in the art all can be without prejudice to this
Under the spirit and the scope of invention, above-described embodiment is modified and changes.Therefore, the scope of the present invention
Should be as the criterion depending on right.
Claims (11)
1. one kind is prevented from the voiceprint authentication method that recording is attacked, it is characterised in that include,
Voiceprint request according to a user generates character combination and the pronunciation rule of character；
The pronunciation rule of described character combination and character is sent to requesting terminal；
Receive the user speech that described requesting terminal inputs according to the pronunciation rule of described character combination and character；
Pronunciation rule according to described user speech, described character combination and character carries out voiceprint；By described vocal print
Authentication result sends to described requesting terminal.
It is prevented from the voiceprint authentication method that recording is attacked the most as claimed in claim 1, it is characterised in that according to institute
State the pronunciation rule of user speech, described character combination and character to carry out voiceprint and farther include,
Judge that whether the voice of described user speech and described user's history input is the sound of same people；
Judge that the character in described user speech is the most identical with the character in described character combination；
Judge whether the articulation type of the character in described user speech mates with the pronunciation rule of described character；
The most described user speech is the character in same people, described user speech with the voice of described user's history input
The articulation type of character in and described user speech identical with the character in described character combination and sending out of described character
During sound rule match, voiceprint just passes through, and other situation voiceprints do not pass through.
It is prevented from the voiceprint authentication method that recording is attacked the most as claimed in claim 2, it is characterised in that judge
Described user speech is the character in same people, described user speech and described word with the voice of described user's history input
Character in symbol combination is identical and the pronunciation rule of the articulation type of character in described user speech and described character
Also include after joining,
Described user speech is stored to history sound bank.
It is prevented from the voiceprint authentication method that recording is attacked the most as claimed in claim 2, it is characterised in that judge
Described user speech is the character in same people, described user speech and described word with the voice of described user's history input
Character in symbol combination is identical and the pronunciation rule of the articulation type of character in described user speech and described character
Also include after joining,
Judge that described user speech is the most consistent with described user voice in history sound bank；
If described user speech is consistent with described user voice in history sound bank, then voiceprint does not passes through；
If described user speech is inconsistent with described user voice in history sound bank, then voiceprint passes through,
Described user speech is stored to history sound bank.
It is prevented from the voiceprint authentication method that recording is attacked the most as claimed in claim 4, it is characterised in that judge institute
State that user speech is the most consistent with described user voice in history sound bank to be farther included,
Extract the characteristic parameter of described user speech；
Calculate the characteristic parameter of described user speech and the characteristic parameter of described user voice in historical data base
Euclidean distance, when described Euclidean distance is less than predetermined threshold, described user speech and described user are in history
Voice in sound bank is consistent, when described Euclidean distance is more than predetermined threshold, and described user speech and described user
Voice in history sound bank is inconsistent.
It is prevented from the voiceprint authentication method that recording is attacked the most as claimed in claim 5, it is characterised in that extract institute
The characteristic parameter stating user speech farther includes,
Described user speech is carried out pretreatment, described user speech is divided into multistage voice by character；
Find the beginning and end of efficient voice part in every section of voice；
Extract the characteristic parameter of efficient voice part.
It is prevented from the voiceprint authentication method that recording is attacked the most as claimed in claim 1, it is characterised in that described word
The pronunciation rule of symbol combination and character is randomly generated.
8. one kind is prevented from the voiceprint authentication method that recording is attacked, it is characterised in that include,
The voiceprint sending a user is asked to server；
Receive and show character combination and the pronunciation rule of character that described server sends；
Receive the user speech that user inputs according to the pronunciation rule of described character combination and character；
Described user speech is sent to described server；
Receive the voiceprint result that described server sends.
9. one kind is prevented from the voiceprint server that recording is attacked, it is characterised in that include,
Signal generating unit, generates character combination and the pronunciation rule of character for the request according to a user；
Transmitting element, for being sent to requesting terminal by the pronunciation rule of described character combination and character, by voiceprint
Result sends to described requesting terminal；
Receive unit, for receiving the use that described requesting terminal inputs according to the pronunciation rule of described character combination and character
Family voice；
Sound detection unit, for carrying out sound according to the pronunciation rule of described user speech, described character combination and character
Stricture of vagina certification.
