CN109243465A - Voiceprint authentication method, device, computer equipment and storage medium - Google Patents
Voiceprint authentication method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109243465A CN109243465A CN201811487395.7A CN201811487395A CN109243465A CN 109243465 A CN109243465 A CN 109243465A CN 201811487395 A CN201811487395 A CN 201811487395A CN 109243465 A CN109243465 A CN 109243465A
- Authority
- CN
- China
- Prior art keywords
- vocal print
- certified
- sound
- groove model
- print feature
- 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 41
- 230000001755 vocal effect Effects 0.000 claims abstract description 248
- 239000000284 extract Substances 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 30
- 238000004590 computer program Methods 0.000 claims description 16
- 238000000605 extraction Methods 0.000 claims description 13
- 230000015654 memory Effects 0.000 claims description 11
- 238000004422 calculation algorithm Methods 0.000 claims description 10
- 238000003064 k means clustering Methods 0.000 claims description 10
- 238000005516 engineering process Methods 0.000 abstract description 8
- 230000004044 response Effects 0.000 abstract description 6
- 238000010586 diagram Methods 0.000 description 13
- 230000006870 function Effects 0.000 description 6
- 239000000203 mixture Substances 0.000 description 6
- 230000008569 process Effects 0.000 description 5
- 230000033764 rhythmic process Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 210000000056 organ Anatomy 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000013139 quantization Methods 0.000 description 3
- 239000011159 matrix material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000035479 physiological effects, processes and functions Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 208000037656 Respiratory Sounds Diseases 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 210000003484 anatomy Anatomy 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000012905 input function Methods 0.000 description 1
- 210000000867 larynx Anatomy 0.000 description 1
- 210000004072 lung Anatomy 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 210000003928 nasal cavity Anatomy 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 210000002105 tongue Anatomy 0.000 description 1
- 210000000515 tooth Anatomy 0.000 description 1
Classifications
-
- 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
- G10L17/04—Training, enrolment or model building
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Collating Specific Patterns (AREA)
Abstract
The embodiment of the invention discloses a kind of voiceprint authentication method, device, computer equipment and storage mediums, wherein the described method includes: pre-establishing sound-groove model library；The voice print database to be certified of user's input is obtained, and extracts corresponding vocal print feature to be certified from the voice print database to be certified；Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified；Calculate the vocal print similarity between the vocal print feature to be certified and target sound-groove model；It obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold；If the vocal print similarity is greater than the preset threshold, determine that the corresponding user identity of the voice print database to be certified matches with target user's identity, and determine that the user identity authentication passes through.The present invention provides a kind of sound groove recognition technology in e, can be improved authentication efficiency, shortens the response time of certification.
Description
Technical field
The present invention relates to field of computer technology more particularly to a kind of voiceprint authentication method, device, computer equipment and
Storage medium.
Background technique
Application on Voiceprint Recognition (Voiceprint Recognize) is one and is spoken human physiology and row according to reflection in speech waveform
The speech parameter being characterized, the technology of automatic identification speaker's identity.With the development of sound groove recognition technology in e, more and more set
Standby that authentication is carried out using Application on Voiceprint Recognition, traditional technology that authentication is carried out using Application on Voiceprint Recognition is a small amount of in processing
Voice print database when can guarantee the efficiency of certification, but for a certain number of voice print databases, low, response that there are authentication efficiencies
The problem of time length.
Summary of the invention
It is situated between in view of this, the embodiment of the present invention provides a kind of voiceprint authentication method, device, computer equipment and storage
Matter can be improved authentication efficiency, shorten the response time of certification.
On the one hand, the embodiment of the invention provides a kind of voiceprint authentication methods, this method comprises:
Pre-establish sound-groove model library；
The voice print database to be certified of user's input is obtained, and is extracted from the voice print database to be certified corresponding to be certified
Vocal print feature；
Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified；
Calculate the vocal print similarity between the vocal print feature to be certified and target sound-groove model；
It obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold；
If the vocal print similarity is greater than the preset threshold, the corresponding user identity of the voice print database to be certified is determined
Match with target user's identity, and determines that the user identity authentication passes through.
On the other hand, the embodiment of the invention provides a kind of voiceprint authentication apparatus, described device includes:
Unit is established, for pre-establishing sound-groove model library；
First extraction unit, for obtaining the voice print database to be certified of user's input, and from the voice print database to be certified
It is middle to extract corresponding vocal print feature to be certified；
Determination unit, for determining target vocal print mould from the sound-groove model library according to the vocal print feature to be certified
Type；
Computing unit, for calculating the vocal print similarity between the vocal print feature to be certified and target sound-groove model；
Judging unit, for obtaining the vocal print similarity and judging whether the vocal print similarity is greater than preset threshold；
Judging unit determines the voice print database to be certified if being greater than the preset threshold for the vocal print similarity
Corresponding user identity matches with target user's identity, and determines that the user identity authentication passes through.
Another aspect the embodiment of the invention also provides a kind of computer equipment, including memory, processor and is stored in
On the memory and the computer program that can run on the processor, when the processor executes the computer program
Realize voiceprint authentication method as described above.
It is described computer-readable to deposit in another aspect, the embodiment of the invention also provides a kind of computer readable storage medium
Storage media is stored with one or more than one computer program, and the one or more computer program can be by one
Or more than one processor executes, to realize voiceprint authentication method as described above.
