CN107086983A - A kind of intelligent automobile voice quick registration method and device - Google Patents
A kind of intelligent automobile voice quick registration method and device Download PDFInfo
- Publication number
- CN107086983A CN107086983A CN201710124156.4A CN201710124156A CN107086983A CN 107086983 A CN107086983 A CN 107086983A CN 201710124156 A CN201710124156 A CN 201710124156A CN 107086983 A CN107086983 A CN 107086983A
- Authority
- CN
- China
- Prior art keywords
- vocal print
- print feature
- user
- memory cell
- extraction
- 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 17
- 230000001755 vocal effect Effects 0.000 claims abstract description 162
- 238000000605 extraction Methods 0.000 claims abstract description 54
- 239000000284 extract Substances 0.000 claims description 10
- 238000000151 deposition Methods 0.000 claims description 3
- 238000005457 optimization Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012067 mathematical method Methods 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 210000000867 larynx Anatomy 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 210000004072 lung Anatomy 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 210000003928 nasal cavity Anatomy 0.000 description 1
- 230000035479 physiological effects, processes and functions Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 210000002105 tongue Anatomy 0.000 description 1
- 210000000515 tooth Anatomy 0.000 description 1
- 238000012795 verification 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
- H04L63/0815—Network architectures or network communication protocols for network security for authentication of entities providing single-sign-on or federations
-
- 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
- G10L15/1807—Speech classification or search using natural language modelling using prosody or stress
-
- 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
- G10L15/1815—Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
-
- 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
- G10L15/28—Constructional details of speech recognition systems
- G10L15/34—Adaptation of a single recogniser for parallel processing, e.g. by use of multiple processors or cloud computing
-
- 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
- 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
-
- 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
-
- 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
- H04L63/0876—Network architectures or network communication protocols for network security for authentication of entities based on the identity of the terminal or configuration, e.g. MAC address, hardware or software configuration or device fingerprint
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Acoustics & Sound (AREA)
- Computing Systems (AREA)
- Human Computer Interaction (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- General Engineering & Computer Science (AREA)
- Computer Security & Cryptography (AREA)
- Computer Hardware Design (AREA)
- Artificial Intelligence (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Mathematical Physics (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Power Engineering (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Telephonic Communication Services (AREA)
Abstract
The present invention relates to a kind of intelligent automobile voice quick registration method and device, when user is logging in, obtain vocal print feature, and the device id or the device id of mobile terminal of the on-vehicle host that user provides are obtained simultaneously, the device id of the on-vehicle host provided according to user or the device id of mobile terminal confirm memory cell in database beyond the clouds, when confirming memory cell of the device id beyond the clouds in database, carry out carrying out matching retrieval to the vocal print feature of extraction in the memory cell, when the match is successful for the vocal print feature, then login successfully, otherwise login failure.The retrieval match time that classification storage greatly reduces volume vocal print feature is carried out to user profile by device id, and login can just be realized by the vocal print feature of user, logged in without inputting user name and login password, add the login safety of the car-mounted terminals such as car running computer, also user is greatly facilitated to be logged in, quick identification user identity, realizes quick registration simultaneously.
Description
Technical field
The present invention relates to intelligent automobile technical field, more particularly to a kind of intelligent automobile voice quick registration method and dress
Put.
Background technology
With continuing to develop for automotive engineering, present automobile is increasingly to energy-conserving and environment-protective direction and intelligent interconnection direction
Function in development, automobile also becomes increasingly complex.And in order to preferably manage these functions, and be easy to interact with driver,
The car-mounted terminals such as car running computer would generally be provided with automobile.The login of the car-mounted terminals such as car running computer, generally passes through control
Key-press input or touch-screen input.
In the prior art, the checking of user is mainly verified by user name with logining password.In some realization sides
In case, also have by speech recognition to realize the input of control instruction.The scheme of the existing control instruction of present speech recognition is main
For:Control instruction collection is prestored in systems, when in use, the user speech collected from sound acquisition module, and
System is digitized processing, extracts the characteristic value of voice, then the condition code that the voice command stored with system is concentrated is carried out
Matching verification, identifies order, and rear and other controllers carry out information interchange and perform action.