10. one kind is prevented from the voiceprint terminal that recording is attacked, it is characterised in that include,
Request unit, asks to server for sending the voiceprint of a user；
Receive unit, for receiving and show character combination and the pronunciation rule of character that described server sends, receive
The voiceprint result that described server sends；
Typing unit, for receiving the user speech that user inputs according to the pronunciation rule of described character combination and character；
Transmitting element, for sending described user speech to described server.
11. 1 kinds are prevented from the voiceprint authentication system that recording is attacked, it is characterised in that include server and request eventually
End, wherein, described server asks to generate the pronunciation rule of character combination and character for the voiceprint according to a user
Then；The pronunciation rule of described character combination and character is sent to requesting terminal；Receive described requesting terminal according to described
The user speech of the pronunciation rule input of character combination and character；According to described user speech, described character combination and word
The pronunciation rule of symbol carries out voiceprint；Described voiceprint result is sent to described requesting terminal；
Described requesting terminal is asked to server for the voiceprint sending a user；Receive and show described server
The character combination sent and the pronunciation rule of character；Receive user defeated according to the pronunciation rule of described character combination and character
The user speech entered；Described user speech is sent to described server；Receive the voiceprint that described server sends
Result.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511020257.4A CN105933272A (en) | 2015-12-30 | 2015-12-30 | Voiceprint recognition method capable of preventing recording attack, server, terminal, and system |
| PCT/CN2016/111714 WO2017114307A1 (en) | 2015-12-30 | 2016-12-23 | Voiceprint authentication method capable of preventing recording attack, server, terminal, and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511020257.4A CN105933272A (en) | 2015-12-30 | 2015-12-30 | Voiceprint recognition method capable of preventing recording attack, server, terminal, and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105933272A true CN105933272A (en) | 2016-09-07 |
Family
ID=56839979
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511020257.4A Pending CN105933272A (en) | 2015-12-30 | 2015-12-30 | Voiceprint recognition method capable of preventing recording attack, server, terminal, and system |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN105933272A (en) |
| WO (1) | WO2017114307A1 (en) |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017114307A1 (en) * | 2015-12-30 | 2017-07-06 | 中国银联股份有限公司 | Voiceprint authentication method capable of preventing recording attack, server, terminal, and system |
| CN109087647A (en) * | 2018-08-03 | 2018-12-25 | 平安科技（深圳）有限公司 | Application on Voiceprint Recognition processing method, device, electronic equipment and storage medium |
| CN109218269A (en) * | 2017-07-05 | 2019-01-15 | 阿里巴巴集团控股有限公司 | Identity authentication method, device, equipment and data processing method |
| CN109935233A (en) * | 2019-01-29 | 2019-06-25 | 天津大学 | A kind of recording attack detection method based on amplitude and phase information |
| CN110169014A (en) * | 2017-01-03 | 2019-08-23 | 诺基亚技术有限公司 | Device, method and computer program product for certification |
| CN111316668A (en) * | 2017-11-14 | 2020-06-19 | 思睿逻辑国际半导体有限公司 | Detection of loudspeaker playback |
| CN111524528A (en) * | 2020-05-28 | 2020-08-11 | Oppo广东移动通信有限公司 | Voice awakening method and device for preventing recording detection |
| US10984083B2 (en) | 2017-07-07 | 2021-04-20 | Cirrus Logic, Inc. | Authentication of user using ear biometric data |