The embodiment of the present invention provides a kind of voiceprint authentication method, device, computer equipment and storage medium, wherein method
It include: to pre-establish sound-groove model library；The voice print database to be certified of user's input is obtained, and from the voice print database to be certified
Extract corresponding vocal print feature to be certified；Target vocal print is determined from the sound-groove model library according to the vocal print feature to be certified
Model；Calculate the vocal print similarity between the vocal print feature to be certified and target sound-groove model；Obtain the vocal print similarity
And judge whether the vocal print similarity is greater than preset threshold；If the vocal print similarity is greater than the preset threshold, institute is determined
It states the corresponding user identity of voice print database to be certified to match with target user's identity, and determines that the user identity authentication is logical
It crosses.The present invention provides a kind of sound groove recognition technology in e, can be improved authentication efficiency, shortens the response time of certification.
Detailed description of the invention
Technical solution in order to illustrate the embodiments of the present invention more clearly, below will be to needed in embodiment description
Attached drawing is briefly described, it should be apparent that, drawings in the following description are some embodiments of the invention, general for this field
For logical technical staff, without creative efforts, it is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of application scenarios schematic diagram of voiceprint authentication method provided in an embodiment of the present invention；
Fig. 2 is a kind of schematic flow diagram of voiceprint authentication method provided in an embodiment of the present invention；
Fig. 3 is a kind of another schematic flow diagram of voiceprint authentication method provided in an embodiment of the present invention；
Fig. 4 is a kind of another schematic flow diagram of voiceprint authentication method provided in an embodiment of the present invention；
Fig. 5 is a kind of schematic block diagram of voiceprint authentication apparatus provided in an embodiment of the present invention；
Fig. 6 is a kind of another schematic block diagram of voiceprint authentication apparatus provided in an embodiment of the present invention；
Fig. 7 is a kind of another schematic block diagram of voiceprint authentication apparatus provided in an embodiment of the present invention；
Fig. 8 is a kind of structure composition schematic diagram of computer equipment provided in an embodiment of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are some of the embodiments of the present invention, instead of all the embodiments.Based on this hair
Embodiment in bright, every other implementation obtained by those of ordinary skill in the art without making creative efforts
Example, shall fall within the protection scope of the present invention.
It should be appreciated that ought use in this specification and in the appended claims, term " includes " and "comprising" instruction
Described feature, entirety, step, operation, the presence of element and/or component, but one or more of the other feature, whole is not precluded
Body, step, operation, the presence or addition of element, component and/or its set.
It is also understood that mesh of the term used in this description of the invention merely for the sake of description specific embodiment
And be not intended to limit the present invention.As description of the invention and it is used in the attached claims, unless on
Other situations are hereafter clearly indicated, otherwise " one " of singular, "one" and "the" are intended to include plural form.
It will be further appreciated that the term "and/or" used in description of the invention and the appended claims is
Refer to any combination and all possible combinations of one or more of associated item listed, and including these combinations.
Fig. 1 and Fig. 2 are please referred to, Fig. 1 is a kind of application scenarios signal of voiceprint authentication method provided in an embodiment of the present invention
Figure, Fig. 2 are a kind of flow diagram of voiceprint authentication method provided in an embodiment of the present invention.The voiceprint authentication method is applied to clothes
It is engaged in device or terminal, wherein terminal can be smart phone, tablet computer, laptop, desktop computer, personal digital assistant
Electronic equipment with wearable device etc. with communication function.As an application, as shown in Figure 1, the voiceprint authentication method application
In server 10, which can be a server in Distributed Services platform, which executes vocal print
Identification/certification instruction, and by implementing result feedback in terminal 20.
It should be noted that only illustrate a terminal 20 in Fig. 1, in the actual operation process, server 10 can be with
Voiceprint result is fed back in more terminals 20.
Referring to Fig. 2, Fig. 2 is a kind of schematic flow diagram of voiceprint authentication method provided in an embodiment of the present invention.Such as Fig. 2 institute
Show, this approach includes the following steps S101~S106.
S101 pre-establishes sound-groove model library.
In embodiments of the present invention, different particular by collecting several sections of voice data in advance as training speech samples
Voice data correspond to different user identity, utilize GMM model (Gaussian Mixture Model, gauss hybrid models)
Training speech samples are trained, obtain the sound-groove model for different user identity, and by obtained different vocal print moulds
Type forms default sound-groove model library；More specifically, it presets and preserves trained speech samples in sound-groove model library and correspond to vocal print at this
Model, the corresponding vocal print feature of training speech samples and different user identity corresponding from training speech samples.Wherein, it uses
Family identity can be identified by identity ID, the user that different identity ID can be different with unique identification.
Further, as shown in figure 3, the step S101 includes step S202~S208.
S202, at least two training speech samples of the acquisition for model training.
In embodiments of the present invention, user can be passed through by the microphone typing voice data in terminal 20, server 10
With terminal 20 and obtain voice data that user inputted as training speech samples, specifically, at least need two it is different
User's typing two different voice data are as training speech samples.
S204 pre-processes each each trained speech samples.
In embodiments of the present invention, it before extracting each trained speech samples, needs to carry out each trained speech samples pre-
Processing, because of mankind's phonatory organ itself and the aliasing as brought by the equipment of acquisition voice data, higher hamonic wave distortion, height
Frequency etc. factor, the influence to quality of speech signal, by pre-processing the voice signal that can be obtained more evenly, smoothly, and
The low-frequency disturbance in voice data can be removed, good characteristic parameter is provided for the extraction of vocal print feature, improves speech processes
Quality.
It is divided into sample quantization specifically, pre-process to each trained speech samples, goes drift, preemphasis and adding window
Four steps.Wherein, sample quantization is handled, and refers to being filtered with voice signal of the sharp filter to training speech samples
Wave makes its nyquist frequency FN 4KHZ；Drift is gone, refers to the average value for calculating the amplitude sequence of quantization, and will be each
Amplitude subtracts average value, obtains the amplitude sequence that average value is 0 after drift；Preemphasis refers to setting digital filter
Transmission function in pre emphasis factor, and the high, medium and low comparable vibration of frequency amplitude of voice signal is obtained by digital filter
Width sequence；Adding window is referred to carrying out each speech signal frame using Hamming window function plus hamming code window is handled.