During the present invention is realized, inventor has found that there are the following problems in the prior art, by user name and stepping on
Enter password to carry out checking and log in easily password occur and forgetting, the generation situations such as leakage of user name and login password.It is existing
In the method for the Voice command of car-mounted terminal, voice control command data are all based on a speech database and are stored and adjusted
With causing system huge, renewal speed is slow, cost is high, system flexibility is low, because the enrollment of user vehicle constantly increases
Plus, the user data stored is very huge, and time-consuming for existing speech recognition schemes, it is impossible in a short time according to user
Pronunciation character, identify the identity of user.
The content of the invention
For this reason, it may be necessary to provide one kind can increase the car-mounted terminals such as car running computer log in conveniently, safely user logged in
With the intelligent automobile voice quick registration method and device of quick identification user identity.
To achieve the above object, a kind of intelligent automobile voice quick registration method is inventor provided, including:
In user's registration, user speech information and device id are obtained, the device id includes the device id of on-vehicle host
With the device id of mobile terminal;According to the device id corresponding memory cell of Database beyond the clouds, and from the user speech
The vocal print feature of user is extracted in information, the vocal print feature of user is associatedly stored to the memory cell corresponding to device id;
In user's registration, user speech information and device id are obtained, the device id includes the device id of on-vehicle host
With the device id of mobile terminal;According to the device id corresponding memory cell of Database beyond the clouds, and from the user speech
The vocal print feature of user is extracted in information, the vocal print feature of user is associatedly stored to the memory cell corresponding to device id
According to the device id corresponding memory cell of database retrieval device id, and the progress pair in the memory cell beyond the clouds
The vocal print feature of extraction carries out retrieval matching, and when the vocal print feature of extraction, the match is successful, then logged in.
Further optimization, the user speech information is sound pick-up by on-vehicle host or the mobile terminal of user
Microphone is obtained, and the user speech information and device id of acquisition are sent to cloud server;
Parsing is carried out to voice messaging by cloud server and extracts vocal print feature, according in device id beyond the clouds database
It is determined that the memory cell of storage vocal print feature, vocal print feature retrieval matching is carried out from memory cell, if the match is successful, is carried out
User logs in, if it fails to match, refuses the User logs in.
Further optimization, the vocal print feature code includes acoustic feature, lexical characteristics and prosodic features.
Further optimization, it is described " and to carry out carrying out retrieval matching to the vocal print feature of extraction in the memory cell, when carrying
The match is successful for the vocal print feature taken " specifically include:
By the way that the vocal print feature of extraction is compared one by one with the vocal print feature in the memory cell, when the memory cell
Middle vocal print feature of the vocal print feature similar value more than setting value existed with extraction, then the match is successful for the vocal print feature extracted, if
It is not present, then it fails to match for the vocal print feature extracted.
Further optimization, in user's registration, is pronounced, so as to obtain by user to the special sound content of push
The vocal print feature of user, and store into memory cell.
The present inventor additionally provides a kind of intelligent automobile voice quick registration device of another technical scheme, including registration list
Member, extraction unit and login unit;
The registering unit is used in user's registration, obtains user speech information and device id, and the device id includes
The device id of on-vehicle host and the device id of mobile terminal, and the corresponding storage of Database is single beyond the clouds according to device id
Member, and from the user speech information extract user vocal print feature, the vocal print feature of user is associatedly stored to equipment
Memory cell corresponding to ID;
The extraction unit is used in User logs in, obtains device id or the shifting of user speech information and on-vehicle host
The device id of dynamic terminal, and parsing extraction vocal print feature is carried out to the user speech information;
The login unit is used for according to the device id corresponding memory cell of database retrieval device id, and at this beyond the clouds
Carry out carrying out retrieval matching to the vocal print feature of extraction in memory cell, when the vocal print feature of extraction, the match is successful, then stepped on
Record.
Further optimization, the user speech information is sound pick-up by on-vehicle host or the mobile terminal of user
Microphone is obtained, and the user speech information and device id of acquisition are sent to cloud server;
The extraction unit carries out parsing to voice messaging by cloud server and extracts vocal print feature;
The unit that logs in is according to the memory cell that storage vocal print feature is determined in device id beyond the clouds database, from storage
Vocal print feature retrieval matching is carried out in unit, if the match is successful, User logs in is carried out, if it fails to match, refuses the user
Log in.
Further optimization, the vocal print feature code includes acoustic feature, lexical characteristics and prosodic features.