| CN112735426A (en) * | 2020-12-24 | 2021-04-30 | 深圳市声扬科技有限公司 | Voice verification method and system, computer device and storage medium |
| US11017252B2 (en) | 2017-10-13 | 2021-05-25 | Cirrus Logic, Inc. | Detection of liveness |
| US11023755B2 (en) | 2017-10-13 | 2021-06-01 | Cirrus Logic, Inc. | Detection of liveness |
| US11037574B2 (en) | 2018-09-05 | 2021-06-15 | Cirrus Logic, Inc. | Speaker recognition and speaker change detection |
| US11042617B2 (en) | 2017-07-07 | 2021-06-22 | Cirrus Logic, Inc. | Methods, apparatus and systems for biometric processes |
| US11042618B2 (en) | 2017-07-07 | 2021-06-22 | Cirrus Logic, Inc. | Methods, apparatus and systems for biometric processes |
| US11042616B2 (en) | 2017-06-27 | 2021-06-22 | Cirrus Logic, Inc. | Detection of replay attack |
| CN113012684A (en) * | 2021-03-04 | 2021-06-22 | 电子科技大学 | Synthesized voice detection method based on voice segmentation |
| US11164588B2 (en) | 2017-06-28 | 2021-11-02 | Cirrus Logic, Inc. | Magnetic detection of replay attack |
| US11264037B2 (en) | 2018-01-23 | 2022-03-01 | Cirrus Logic, Inc. | Speaker identification |
| US11270707B2 (en) | 2017-10-13 | 2022-03-08 | Cirrus Logic, Inc. | Analysing speech signals |
| US11276409B2 (en) | 2017-11-14 | 2022-03-15 | Cirrus Logic, Inc. | Detection of replay attack |
| CN114826709A (en) * | 2022-04-15 | 2022-07-29 | 马上消费金融股份有限公司 | Identity authentication and acoustic environment detection method, system, electronic device and medium |
| US11475899B2 (en) | 2018-01-23 | 2022-10-18 | Cirrus Logic, Inc. | Speaker identification |
| US11631402B2 (en) | 2018-07-31 | 2023-04-18 | Cirrus Logic, Inc. | Detection of replay attack |
| US11704397B2 (en) | 2017-06-28 | 2023-07-18 | Cirrus Logic, Inc. | Detection of replay attack |
| US11705135B2 (en) | 2017-10-13 | 2023-07-18 | Cirrus Logic, Inc. | Detection of liveness |
| US11735189B2 (en) | 2018-01-23 | 2023-08-22 | Cirrus Logic, Inc. | Speaker identification |
| US11748462B2 (en) | 2018-08-31 | 2023-09-05 | Cirrus Logic Inc. | Biometric authentication |
| US11755701B2 (en) | 2017-07-07 | 2023-09-12 | Cirrus Logic Inc. | Methods, apparatus and systems for authentication |
| US11829461B2 (en) | 2017-07-07 | 2023-11-28 | Cirrus Logic Inc. | Methods, apparatus and systems for audio playback |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109754817A (en) * | 2017-11-02 | 2019-05-14 | 北京三星通信技术研究有限公司 | signal processing method and terminal device |
| CN112365895B (en) * | 2020-10-09 | 2024-04-19 | 深圳前海微众银行股份有限公司 | Audio processing method, device, computing equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1808567A (en) * | 2006-01-26 | 2006-07-26 | 覃文华 | Voice-print authentication device and method of authenticating people presence |
| CN102457845A (en) * | 2010-10-14 | 2012-05-16 | 阿里巴巴集团控股有限公司 | Wireless service identity authentication method, equipment and system |
| CN102543084A (en) * | 2010-12-29 | 2012-07-04 | 盛乐信息技术（上海）有限公司 | Online voiceprint recognition system and implementation method thereof |
| CN102737634A (en) * | 2012-05-29 | 2012-10-17 | 百度在线网络技术（北京）有限公司 | Authentication method and device based on voice |
| CN104717219A (en) * | 2015-03-20 | 2015-06-17 | 百度在线网络技术（北京）有限公司 | Vocal print login method and device based on artificial intelligence |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9202460B2 (en) * | 2008-05-14 | 2015-12-01 | At&T Intellectual Property I, Lp | Methods and apparatus to generate a speech recognition library |
| CN104901808A (en) * | 2015-04-14 | 2015-09-09 | 时代亿宝(北京)科技有限公司 | Voiceprint authentication system and method based on time type dynamic password |
| CN105185379B (en) * | 2015-06-17 | 2017-08-18 | 百度在线网络技术（北京）有限公司 | voiceprint authentication method and device |
| CN105096121B (en) * | 2015-06-25 | 2017-07-25 | 百度在线网络技术（北京）有限公司 | voiceprint authentication method and device |