S206 extracts the vocal print feature of pretreated each trained speech samples.
In embodiments of the present invention, so-called vocal print (Voiceprint) is the carrying speech letter that electricity consumption acoustic instrument is shown
The sound wave spectrum of breath.The generation of human language is a complicated physiology physics mistake between Body Languages maincenter and vocal organs
Journey, the phonatory organ (such as: tongue, tooth, larynx, lung, nasal cavity) that people uses in speech in terms of size and form everyone
It is widely different, so the voiceprint map of any two people is all variant.Different speak can be recognized and confirmed by vocal print
People.The vocal print feature includes one of acoustic feature, lexical characteristics, prosodic features and language feature or a variety of；This reality
Example is applied and realizes based on the vocal print feature of different trained speech samples the training of different sound-groove models, and based on different vocal prints
The determination of trained objectives sound-groove model also may be implemented in feature, it is therefore desirable to extract from training speech samples in advance
Out the vocal print feature of different trained speech samples and correspondence saved；Specifically, being extracted from different trained speech samples
It corresponding vocal print feature, in particular to is extracted from sound and selects there is the vocal print of speaker that separability is strong, stability is high
Etc. characteristics acoustics, morphology, the rhythm or language feature, extraction in the prior art can be passed through for the extraction of vocal print feature
Mode realizes that this is no longer described in detail in the embodiment of the present invention.
It should be noted that the vocal print feature extracted to training speech samples may be one, it is also possible to multiple.
Each vocal print feature of extraction is carried out gauss hybrid models training by S208, obtains different sound-groove models, and by
The different sound-groove model forms the sound-groove model library.
In embodiments of the present invention, GMM model is carried out to the vocal print feature for belonging to a trained speech samples
(Gaussian Mixture Model, gauss hybrid models) training is to establish corresponding sound-groove model.More specifically, can first really
Surely belong to the corresponding speech signal frame of all vocal print features of the same trained speech samples, it later, can be to belonging to the same instruction
All speech signal frames for practicing speech samples are trained to obtain corresponding GMM model, and by different training speech samples
The different sound-groove models that are trained of vocal print feature form the sound-groove model library.
S102, obtains the voice print database to be certified of user's input, and extracts from the voice print database to be certified corresponding
Vocal print feature to be certified.
In embodiments of the present invention, the voice print database to be certified of user's input refers to that unverified user passes through terminal 20
Microphone input for certification voice data.After detecting the voice print database to be certified of active user's input, it can mention
Take the vocal print feature to be certified of current voice print database, wherein the vocal print feature to be certified include acoustic feature, lexical characteristics,
One of prosodic features and language feature are a variety of.Vocal print feature, the i.e. anatomical structure with the pronunciation mechanism of the mankind have
The feature of pass, including such as nasal sound, band deep breathing sound, hoarse sound, laugh, thus analysis is available such as frequency spectrum, cepstrum, resonance
The features such as peak, fundamental tone, reflection coefficient；Lexical characteristics, i.e., the language influenced by socioeconomic status, education level, birthplace etc.
The features such as justice, rhetoric, pronunciation, speech habit；Prosodic features, that is, the features such as the rhythm spoken, rhythm, speed, intonation, volume.
The features such as language feature, i.e. languages, dialect, accent.
S103 determines target sound-groove model according to the vocal print feature to be certified from the sound-groove model library.
In embodiments of the present invention, using in extracted vocal print feature to be certified and the default sound-groove model library established
The corresponding vocal print feature parameter of each sound-groove model compares, and the target sound-groove model is determined according to comparing result, specific
During realization, it is contemplated that the sound of each individual subject may change within a certain period of time, in order to improve contrast effect,
It is a predetermined threshold that comparison degree, which can be preset, when in extracted vocal print feature to be certified and default sound-groove model library
When the comparison degree of the corresponding vocal print feature parameter of each sound-groove model is more than institute's scheduled threshold value, it is determined that the vocal print compared is special
The corresponding sound-groove model of sign parameter is the vocal print feature extracted in target sound-groove model, such as current voice print database and default vocal print
Vocal print feature parameter A matching degree in model library reaches 90% or more, that is, determines the corresponding vocal print mould of vocal print feature parameter A
Type is target sound-groove model.
In one embodiment, as shown in figure 4, the step S103 includes step S302~S306.
S302 determines the quantity of extracted vocal print feature to be certified.
In embodiments of the present invention, extracted vocal print feature to be certified may be one in current voice print database,
It may be multiple, it is thus necessary to determine that the quantity of extracted vocal print feature to be certified.
S304 determines the priority of each vocal print feature to be certified if extracted vocal print feature to be certified is multiple.
In embodiments of the present invention, when extracted vocal print feature to be certified be it is multiple when, it is thus necessary to determine that comparison wait recognize
The priority of vocal print feature is demonstrate,proved, for example, the vocal print feature to be certified currently extracted includes acoustic feature, lexical characteristics, rhythm spy
It levies, vocal print feature priority corresponding to acoustic feature is higher than lexical characteristics, vocal print feature priority corresponding to lexical characteristics
Higher than prosodic features, select acoustic feature as the vocal print feature to be certified for comparison at this time.Vocal print feature knowledge can be improved
Other accuracy.
S306 determines target vocal print mould according to the vocal print feature to be certified of highest priority from the sound-groove model library
Type.
In embodiments of the present invention, the vocal print feature to be certified of highest priority and the default sound-groove model established are selected
The corresponding vocal print feature parameter of each sound-groove model compares in library, when extracted vocal print feature to be certified and default vocal print mould
When the comparison degree of the corresponding vocal print feature parameter of each sound-groove model is more than institute's scheduled threshold value in type library, it is determined that compared
The corresponding sound-groove model of vocal print feature parameter is target sound-groove model.The accuracy of vocal print feature identification can be improved.