Further optimization, it is described " and to carry out carrying out retrieval matching to the vocal print feature of extraction in the memory cell, when carrying
The match is successful for the vocal print feature taken " specifically include:
By the way that the vocal print feature of extraction is compared one by one with the vocal print feature in the memory cell, when the memory cell
Middle vocal print feature of the vocal print feature similar value more than setting value existed with extraction, then the match is successful for the vocal print feature extracted, if
It is not present, then it fails to match for the vocal print feature extracted.
Further optimization, the registering unit is additionally operable in user's registration, by user in the special sound of push
Hold to be pronounced, so as to obtain the vocal print feature of user, and store into memory cell.
Prior art is different from, above-mentioned technical proposal is by user's registration, obtaining the sound provided during user's registration
Information and device id, and the vocal print feature that user is extracted in parsing is carried out to acoustic information, while the device id provided according to user
Advise memory cell corresponding with device id in database beyond the clouds, the vocal print feature of the user of extraction is then stored in this and deposited
In storage unit, when user is logging in, the voice messaging that identification user provides, parsing obtains vocal print feature, and obtains simultaneously
The device id or the device id of mobile terminal of the on-vehicle host of family offer, the equipment of the on-vehicle host provided according to user are provided
The device id of ID or mobile terminal confirms memory cell in database beyond the clouds, when confirming the device id beyond the clouds in database
Memory cell, carry out carrying out matching retrieval to the vocal print feature of extraction in the memory cell, when the vocal print feature is matched into
During work(, then login successfully, otherwise login failure.By confirming corresponding memory cell in device id beyond the clouds database, and from
Retrieval vocal print feature matching is carried out in memory cell, classification storage is carried out to user profile by device id greatly reduces volume vocal print
The retrieval match time of feature, and login can just be realized by the vocal print feature of user, without inputting user name and logging in close
Code is logged in, that is, is added the login safety of the car-mounted terminals such as car running computer, also greatly facilitated user and logged in, simultaneously
Quick identification user identity, realizes quick registration.
Brief description of the drawings
Fig. 1 is a kind of schematic flow sheet of intelligent automobile voice quick registration method described in embodiment;
Fig. 2 is a kind of structural representation of intelligent automobile voice quick registration device described in embodiment.
Description of reference numerals:
210th, registering unit, 220, extraction unit, 230, log in unit, 240, memory cell.
Embodiment
To describe the technology contents of technical scheme in detail, feature, the objects and the effects being constructed, below in conjunction with specific reality
Apply example and coordinate accompanying drawing to be explained in detail.
Referring to Fig. 1, the present embodiment intelligent automobile voice quick registration method, including:
Step S110:In user's registration, user speech information and device id are obtained, the device id includes on-vehicle host
Device id and mobile terminal device id;According to the device id corresponding memory cell of Database beyond the clouds, and from described
The vocal print feature of user is extracted in user speech information, the vocal print feature of user is associatedly stored to depositing corresponding to device id
Storage unit;
Step S120:In User logs in, the device id or mobile terminal of user speech information and on-vehicle host are obtained
Device id, and to the user speech information carry out parsing extract vocal print feature;
Step S130:According to the device id corresponding memory cell of database retrieval device id, and in the storage list beyond the clouds
Carry out carrying out retrieval matching to the vocal print feature of extraction in member, when the vocal print feature of extraction, the match is successful, then logged in.
Vocal print not only has specificity, and the characteristics of have relative stability.After adult, the sound of people can keep long-term
Stablize relatively constant.It is demonstrated experimentally that no matter talker is deliberately to imitate other people sound and the tone, or whisper in sb.'s ear is softly talked, i.e.,
Make to imitate remarkably true to life, its vocal print is but differed all the time.And Application on Voiceprint Recognition, one kind of biological identification technology.Application on Voiceprint Recognition is just
It is that acoustical signal is converted into electric signal, then is identified with computer.The generation of human language is Body Languages maincenter and pronunciation
A complicated physiology physical process between organ, phonatory organ -- tongue, tooth, larynx, lung, nasal cavity that people uses in speech
Everyone is widely different in terms of size and form, so the voiceprint map of any two people is all variant.