| CN105933272A (en) * | 2015-12-30 | 2016-09-07 | 中国银联股份有限公司 | Voiceprint recognition method capable of preventing recording attack, server, terminal, and system |
-
2015
- 2015-12-30 CN CN201511020257.4A patent/CN105933272A/en active Pending
-
2016
- 2016-12-23 WO PCT/CN2016/111714 patent/WO2017114307A1/en active Application Filing
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1808567A (en) * | 2006-01-26 | 2006-07-26 | 覃文华 | Voice-print authentication device and method of authenticating people presence |
| CN102457845A (en) * | 2010-10-14 | 2012-05-16 | 阿里巴巴集团控股有限公司 | Wireless service identity authentication method, equipment and system |
| CN102543084A (en) * | 2010-12-29 | 2012-07-04 | 盛乐信息技术（上海）有限公司 | Online voiceprint recognition system and implementation method thereof |
| CN102737634A (en) * | 2012-05-29 | 2012-10-17 | 百度在线网络技术（北京）有限公司 | Authentication method and device based on voice |
| CN104717219A (en) * | 2015-03-20 | 2015-06-17 | 百度在线网络技术（北京）有限公司 | Vocal print login method and device based on artificial intelligence |
Non-Patent Citations (1)
| Title |
|---|
| 赵力: "《语音信号处理》", 31 May 2009 * |
Cited By (36)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017114307A1 (en) * | 2015-12-30 | 2017-07-06 | 中国银联股份有限公司 | Voiceprint authentication method capable of preventing recording attack, server, terminal, and system |
| US11283631B2 (en) | 2017-01-03 | 2022-03-22 | Nokia Technologies Oy | Apparatus, method and computer program product for authentication |
| CN110169014A (en) * | 2017-01-03 | 2019-08-23 | 诺基亚技术有限公司 | Device, method and computer program product for certification |
| US11042616B2 (en) | 2017-06-27 | 2021-06-22 | Cirrus Logic, Inc. | Detection of replay attack |
| US12026241B2 (en) | 2017-06-27 | 2024-07-02 | Cirrus Logic Inc. | Detection of replay attack |
| US11704397B2 (en) | 2017-06-28 | 2023-07-18 | Cirrus Logic, Inc. | Detection of replay attack |
| US11164588B2 (en) | 2017-06-28 | 2021-11-02 | Cirrus Logic, Inc. | Magnetic detection of replay attack |
| CN109218269A (en) * | 2017-07-05 | 2019-01-15 | 阿里巴巴集团控股有限公司 | Identity authentication method, device, equipment and data processing method |
| US11042618B2 (en) | 2017-07-07 | 2021-06-22 | Cirrus Logic, Inc. | Methods, apparatus and systems for biometric processes |
| US10984083B2 (en) | 2017-07-07 | 2021-04-20 | Cirrus Logic, Inc. | Authentication of user using ear biometric data |
| US11755701B2 (en) | 2017-07-07 | 2023-09-12 | Cirrus Logic Inc. | Methods, apparatus and systems for authentication |
| US11714888B2 (en) | 2017-07-07 | 2023-08-01 | Cirrus Logic Inc. | Methods, apparatus and systems for biometric processes |
| US11042617B2 (en) | 2017-07-07 | 2021-06-22 | Cirrus Logic, Inc. | Methods, apparatus and systems for biometric processes |
| US11829461B2 (en) | 2017-07-07 | 2023-11-28 | Cirrus Logic Inc. | Methods, apparatus and systems for audio playback |
| US12135774B2 (en) | 2017-07-07 | 2024-11-05 | Cirrus Logic Inc. | Methods, apparatus and systems for biometric processes |
| US11270707B2 (en) | 2017-10-13 | 2022-03-08 | Cirrus Logic, Inc. | Analysing speech signals |
| US11023755B2 (en) | 2017-10-13 | 2021-06-01 | Cirrus Logic, Inc. | Detection of liveness |
| US11705135B2 (en) | 2017-10-13 | 2023-07-18 | Cirrus Logic, Inc. | Detection of liveness |
| US11017252B2 (en) | 2017-10-13 | 2021-05-25 | Cirrus Logic, Inc. | Detection of liveness |
| US11051117B2 (en) | 2017-11-14 | 2021-06-29 | Cirrus Logic, Inc. | Detection of loudspeaker playback |
| CN111316668A (en) * | 2017-11-14 | 2020-06-19 | 思睿逻辑国际半导体有限公司 | Detection of loudspeaker playback |
| US11276409B2 (en) | 2017-11-14 | 2022-03-15 | Cirrus Logic, Inc. | Detection of replay attack |
| CN111316668B (en) * | 2017-11-14 | 2021-09-28 | 思睿逻辑国际半导体有限公司 | Detection of loudspeaker playback |
| US11264037B2 (en) | 2018-01-23 | 2022-03-01 | Cirrus Logic, Inc. | Speaker identification |
| US11694695B2 (en) | 2018-01-23 | 2023-07-04 | Cirrus Logic, Inc. | Speaker identification |