S104 calculates the vocal print similarity between the vocal print feature to be certified and target sound-groove model.
In embodiments of the present invention, the vocal print similarity between the vocal print feature to be certified and target sound-groove model is calculated
Calculation method may include:
If the target sound-groove model has n vocal print feature, respectively to the vocal print feature to be certified and the mesh
N vocal print feature for marking sound-groove model carries out vectorization, calculates the vocal print feature to be certified and institute by included angle cosine function
The similarity between n vocal print feature of target sound-groove model is stated, and obtains similarity matrix K；Similarity matrix K is asked
With to calculate the vocal print similarity between the vocal print feature to be certified and the target sound-groove model:
If K=[sim (x, yi)]n, i=1 ..., n, wherein sim (x, yi) indicate vocal print feature to be certified and the target
Similarity between n vocal print feature of sound-groove model, sum formula are as follows:
S=1 ..., n.
In one embodiment, if the vocal print feature to be certified be it is multiple, using K-means clustering algorithm calculate described in
Vocal print similarity between the vocal print feature to be certified of highest priority and each vocal print feature of the target sound-groove model.
In one embodiment, if the vocal print feature to be certified is one, using described in the calculating of K-means clustering algorithm
Vocal print similarity between vocal print feature to be certified and each vocal print feature of the target sound-groove model.
S105 obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold.
In embodiments of the present invention, the preset threshold for vocal print similarity comparison can be preset, when calculated
The vocal print similarity between target sound-groove model in vocal print feature to be certified and default sound-groove model library is more than the default threshold
When value, then by the certification of the vocal print feature to be certified, and user identity corresponding to the vocal print feature to be certified is determined
Target user's identity corresponding with target sound-groove model matches and passes through authentication.
S106 determines the corresponding use of the voice print database to be certified if the vocal print similarity is greater than the preset threshold
Family identity matches with target user's identity, and determines that the user identity authentication passes through.
In embodiments of the present invention, only just allow user's further operating after user identity is by certification, for example,
Management is operated by the fund of the user of certification, typing user information or the personal information for modifying user etc. at the terminal.
As seen from the above, the embodiment of the present invention is by pre-establishing sound-groove model library；Obtain the sound to be certified of user's input
Line data, and corresponding vocal print feature to be certified is extracted from the voice print database to be certified；It is special according to the vocal print to be certified
Sign determines target sound-groove model from the sound-groove model library；It calculates between the vocal print feature to be certified and target sound-groove model
Vocal print similarity；It obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold；If the sound
Line similarity is greater than the preset threshold, determines the corresponding user identity of the voice print database to be certified and target user's identity phase
Matching, and determine that the user identity authentication passes through.The present invention provides a kind of sound groove recognition technology in e, can be improved authentication effect
Rate shortens the response time of certification.
Referring to Fig. 5, a kind of corresponding above-mentioned voiceprint authentication method, the embodiment of the present invention also proposes a kind of voiceprint dress
It sets, which includes: to establish unit 101, the first extraction unit 102, determination unit 103, computing unit 104, judging unit
105, judging unit 106.
Wherein, described to establish unit 101, for pre-establishing sound-groove model library.
First extraction unit 102, for obtaining the voice print database to be certified of user's input, and from the vocal print number to be certified
Corresponding vocal print feature to be certified is extracted according to middle.
Determination unit 103, for determining target vocal print from the sound-groove model library according to the vocal print feature to be certified
Model.
Computing unit 104, for calculating the vocal print similarity between the vocal print feature to be certified and target sound-groove model.
Judging unit 105, for obtaining the vocal print similarity and judging whether the vocal print similarity is greater than default threshold
Value.
Judging unit 106 determines the vocal print number to be certified if being greater than the preset threshold for the vocal print similarity
Match according to corresponding user identity and target user's identity, and determines that the user identity authentication passes through.
In one embodiment, the computing unit 104, if be specifically used for the vocal print feature to be certified be it is multiple, make
The vocal print feature to be certified of the highest priority and each vocal print of the target sound-groove model are calculated with K-means clustering algorithm
Vocal print similarity between feature.
In one embodiment, the computing unit 104, if being one also particularly useful for the vocal print feature to be certified,
It is calculated between the vocal print feature to be certified and each vocal print feature of the target sound-groove model using K-means clustering algorithm
Vocal print similarity.
As seen from the above, the embodiment of the present invention is by pre-establishing sound-groove model library；Obtain the sound to be certified of user's input
Line data, and corresponding vocal print feature to be certified is extracted from the voice print database to be certified；It is special according to the vocal print to be certified
Sign determines target sound-groove model from the sound-groove model library；It calculates between the vocal print feature to be certified and target sound-groove model
Vocal print similarity；It obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold；If the sound
Line similarity is greater than the preset threshold, determines the corresponding user identity of the voice print database to be certified and target user's identity phase
Matching, and determine that the user identity authentication passes through.The present invention provides a kind of sound groove recognition technology in e, can be improved authentication effect
Rate shortens the response time of certification.
Referring to Fig. 6, described establish unit 101, comprising:
Acquisition unit 101a, for acquiring at least two training voices for being used for model training.
Pretreatment unit 101b, for being pre-processed to each each trained speech samples.
Second extraction unit 101c, for extracting the vocal print feature of pretreated each trained speech samples.
Subelement 101d is established, each vocal print feature for that will extract carries out gauss hybrid models training, obtains difference
Sound-groove model, and the sound-groove model library is formed by the different sound-groove model.
Referring to Fig. 7, the determination unit 103, comprising:
First determines subelement 103a, for determining the quantity of extracted vocal print feature to be certified.