By in user's registration, obtaining the acoustic information and device id provided during user's registration, and acoustic information is entered
The vocal print feature of user is extracted in row parsing, while the device id provided according to user suggestion and device id pair in database beyond the clouds
The vocal print feature of the user of extraction, is then stored in the memory cell by the memory cell answered, when user is logging in,
The voice messaging that user provides is recognized, parsing obtains vocal print feature, and obtains the device id for the on-vehicle host that user provides simultaneously
Or the device id of mobile terminal, the device id of the on-vehicle host provided according to user or the device id of mobile terminal are beyond the clouds
Memory cell is confirmed in database, when confirming memory cell of the device id beyond the clouds in database, is entered in the memory cell
Row carries out matching retrieval to the vocal print feature of extraction, when the match is successful for the vocal print feature, then logins successfully, and otherwise logs in and loses
Lose.By confirming corresponding memory cell, and the progress retrieval vocal print feature from memory cell in device id beyond the clouds database
Matching, carries out classification storage to user profile by device id and greatly reduces the retrieval match time of volume vocal print feature, and pass through
The vocal print feature of user can just realize login, be logged in without inputting user name and login password, that is, add driving electricity
The login safety of the car-mounted terminals such as brain, also greatly facilitates user and is logged in, while quickly recognizing user identity, realizes quick
Log in
In the present embodiment, the user speech information is sound pick-up by on-vehicle host or the mobile terminal of user
Microphone obtain, and the user speech information and device id of acquisition are sent to cloud server;Pass through cloud server pair
Voice messaging carries out parsing and extracts vocal print feature, according to the storage list that storage vocal print feature is determined in device id beyond the clouds database
Member, vocal print feature retrieval matching is carried out from memory cell, if the match is successful, User logs in is carried out, if it fails to match, refuses
The exhausted User logs in.When user carries out login on-vehicle host using new intelligent automobile, pass through the Mike of mobile terminal
Wind obtains voice messaging, and obtains the device id of the mobile terminal of user simultaneously, is counted beyond the clouds by the device id of mobile terminal
According to corresponding memory cell, and progress vocal print feature retrieval matching in the memory unit is confirmed in storehouse, allow users to quickly step on
Record, it is not necessary to re-register.Or when user is logging in on-vehicle host using new mobile device, pass through picking up for car-mounted terminal
Sound device obtains the acoustic information of user and the device id of car-mounted terminal, and by the device id of car-mounted terminal beyond the clouds database
Confirm corresponding memory cell, and vocal print feature retrieval matching is carried out in the memory cell, equally facilitate user to be logged in.
In the present embodiment, in order to the vocal print feature of simple and fast resolving user, the vocal print feature includes:Sound
Learn feature, lexical characteristics, prosodic features and channel information.From the angle that can be modeled using mathematical method, vocal print is automatic
The feature that identification model can be used at present includes:(1) acoustic feature (cepstrum);(2) lexical characteristics (the related word of speaker
N-gram, phoneme n-gram);(3) prosodic features (fundamental tone and energy " posture " that are described using n-gram);(4) channel information
(which kind of passage used);N-Gram is a kind of language model commonly used in large vocabulary continuous speech recognition.By using mathematics side
Method is modeled, parsing voice messaging that can be quickly, so as to obtain the vocal print feature of user.
In the present embodiment, described " and carry out carrying out retrieval matching to the vocal print feature of extraction in the memory cell, when
The match is successful for the vocal print feature of extraction " specifically include:By by the vocal print feature in the vocal print feature of extraction and the memory cell
Compared one by one, when there is vocal print feature of the vocal print feature similar value more than setting value with extraction in the memory cell, then
The match is successful for the vocal print feature of extraction, if being not present, and it fails to match for the vocal print feature extracted.Influence due to various environment etc.
Vocal print feature in reason, the vocal print feature and memory cell of user 100% can not coincide, by by the vocal print feature of user with
Vocal print feature in memory cell is compared, and when similar value exceedes setting value, then confirms to coincide, by setting similar value to set
Definite value, then efficiently identify vocal print feature code, and reduce the possibility of mistake.Wherein setting value can be 80%.
In the present embodiment, in user's registration, the special sound content of push is pronounced by user, so as to obtain
The vocal print feature at family is taken, and is stored into memory cell.Occurred by the content of setting, such as " cloud degree automobile changes flat
It is all ", " cloud degree pursued vehicle remarkable ", the content provided of " cloud degree bus arrival is farther " pronounced and then or vocal print spy
Levy, stored into memory cell, in the memory unit, the foundation that sound-groove model can be refined one by one during identification, passes through
Pronounce by defined content, can so reach more preferable recognition effect.