| US11735189B2 (en) | 2018-01-23 | 2023-08-22 | Cirrus Logic, Inc. | Speaker identification |
| US11475899B2 (en) | 2018-01-23 | 2022-10-18 | Cirrus Logic, Inc. | Speaker identification |
| US11631402B2 (en) | 2018-07-31 | 2023-04-18 | Cirrus Logic, Inc. | Detection of replay attack |
| CN109087647A (en) * | 2018-08-03 | 2018-12-25 | 平安科技（深圳）有限公司 | Application on Voiceprint Recognition processing method, device, electronic equipment and storage medium |
| US11748462B2 (en) | 2018-08-31 | 2023-09-05 | Cirrus Logic Inc. | Biometric authentication |
| US11037574B2 (en) | 2018-09-05 | 2021-06-15 | Cirrus Logic, Inc. | Speaker recognition and speaker change detection |
| CN109935233A (en) * | 2019-01-29 | 2019-06-25 | 天津大学 | A kind of recording attack detection method based on amplitude and phase information |
| CN111524528A (en) * | 2020-05-28 | 2020-08-11 | Oppo广东移动通信有限公司 | Voice awakening method and device for preventing recording detection |
| CN112735426A (en) * | 2020-12-24 | 2021-04-30 | 深圳市声扬科技有限公司 | Voice verification method and system, computer device and storage medium |
| CN113012684A (en) * | 2021-03-04 | 2021-06-22 | 电子科技大学 | Synthesized voice detection method based on voice segmentation |
| CN114826709A (en) * | 2022-04-15 | 2022-07-29 | 马上消费金融股份有限公司 | Identity authentication and acoustic environment detection method, system, electronic device and medium |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2017114307A1 (en) | 2017-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105933272A (en) | Voiceprint recognition method capable of preventing recording attack, server, terminal, and system | |
| Yu et al. | Spoofing detection in automatic speaker verification systems using DNN classifiers and dynamic acoustic features | |
| Reynolds | An overview of automatic speaker recognition technology | |
| CN107104803B (en) | User identity authentication method based on digital password and voiceprint joint confirmation | |
| Mukhopadhyay et al. | All your voices are belong to us: Stealing voices to fool humans and machines | |
| CN105913855B (en) | A kind of voice playback attack detecting algorithm based on long window scale factor | |
| US11979398B2 (en) | Privacy-preserving voiceprint authentication apparatus and method | |
| Reynolds | Automatic speaker recognition: Current approaches and future trends | |
| Chen et al. | Towards understanding and mitigating audio adversarial examples for speaker recognition | |
| CN105933323B (en) | Voiceprint registration, authentication method and device | |
| WO2017162053A1 (en) | Identity authentication method and device | |
| CN1808567A (en) | Voice-print authentication device and method of authenticating people presence | |
| JPWO2005013263A1 (en) | Voice authentication system | |
| CN110459226A (en) | A method of voice is detected by vocal print engine or machine sound carries out identity veritification | |
| Turner et al. | Attacking speaker recognition systems with phoneme morphing | |
| CN110379433A (en) | Method, apparatus, computer equipment and the storage medium of authentication | |
| CN109273012A (en) | A kind of identity identifying method based on Speaker Identification and spoken digit recognition | |
| Yuan et al. | Overview of the development of speaker recognition | |
| Al-Shayea et al. | Speaker identification: A novel fusion samples approach | |
| Reynolds et al. | Automatic speaker recognition | |
| EP4170526A1 (en) | An authentication system and method | |
| RU2351023C2 (en) | User verification method in authorised access systems | |
| CN113012684B (en) | Synthesized voice detection method based on voice segmentation | |
| Chen et al. | Personal threshold in a small scale text-dependent speaker recognition | |
| Mishra et al. | Speaker identification, differentiation and verification using deep learning for human machine interface |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into 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: 20160907 |