Second determines subelement 103b, if be multiple for extracted vocal print feature to be certified, determines each wait recognize
Demonstrate,prove the priority of vocal print feature.
Third determines subelement 103c, for according to the vocal print feature to be certified of highest priority from the sound-groove model library
Middle determining target sound-groove model.
Above-mentioned voiceprint authentication apparatus and above-mentioned voiceprint authentication method one-to-one correspondence, specific principle and process and above-mentioned reality
It is identical to apply the method, repeats no more.
Above-mentioned voiceprint authentication apparatus can be implemented as a kind of form of computer program, and computer program can be in such as Fig. 8
Shown in run in computer equipment.
Fig. 8 is a kind of structure composition schematic diagram of computer equipment of the present invention.The equipment can be terminal, be also possible to take
Business device, wherein terminal can be smart phone, tablet computer, laptop, desktop computer, personal digital assistant and wearing
Formula device etc. has the electronic device of communication function and speech voice input function.Server can be independent server, can also be with
It is the server cluster of multiple server compositions, server can obtain the vocal print to be certified of user's input by communicating with terminal
Data.Referring to Fig. 8, which includes the processor 502 connected by system bus 501, non-volatile memories Jie
Matter 503, built-in storage 504 and network interface 505.Wherein, the non-volatile memory medium 503 of the computer equipment 500 can be deposited
Storage operating system 5031 and computer program 5032, the computer program 5032 are performed, and processor 502 may make to execute one
Kind voiceprint authentication method.The processor 502 of the computer equipment 500 supports entire calculate for providing calculating and control ability
The operation of machine equipment 500.The built-in storage 504 is that the operation of the computer program 5032 in non-volatile memory medium 503 mentions
For environment, when which is executed by processor, processor 502 may make to execute a kind of voiceprint authentication method.Computer
The network interface 505 of equipment 500 is for carrying out network communication.It will be understood by those skilled in the art that structure shown in Fig. 8,
The only block diagram of part-structure relevant to application scheme, does not constitute the calculating being applied thereon to application scheme
The restriction of machine equipment, specific computer equipment may include more certain than more or fewer components as shown in the figure, or combination
Component, or with different component layouts.
Wherein, following operation is realized when the processor 502 executes the computer program:
Pre-establish sound-groove model library；
The voice print database to be certified of user's input is obtained, and is extracted from the voice print database to be certified corresponding to be certified
Vocal print feature；
Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified；
Calculate the vocal print similarity between the vocal print feature to be certified and target sound-groove model；
It obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold；
If the vocal print similarity is greater than the preset threshold, the corresponding user identity of the voice print database to be certified is determined
Match with target user's identity, and determines that the user identity authentication passes through.
It is in one embodiment, described to pre-establish sound-groove model library, comprising:
At least two training speech samples of the acquisition for model training；
Each each trained speech samples are pre-processed；
Extract the vocal print feature of pretreated each trained speech samples；
Each vocal print feature of extraction is subjected to gauss hybrid models training, obtains different sound-groove models, and by described
Different sound-groove models forms the sound-groove model library.
In one embodiment, described to determine target sound from the sound-groove model library according to the vocal print feature to be certified
Line model, comprising:
Determine the quantity of extracted vocal print feature to be certified；
If extracted vocal print feature to be certified is multiple, the priority of each vocal print feature to be certified is determined；
Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified of highest priority.
In one embodiment, the calculating vocal print feature to be certified is similar to the vocal print between target sound-groove model
Degree, comprising:
If the vocal print feature to be certified be it is multiple, using K-means clustering algorithm calculate the highest priority to
Authenticate the vocal print similarity between vocal print feature and each vocal print feature of the target sound-groove model.
In one embodiment, the calculating vocal print feature to be certified is similar to the vocal print between target sound-groove model
Degree, comprising:
If the vocal print feature to be certified is one, the vocal print feature to be certified is calculated using K-means clustering algorithm
Vocal print similarity between each vocal print feature of the target sound-groove model.
It will be understood by those skilled in the art that the embodiment of computer equipment shown in Fig. 8 is not constituted to computer
The restriction of equipment specific composition, in other embodiments, computer equipment may include components more more or fewer than diagram, or
Person combines certain components or different component layouts.For example, in some embodiments, computer equipment only includes memory
And processor, in such embodiments, the structure and function of memory and processor are consistent with embodiment illustrated in fig. 8, herein
It repeats no more.
The present invention provides a kind of computer readable storage medium, computer-readable recording medium storage has one or one
A above computer program, the one or more computer program can be held by one or more than one processor
Row, to perform the steps of
Pre-establish sound-groove model library；
The voice print database to be certified of user's input is obtained, and is extracted from the voice print database to be certified corresponding to be certified
Vocal print feature；
Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified；
Calculate the vocal print similarity between the vocal print feature to be certified and target sound-groove model；
It obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold；
If the vocal print similarity is greater than the preset threshold, the corresponding user identity of the voice print database to be certified is determined
Match with target user's identity, and determines that the user identity authentication passes through.
It is in one embodiment, described to pre-establish sound-groove model library, comprising:
At least two training speech samples of the acquisition for model training；
Each each trained speech samples are pre-processed；
Extract the vocal print feature of pretreated each trained speech samples；
Each vocal print feature of extraction is subjected to gauss hybrid models training, obtains different sound-groove models, and by described
Different sound-groove models forms the sound-groove model library.
In one embodiment, described to determine target sound from the sound-groove model library according to the vocal print feature to be certified
Line model, comprising:
Determine the quantity of extracted vocal print feature to be certified；
If extracted vocal print feature to be certified is multiple, the priority of each vocal print feature to be certified is determined；
Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified of highest priority.