Referring to Fig. 2, in another embodiment intelligent automobile voice quick registration device, including registering unit 210, extraction
Unit 220 and login unit 230;
The registering unit 210 is used in user's registration, obtains user speech information and device id, the device id bag
The device id of on-vehicle host and the device id of mobile terminal are included, and the corresponding storage of Database is single beyond the clouds according to device id
Member 240, and the vocal print feature of user is extracted from the user speech information, the vocal print feature of user is associatedly stored to setting
Memory cell 240 corresponding to standby ID;
The extraction unit 220 is used in User logs in, obtain user speech information and on-vehicle host device id or
The device id of person's mobile terminal, and parsing extraction vocal print feature is carried out to the user speech information;
The login unit 230 is used for according to the device id corresponding memory cell 240 of database retrieval device id beyond the clouds,
And carry out carrying out retrieval matching to the vocal print feature of extraction in the memory cell 240, when the vocal print feature of extraction, the match is successful,
Then logged in.
By registering unit 210, in user's registration, the acoustic information and device id provided during user's registration is obtained, and
The vocal print feature that user is extracted in parsing is carried out to acoustic information, while the device id provided according to user is built in database beyond the clouds
The vocal print feature of the user of extraction, is then stored in the memory cell 240 by view memory cell 240 corresponding with device id,
When user is logging in, the voice messaging that user provides is recognized by extraction unit 220, parsing obtains vocal print feature, and
The device id or the device id of mobile terminal for the on-vehicle host that user provides are obtained simultaneously, by logging in the basis of unit 230
The device id for the on-vehicle host that user provides or the device id of mobile terminal confirm memory cell 240 in database beyond the clouds,
When confirming memory cell 240 of the device id beyond the clouds in database, carried out in the memory cell 240 special to the vocal print of extraction
Progress matching retrieval is levied, when the match is successful for the vocal print feature, is then logined successfully, otherwise login failure.By device id in cloud
Corresponding memory cell 240 is confirmed in client database, and retrieval vocal print feature matching is carried out from memory cell 240, by setting
Standby ID is carried out the retrieval match time that classification storage greatly reduces volume vocal print feature to user profile, and special by the vocal print of user
Login can just be realized by levying, and logged in without inputting user name and login password, that is, add the car-mounted terminals such as car running computer
Login safety, also greatly facilitate user and logged in, while quickly recognizing user identity, realize quick registration.
In the present embodiment, the user speech information is sound pick-up by on-vehicle host or the mobile terminal of user
Microphone obtain, and the user speech information and device id of acquisition are sent to cloud server;Pass through cloud server pair
Voice messaging carries out parsing and extracts vocal print feature, according to the storage list that storage vocal print feature is determined in device id beyond the clouds database
Member 240, vocal print feature retrieval matching is carried out from memory cell 240, if the match is successful, User logs in is carried out, if matching is lost
Lose, then refuse the User logs in.When user carries out login on-vehicle host using new intelligent automobile, pass through mobile terminal
Microphone obtain voice messaging, and simultaneously obtain user mobile terminal device id, existed by the device id of mobile terminal
Corresponding memory cell 240 is confirmed in cloud database, and vocal print feature retrieval matching is carried out in memory cell 240, is used
Family being capable of quick registration, it is not necessary to re-register.Or when user is logging in on-vehicle host using new mobile device, pass through
The sound pick-up of car-mounted terminal obtains the acoustic information of user and the device id of car-mounted terminal, and is existed by the device id of car-mounted terminal
Corresponding memory cell 240 is confirmed in cloud database, and vocal print feature retrieval matching is carried out in the memory cell 240, equally
User is facilitated to be logged in.
In the present embodiment, in order to the vocal print feature of simple and fast resolving user, the vocal print feature includes:Sound
Learn feature, lexical characteristics, prosodic features and channel information.From the angle that can be modeled using mathematical method, vocal print is automatic
The feature that identification model can be used at present includes:(1) acoustic feature (cepstrum);(2) lexical characteristics (the related word of speaker
N-gram, phoneme n-gram);(3) prosodic features (fundamental tone and energy " posture " that are described using n-gram);(4) channel information
(which kind of passage used);N-Gram is a kind of language model commonly used in large vocabulary continuous speech recognition.By using mathematics side
Method is modeled, parsing voice messaging that can be quickly, so as to obtain the vocal print feature of user.
In the present embodiment, described " and carry out carrying out retrieval to the vocal print feature of extraction in the memory cell 240
Match somebody with somebody, when the vocal print feature of extraction, the match is successful " specifically include:By by the vocal print feature of extraction and the memory cell 240
Vocal print feature is compared one by one, exceedes setting value with the vocal print feature similar value of extraction when existing in the memory cell 240
Vocal print feature, then the match is successful for the vocal print feature extracted, if being not present, and it fails to match for the vocal print feature extracted.Due to various
The reasons such as the influence of environment, the vocal print feature of user and the vocal print feature in memory cell 240 100% can not coincide, and passing through will
The vocal print feature of user is compared with the vocal print feature in memory cell 240, when similar value exceedes setting value, then confirms kiss
Close, vocal print feature code is efficiently identified as setting value, then by setting similar value, and reduce the possibility of mistake.Wherein setting value
Can be 80%.
In the present embodiment, the registering unit 210 is additionally operable in user's registration, passes through specific language of the user to push
Sound content is pronounced, so as to obtain the vocal print feature of user, and is stored into memory cell 240.Entered by the content of setting
Row occurs, such as " cloud degree automobile change ordinary ", " cloud degree pursued vehicle is remarkable ", " cloud degree bus arrival is farther " provide it is interior
Hold pronounced and then or vocal print feature, stored into memory cell 240, in memory cell 240, sound-groove model energy
The enough foundation being refined one by one, during identification, by pronouncing by defined content, can so reach more preferable recognition effect.
Although the various embodiments described above are described, those skilled in the art once know basic wound
The property made concept, then can make other change and modification to these embodiments, so embodiments of the invention are the foregoing is only,
Not thereby the scope of patent protection of the present invention, the equivalent structure that every utilization description of the invention and accompanying drawing content are made are limited
Or equivalent flow conversion, or other related technical fields are directly or indirectly used in, similarly it is included in the patent of the present invention
Within protection domain.
Claims (10)
1. a kind of intelligent automobile voice quick registration method, it is characterised in that including:
In user's registration, user speech information and device id are obtained, the device id includes device id and the shifting of on-vehicle host
The device id of dynamic terminal;According to the device id corresponding memory cell of Database beyond the clouds, and from the user speech information
The middle vocal print feature for extracting user, the vocal print feature of user is associatedly stored to the memory cell corresponding to device id;
In User logs in, the device id or the device id of mobile terminal of user speech information and on-vehicle host are obtained, and it is right
The user speech information carries out parsing and extracts vocal print feature;
According to the device id corresponding memory cell of database retrieval device id, and being carried out in the memory cell to extracting beyond the clouds
Vocal print feature carry out retrieval matching, when the vocal print feature of extraction, the match is successful, then logged in.
2. intelligent automobile voice quick registration method according to claim 1, it is characterised in that the user speech information is
Obtained, and by the user speech information of acquisition and set by the microphone of the sound pick-up of on-vehicle host or the mobile terminal of user
Standby ID is sent to cloud server;
Parsing is carried out to voice messaging by cloud server and extracts vocal print feature, is determined according in device id beyond the clouds database
The memory cell of vocal print feature is stored, vocal print feature retrieval matching is carried out from memory cell, if the match is successful, user is carried out
Log in, if it fails to match, refuse the User logs in.
3. intelligent automobile voice quick registration method according to claim 1, it is characterised in that the vocal print feature code includes
Acoustic feature, lexical characteristics and prosodic features.
4. intelligent automobile voice quick registration method according to claim 1, it is characterised in that described " and in the storage list
Carry out carrying out retrieval matching to the vocal print feature of extraction in member, when the vocal print feature of extraction, the match is successful " specifically include:
By the way that the vocal print feature of extraction is compared one by one with the vocal print feature in the memory cell, deposited when in the memory cell
Exceeding the vocal print feature of setting value with the vocal print feature similar value of extraction, then the match is successful for the vocal print feature extracted, if not depositing
It fails to match for the vocal print feature then extracted.
5. intelligent automobile voice quick registration method according to claim 1, it is characterised in that in user's registration, pass through
User pronounces to the special sound content of push, so as to obtain the vocal print feature of user, and stores into memory cell.
6. a kind of intelligent automobile voice quick registration device, it is characterised in that single including registering unit, extraction unit and login
Member;
The registering unit is used in user's registration, obtains user speech information and device id, and the device id includes vehicle-mounted
The device id of main frame and the device id of mobile terminal, and according to the device id corresponding memory cell of Database beyond the clouds, and
The vocal print feature of user is extracted from the user speech information, the vocal print feature of user is associatedly stored right to device id institute
The memory cell answered;
The extraction unit is used in User logs in, obtains the device id or mobile whole of user speech information and on-vehicle host
The device id at end, and parsing extraction vocal print feature is carried out to the user speech information;
The login unit is used for according to the device id corresponding memory cell of database retrieval device id, and in the storage beyond the clouds
Carry out carrying out retrieval matching to the vocal print feature of extraction in unit, when the vocal print feature of extraction, the match is successful, then logged in.