In one embodiment, the calculating vocal print feature to be certified is similar to the vocal print between target sound-groove model
Degree, comprising:
If the vocal print feature to be certified be it is multiple, using K-means clustering algorithm calculate the highest priority to
Authenticate the vocal print similarity between vocal print feature and each vocal print feature of the target sound-groove model.
In one embodiment, the calculating vocal print feature to be certified is similar to the vocal print between target sound-groove model
Degree, comprising:
If the vocal print feature to be certified is one, the vocal print feature to be certified is calculated using K-means clustering algorithm
Vocal print similarity between each vocal print feature of the target sound-groove model.
Present invention storage medium above-mentioned include: magnetic disk, CD, read-only memory (Read-Only Memory,
The various media that can store program code such as ROM).
Unit in all embodiments of the invention can pass through universal integrated circuit, such as CPU (Central
Processing Unit, central processing unit), or pass through ASIC (Application Specific Integrated
Circuit, specific integrated circuit) it realizes.
Step in voiceprint authentication method of the embodiment of the present invention can according to actual needs the adjustment of carry out sequence, merge and delete
Subtract.
Unit in voiceprint authentication apparatus of the embodiment of the present invention can be combined, divided and deleted according to actual needs.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, any
Those familiar with the art in the technical scope disclosed by the present invention, can readily occur in various equivalent modifications or replace
It changes, these modifications or substitutions should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with right
It is required that protection scope subject to.
Claims (10)
1. a kind of voiceprint authentication method, which is characterized in that the described method includes:
Pre-establish sound-groove model library；
The voice print database to be certified of user's input is obtained, and extracts corresponding vocal print to be certified from the voice print database to be certified
Feature；
Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified；
Calculate the vocal print similarity between the vocal print feature to be certified and target sound-groove model；
It obtains the vocal print similarity and judges whether the vocal print similarity is greater than preset threshold；
If the vocal print similarity is greater than the preset threshold, the corresponding user identity of the voice print database to be certified and mesh are determined
Mark user identity matches, and determines that the user identity authentication passes through.
2. the method as described in claim 1, which is characterized in that described to pre-establish sound-groove model library, comprising:
At least two training speech samples of the acquisition for model training；
Each each trained speech samples are pre-processed；
Extract the vocal print feature of pretreated each trained speech samples；
Each vocal print feature of extraction is subjected to gauss hybrid models training, obtains different sound-groove models, and by the difference
Sound-groove model form the sound-groove model library.
3. the method as described in claim 1, which is characterized in that it is described according to the vocal print feature to be certified from the vocal print mould
Target sound-groove model is determined in type library, comprising:
Determine the quantity of extracted vocal print feature to be certified；
If extracted vocal print feature to be certified is multiple, the priority of each vocal print feature to be certified is determined；
Target sound-groove model is determined from the sound-groove model library according to the vocal print feature to be certified of highest priority.
4. method as claimed in claim 3, which is characterized in that described to calculate the vocal print feature to be certified and target vocal print mould
Vocal print similarity between type, comprising:
If the vocal print feature to be certified be it is multiple, calculate the to be certified of the highest priority using K-means clustering algorithm
Vocal print similarity between vocal print feature and each vocal print feature of the target sound-groove model.
5. the method as described in claim 1, which is characterized in that described to calculate the vocal print feature to be certified and target vocal print mould
Vocal print similarity between type, comprising:
If the vocal print feature to be certified is one, the vocal print feature to be certified and institute are calculated using K-means clustering algorithm
State the vocal print similarity between each vocal print feature of target sound-groove model.
6. a kind of voiceprint authentication apparatus, which is characterized in that described device includes:
Unit is established, for pre-establishing sound-groove model library；
First extraction unit for obtaining the voice print database to be certified of user's input, and is mentioned from the voice print database to be certified
Take corresponding vocal print feature to be certified；
Determination unit, for determining target sound-groove model from the sound-groove model library according to the vocal print feature to be certified；
Computing unit, for calculating the vocal print similarity between the vocal print feature to be certified and target sound-groove model；
Judging unit, for obtaining the vocal print similarity and judging whether the vocal print similarity is greater than preset threshold；
Judging unit determines that the voice print database to be certified is corresponding if being greater than the preset threshold for the vocal print similarity
User identity match with target user's identity, and determine that the user identity authentication passes through.
7. device as claimed in claim 6, which is characterized in that described to establish unit, comprising:
Acquisition unit, for acquiring at least two training speech samples for being used for model training；
Pretreatment unit, for being pre-processed to each each trained speech samples；
Second extraction unit, for extracting the vocal print feature of pretreated each trained speech samples；
Subelement is established, each vocal print feature for that will extract carries out gauss hybrid models training, obtains different vocal print moulds
Type, and the sound-groove model library is formed by the different sound-groove model.
8. device as claimed in claim 6, which is characterized in that the determination unit, comprising:
First determines subelement, for determining the quantity of extracted vocal print feature to be certified；
Second determines subelement, if be multiple for extracted vocal print feature to be certified, determines that each vocal print to be certified is special
The priority of sign；
Third determines subelement, and mesh is determined from the sound-groove model library for the vocal print feature to be certified according to highest priority
Mark sound-groove model.
9. a kind of computer equipment, including memory, processor and it is stored on the memory and can be on the processor
The computer program of operation, which is characterized in that the processor realizes that claim 1-5 such as appoints when executing the computer program
Voiceprint authentication method described in one.