7. intelligent automobile voice quick registration device according to claim 6, it is characterised in that the user speech information is
Obtained, and by the user speech information of acquisition and set by the microphone of the sound pick-up of on-vehicle host or the mobile terminal of user
Standby ID is sent to cloud server;
The extraction unit carries out parsing to voice messaging by cloud server and extracts vocal print feature;
The unit that logs in is according to the memory cell that storage vocal print feature is determined in device id beyond the clouds database, from memory cell
Middle progress vocal print feature retrieval matching, if the match is successful, carries out User logs in, if it fails to match, refuses the User logs in.
8. intelligent automobile voice quick registration device according to claim 6, it is characterised in that the vocal print feature code includes
Acoustic feature, lexical characteristics and prosodic features.
9. intelligent automobile voice quick registration device according to claim 6, it is characterised in that described " and in the storage list
Carry out carrying out retrieval matching to the vocal print feature of extraction in member, when the vocal print feature of extraction, the match is successful " specifically include:
By the way that the vocal print feature of extraction is compared one by one with the vocal print feature in the memory cell, deposited when in the memory cell
Exceeding the vocal print feature of setting value with the vocal print feature similar value of extraction, then the match is successful for the vocal print feature extracted, if not depositing
It fails to match for the vocal print feature then extracted.
10. intelligent automobile voice quick registration device according to claim 6, it is characterised in that the registering unit is also used
When in user's registration, the special sound content of push is pronounced by user, so that the vocal print feature of user is obtained, and
Storage is into memory cell.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710124156.4A CN107086983A (en) | 2017-03-03 | 2017-03-03 | A kind of intelligent automobile voice quick registration method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710124156.4A CN107086983A (en) | 2017-03-03 | 2017-03-03 | A kind of intelligent automobile voice quick registration method and device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107086983A true CN107086983A (en) | 2017-08-22 |
Family
ID=59614499
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710124156.4A Pending CN107086983A (en) | 2017-03-03 | 2017-03-03 | A kind of intelligent automobile voice quick registration method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107086983A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108305621A (en) * | 2017-08-25 | 2018-07-20 | 维沃移动通信有限公司 | A kind of phonetic order processing method and electronic equipment |
CN108614963A (en) * | 2018-04-28 | 2018-10-02 | 福建省汽车工业集团云度新能源汽车股份有限公司 | A kind of method and system of the Telnet onboard system based on user right |
CN109714306A (en) * | 2017-10-25 | 2019-05-03 | 现代自动车株式会社 | Customer certification system, user authen method and server |
CN110932862A (en) * | 2019-10-22 | 2020-03-27 | 安信通科技(澳门)有限公司 | Vehicle-mounted equipment login management method and device in vehicle |
CN111147484A (en) * | 2019-12-25 | 2020-05-12 | 秒针信息技术有限公司 | Account login method and device |
CN112910850A (en) * | 2021-01-16 | 2021-06-04 | 浙江吉利控股集团有限公司 | Safety verification method and system for controlling vehicle |
CN114724566A (en) * | 2022-04-18 | 2022-07-08 | 中国第一汽车股份有限公司 | Voice processing method, device, storage medium and electronic equipment |
CN115529163A (en) * | 2022-08-25 | 2022-12-27 | 江苏电力信息技术有限公司 | Login identification encryption method for digital business hall of electric power system |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007238058A (en) * | 2006-03-13 | 2007-09-20 | Denso Corp | Smart entry system for vehicle |
CN105279416A (en) * | 2015-10-27 | 2016-01-27 | 上海川织金融信息服务有限公司 | Identity recognition method and system based on multi-biometric feature in combination with device fingerprint |
CN105447364A (en) * | 2015-04-30 | 2016-03-30 | 北京天诚盛业科技有限公司 | Remote biological recognition registering method, device and system |
CN105721460A (en) * | 2016-01-29 | 2016-06-29 | 大连楼兰科技股份有限公司 | Voiceprint password vehicle control system and method |
-
2017