10. a kind of computer readable storage medium, which is characterized in that the computer-readable recording medium storage have one or
More than one computer program, the one or more computer program can be by one or more than one processors
It executes, to realize voiceprint authentication method as described in any one in claim 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811487395.7A CN109243465A (en) | 2018-12-06 | 2018-12-06 | Voiceprint authentication method, device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811487395.7A CN109243465A (en) | 2018-12-06 | 2018-12-06 | Voiceprint authentication method, device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109243465A true CN109243465A (en) | 2019-01-18 |
Family
ID=65073890
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811487395.7A Pending CN109243465A (en) | 2018-12-06 | 2018-12-06 | Voiceprint authentication method, device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109243465A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109902957A (en) * | 2019-02-28 | 2019-06-18 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN110197172A (en) * | 2019-06-10 | 2019-09-03 | 清华大学 | One kind carrying out identity authentication method and device based on photoelectricity capacity of blood vessel information |
| CN111243601A (en) * | 2019-12-31 | 2020-06-05 | 北京捷通华声科技股份有限公司 | Voiceprint clustering method and device, electronic equipment and computer-readable storage medium |
| WO2020199473A1 (en) * | 2019-04-04 | 2020-10-08 | 平安科技（深圳）有限公司 | Voice password verification method and apparatus, storage medium, and computer device |
| CN111833068A (en) * | 2020-07-31 | 2020-10-27 | 重庆富民银行股份有限公司 | Identity verification system and method based on voiceprint recognition |
| CN111833882A (en) * | 2019-03-28 | 2020-10-27 | 阿里巴巴集团控股有限公司 | Voiceprint information management method, device and system, computing equipment and storage medium |
| CN111933147A (en) * | 2020-06-22 | 2020-11-13 | 厦门快商通科技股份有限公司 | Voiceprint recognition method, system, mobile terminal and storage medium |
| CN112164404A (en) * | 2020-10-28 | 2021-01-01 | 广西电网有限责任公司贺州供电局 | Remote identity authentication method and system based on voiceprint recognition technology |
| CN112992154A (en) * | 2021-05-08 | 2021-06-18 | 北京远鉴信息技术有限公司 | Voice identity determination method and system based on enhanced voiceprint library |
| WO2021139589A1 (en) * | 2020-01-10 | 2021-07-15 | 华为技术有限公司 | Voice processing method, medium, and system |
| CN113327617A (en) * | 2021-05-17 | 2021-08-31 | 西安讯飞超脑信息科技有限公司 | Voiceprint distinguishing method and device, computer equipment and storage medium |
| CN113327618A (en) * | 2021-05-17 | 2021-08-31 | 西安讯飞超脑信息科技有限公司 | Voiceprint distinguishing method and device, computer equipment and storage medium |
| CN113366567A (en) * | 2021-05-08 | 2021-09-07 | 腾讯音乐娱乐科技（深圳）有限公司 | Voiceprint identification method, singer authentication method, electronic equipment and storage medium |
| CN113457096A (en) * | 2020-03-31 | 2021-10-01 | 荣耀终端有限公司 | Method for detecting basketball movement based on wearable device and wearable device |
| CN115021937A (en) * | 2022-06-21 | 2022-09-06 | 中国银行股份有限公司 | User identity authentication method, system, electronic equipment and storage medium |
| CN113366567B (en) * | 2021-05-08 | 2024-06-04 | 腾讯音乐娱乐科技（深圳）有限公司 | Voiceprint recognition method, singer authentication method, electronic equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1447278A (en) * | 2002-11-15 | 2003-10-08 | 郑方 | Method for recognizing voice print |
| CN103714817A (en) * | 2013-12-31 | 2014-04-09 | 厦门天聪智能软件有限公司 | Satisfaction survey cheating screening method based on voiceprint recognition technology |
| CN107680582A (en) * | 2017-07-28 | 2018-02-09 | 平安科技（深圳）有限公司 | Acoustic training model method, audio recognition method, device, equipment and medium |
| CN107993663A (en) * | 2017-09-11 | 2018-05-04 | 北京航空航天大学 | A kind of method for recognizing sound-groove based on Android |
| CN108040032A (en) * | 2017-11-02 | 2018-05-15 | 阿里巴巴集团控股有限公司 | A kind of voiceprint authentication method, account register method and device |
| CN108297108A (en) * | 2018-02-06 | 2018-07-20 | 上海交通大学 | A kind of spherical shape follows robot and its follow-up control method |
-
2018
- 2018-12-06 CN CN201811487395.7A patent/CN109243465A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1447278A (en) * | 2002-11-15 | 2003-10-08 | 郑方 | Method for recognizing voice print |
| CN103714817A (en) * | 2013-12-31 | 2014-04-09 | 厦门天聪智能软件有限公司 | Satisfaction survey cheating screening method based on voiceprint recognition technology |
| CN107680582A (en) * | 2017-07-28 | 2018-02-09 | 平安科技（深圳）有限公司 | Acoustic training model method, audio recognition method, device, equipment and medium |
| CN107993663A (en) * | 2017-09-11 | 2018-05-04 | 北京航空航天大学 | A kind of method for recognizing sound-groove based on Android |
| CN108040032A (en) * | 2017-11-02 | 2018-05-15 | 阿里巴巴集团控股有限公司 | A kind of voiceprint authentication method, account register method and device |
| CN108297108A (en) * | 2018-02-06 | 2018-07-20 | 上海交通大学 | A kind of spherical shape follows robot and its follow-up control method |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109902957A (en) * | 2019-02-28 | 2019-06-18 | 腾讯科技（深圳）有限公司 | A kind of data processing method and device |
| CN109902957B (en) * | 2019-02-28 | 2022-12-09 | 腾讯科技（深圳）有限公司 | Data processing method and device |
| CN111833882A (en) * | 2019-03-28 | 2020-10-27 | 阿里巴巴集团控股有限公司 | Voiceprint information management method, device and system, computing equipment and storage medium |