- 2017-03-03 CN CN201710124156.4A patent/CN107086983A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2007238058A (en) * | 2006-03-13 | 2007-09-20 | Denso Corp | Smart entry system for vehicle |
CN105447364A (en) * | 2015-04-30 | 2016-03-30 | 北京天诚盛业科技有限公司 | Remote biological recognition registering method, device and system |
CN105279416A (en) * | 2015-10-27 | 2016-01-27 | 上海川织金融信息服务有限公司 | Identity recognition method and system based on multi-biometric feature in combination with device fingerprint |
CN105721460A (en) * | 2016-01-29 | 2016-06-29 | 大连楼兰科技股份有限公司 | Voiceprint password vehicle control system and method |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108305621A (en) * | 2017-08-25 | 2018-07-20 | 维沃移动通信有限公司 | A kind of phonetic order processing method and electronic equipment |
CN108305621B (en) * | 2017-08-25 | 2020-05-05 | 维沃移动通信有限公司 | Voice instruction processing method and electronic equipment |
CN109714306A (en) * | 2017-10-25 | 2019-05-03 | 现代自动车株式会社 | Customer certification system, user authen method and server |
CN108614963A (en) * | 2018-04-28 | 2018-10-02 | 福建省汽车工业集团云度新能源汽车股份有限公司 | A kind of method and system of the Telnet onboard system based on user right |
CN110932862A (en) * | 2019-10-22 | 2020-03-27 | 安信通科技(澳门)有限公司 | Vehicle-mounted equipment login management method and device in vehicle |
CN111147484A (en) * | 2019-12-25 | 2020-05-12 | 秒针信息技术有限公司 | Account login method and device |
CN111147484B (en) * | 2019-12-25 | 2022-06-14 | 秒针信息技术有限公司 | Account login method and device |
CN112910850A (en) * | 2021-01-16 | 2021-06-04 | 浙江吉利控股集团有限公司 | Safety verification method and system for controlling vehicle |
CN114724566A (en) * | 2022-04-18 | 2022-07-08 | 中国第一汽车股份有限公司 | Voice processing method, device, storage medium and electronic equipment |
CN115529163A (en) * | 2022-08-25 | 2022-12-27 | 江苏电力信息技术有限公司 | Login identification encryption method for digital business hall of electric power system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107086983A (en) | A kind of intelligent automobile voice quick registration method and device | |
CN110265040B (en) | Voiceprint model training method and device, storage medium and electronic equipment | |
KR101963993B1 (en) | Identification system and method with self-learning function based on dynamic password voice | |
CN103700371B (en) | A kind of caller identity identifying system and its recognition methods based on Application on Voiceprint Recognition | |
WO2016150001A1 (en) | Speech recognition method, device and computer storage medium | |
CN108766441A (en) | A kind of sound control method and device based on offline Application on Voiceprint Recognition and speech recognition | |
ATE158889T1 (en) | SYSTEM FOR PROCESSING CONTINUOUS SPEECH | |
CN104143326A (en) | Voice command recognition method and device | |
CN110348002A (en) | The system and method for realizing voice request | |
CN113129867B (en) | Training method of voice recognition model, voice recognition method, device and equipment | |
CN111916088B (en) | Voice corpus generation method and device and computer readable storage medium | |
CN107068146A (en) | A kind of intelligent automobile voice login method and system | |
CN109272991A (en) | Method, apparatus, equipment and the computer readable storage medium of interactive voice | |
CN110718217B (en) | Control method, terminal and computer readable storage medium | |
CN110539721A (en) | vehicle control method and device | |
US20150310853A1 (en) | Systems and methods for speech artifact compensation in speech recognition systems | |
CN107818783A (en) | A kind of mutual method and device of man-machine multi-modal on-vehicle safety sexual intercourse based on vocal print technology | |
CN109074809B (en) | Information processing apparatus, information processing method, and computer-readable storage medium | |
KR20230156145A (en) | Hybrid multilingual text-dependent and text-independent speaker verification | |
CN112927688B (en) | Voice interaction method and system for vehicle | |
CN109979445A (en) | It is a kind of for understanding the system and method for standard language and dialect | |
CN106850673A (en) | User vehicle accessing method and device based on Application on Voiceprint Recognition | |
CN107010009A (en) | A kind of safe accessing method of intelligent automobile voice and device | |
CN110503943A (en) | A kind of voice interactive method and voice interactive system | |
CN112309396A (en) | AI virtual robot state dynamic setting system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170822 |