| WO2020199473A1 (en) * | 2019-04-04 | 2020-10-08 | 平安科技（深圳）有限公司 | Voice password verification method and apparatus, storage medium, and computer device |
| CN110197172A (en) * | 2019-06-10 | 2019-09-03 | 清华大学 | One kind carrying out identity authentication method and device based on photoelectricity capacity of blood vessel information |
| CN111243601A (en) * | 2019-12-31 | 2020-06-05 | 北京捷通华声科技股份有限公司 | Voiceprint clustering method and device, electronic equipment and computer-readable storage medium |
| CN111243601B (en) * | 2019-12-31 | 2023-04-07 | 北京捷通华声科技股份有限公司 | Voiceprint clustering method and device, electronic equipment and computer-readable storage medium |
| WO2021139589A1 (en) * | 2020-01-10 | 2021-07-15 | 华为技术有限公司 | Voice processing method, medium, and system |
| CN113457096B (en) * | 2020-03-31 | 2022-06-24 | 荣耀终端有限公司 | Method for detecting basketball movement based on wearable device and wearable device |
| CN113457096A (en) * | 2020-03-31 | 2021-10-01 | 荣耀终端有限公司 | Method for detecting basketball movement based on wearable device and wearable device |
| CN111933147B (en) * | 2020-06-22 | 2023-02-14 | 厦门快商通科技股份有限公司 | Voiceprint recognition method, system, mobile terminal and storage medium |
| CN111933147A (en) * | 2020-06-22 | 2020-11-13 | 厦门快商通科技股份有限公司 | Voiceprint recognition method, system, mobile terminal and storage medium |
| CN111833068A (en) * | 2020-07-31 | 2020-10-27 | 重庆富民银行股份有限公司 | Identity verification system and method based on voiceprint recognition |
| CN112164404A (en) * | 2020-10-28 | 2021-01-01 | 广西电网有限责任公司贺州供电局 | Remote identity authentication method and system based on voiceprint recognition technology |
| CN112992154A (en) * | 2021-05-08 | 2021-06-18 | 北京远鉴信息技术有限公司 | Voice identity determination method and system based on enhanced voiceprint library |
| CN113366567A (en) * | 2021-05-08 | 2021-09-07 | 腾讯音乐娱乐科技（深圳）有限公司 | Voiceprint identification method, singer authentication method, electronic equipment and storage medium |
| CN113366567B (en) * | 2021-05-08 | 2024-06-04 | 腾讯音乐娱乐科技（深圳）有限公司 | Voiceprint recognition method, singer authentication method, electronic equipment and storage medium |
| CN113327617A (en) * | 2021-05-17 | 2021-08-31 | 西安讯飞超脑信息科技有限公司 | Voiceprint distinguishing method and device, computer equipment and storage medium |
| CN113327617B (en) * | 2021-05-17 | 2024-04-19 | 西安讯飞超脑信息科技有限公司 | Voiceprint discrimination method, voiceprint discrimination device, computer device and storage medium |
| CN113327618B (en) * | 2021-05-17 | 2024-04-19 | 西安讯飞超脑信息科技有限公司 | Voiceprint discrimination method, voiceprint discrimination device, computer device and storage medium |
| CN113327618A (en) * | 2021-05-17 | 2021-08-31 | 西安讯飞超脑信息科技有限公司 | Voiceprint distinguishing method and device, computer equipment and storage medium |
| CN115021937A (en) * | 2022-06-21 | 2022-09-06 | 中国银行股份有限公司 | User identity authentication method, system, electronic equipment and storage medium |
| CN115021937B (en) * | 2022-06-21 | 2024-02-09 | 中国银行股份有限公司 | User identity authentication method, system, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109243465A (en) | Voiceprint authentication method, device, computer equipment and storage medium | |
| CN107492382B (en) | Voiceprint information extraction method and device based on neural network | |
| US9990915B2 (en) | Systems and methods for multi-style speech synthesis | |
| CN113470615B (en) | Cross-speaker style transfer speech synthesis | |
| CN109817246A (en) | Training method, emotion identification method, device, equipment and the storage medium of emotion recognition model | |
| CN108460081A (en) | Voice data base establishing method, voiceprint registration method, apparatus, equipment and medium | |
| CN107767869A (en) | Method and apparatus for providing voice service | |
| US10255903B2 (en) | Method for forming the excitation signal for a glottal pulse model based parametric speech synthesis system | |
| CN110197658A (en) | Method of speech processing, device and electronic equipment | |
| CN110570876A (en) | Singing voice synthesis method and device, computer equipment and storage medium | |
| CN112735371B (en) | Method and device for generating speaker video based on text information | |
| CN112382300A (en) | Voiceprint identification method, model training method, device, equipment and storage medium | |
| Rashmi | Review of algorithms and applications in speech recognition system | |
| US10014007B2 (en) | Method for forming the excitation signal for a glottal pulse model based parametric speech synthesis system | |
| EP3363015A1 (en) | Method for forming the excitation signal for a glottal pulse model based parametric speech synthesis system | |
| CN111599339A (en) | Speech splicing synthesis method, system, device and medium with high naturalness | |
| CN114283783A (en) | Speech synthesis method, model training method, device and storage medium | |
| CN108880815A (en) | Auth method, device and system | |
| CN113782032A (en) | Voiceprint recognition method and related device | |
| CN108665901A (en) | A kind of phoneme/syllable extracting method and device | |
| CN110298150B (en) | Identity verification method and system based on voice recognition | |
| CA3178027A1 (en) | Method for forming the excitation signal for a glottal pulse model based parametric speech synthesis system | |
| CN113555003B (en) | Speech synthesis method, device, electronic equipment and storage medium | |
| Gao | Audio deepfake detection based on differences in human and machine generated speech | |
| Nose et al. | HMM-based speech synthesis with unsupervised labeling of accentual context based on F0 quantization and average voice model |
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 |