CN115238260A - Dual-identification authentication method and device, computer equipment and storage medium - Google Patents

Dual-identification authentication method and device, computer equipment and storage medium Download PDF

Info

Publication number
CN115238260A
CN115238260A CN202110437704.5A CN202110437704A CN115238260A CN 115238260 A CN115238260 A CN 115238260A CN 202110437704 A CN202110437704 A CN 202110437704A CN 115238260 A CN115238260 A CN 115238260A
Authority
CN
China
Prior art keywords
face
person
information
voiceprint
recognized
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
Application number
CN202110437704.5A
Other languages
Chinese (zh)
Inventor
李瑞敏
吴云清
王盛
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shanghai Aviation Electric Co Ltd
Original Assignee
Shanghai Aviation Electric Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shanghai Aviation Electric Co Ltd filed Critical Shanghai Aviation Electric Co Ltd
Priority to CN202110437704.5A priority Critical patent/CN115238260A/en
Publication of CN115238260A publication Critical patent/CN115238260A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/44—Program or device authentication
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
    • G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
    • G06F21/31—User authentication
    • G06F21/32—User authentication using biometric data, e.g. fingerprints, iris scans or voiceprints
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Security & Cryptography (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Collating Specific Patterns (AREA)

Abstract

The invention discloses a double-identification authentication method, a double-identification authentication device, computer equipment and a storage medium. The double identification and authentication method comprises the steps of obtaining face information and voice information of a person to be identified; extracting face key features in face information of a person to be identified; extracting the voice print characteristics of a speaker in the voice information of a person to be identified; and matching the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information with the voiceprint information base of the pilot face, and if the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information of the person to be recognized both point to the same preset pilot in the voiceprint information base of the pilot face, indicating that the person to be recognized passes the authentication. The invention has the beneficial effects that: identity authentication can be carried out through two biological indicators, namely the face and the voiceprint of a person to be identified, and the accuracy of identity identification is greatly improved.

Description

Dual-identification authentication method and device, computer equipment and storage medium
Technical Field
The invention relates to a dual identification authentication method, a dual identification authentication device, a computer device and a storage medium.
Background
At present, the face and voiceprint recognition method based on the onboard equipment mainly has the following limitations: (1) The noise under the flying environment is large, about 70 decibels, the noise can cause interference to audio signals, and is not beneficial to extracting speaker information contained in the voice; (2) Face recognition/voiceprint recognition has the possibility of being broken by materials such as photographs, sound recordings, and the like. Therefore, there is a need for an authentication method that can accurately identify the identity of a pilot and is difficult to be broken by materials such as photographs and recordings.
Disclosure of Invention
The present invention is directed to solve the problems in the prior art, and provides a novel dual-recognition authentication method and apparatus, a computer device, and a storage medium.
In order to achieve the purpose, the technical scheme of the invention is as follows: the double identification and authentication method is used for the identity authentication of the aircraft to the pilot and comprises the following steps,
acquiring face information and voice information of a person to be identified;
extracting face key features in face information of a person to be recognized;
extracting the voice print characteristics of a speaker in the voice information of a person to be identified;
matching the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information with a voiceprint information base of the pilot human face, and if the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information point to the same preset pilot in the voiceprint information base of the pilot human face, indicating that the person to be recognized passes the authentication; otherwise, the authentication of the person to be identified is not passed.
As an optimal scheme of the dual recognition authentication method, the extraction of the key human face features in the human face information of the person to be recognized is completed by using a human face recognition model.
The method for acquiring the face recognition model comprises the following steps,
providing a face data set for training, wherein the face data set for training comprises published face data and/or face data collected from a preset pilot;
and (3) building a convolutional neural network for face recognition, and training by using a face data set for training to obtain a face recognition model.
As a preferred scheme of the double recognition authentication method, the voice print feature of the speaker in the voice information of the person to be recognized is extracted by using a voice print recognition model.
The method for training and acquiring the voiceprint recognition model comprises the following steps,
providing a voiceprint data set for training, wherein the voiceprint data set for training comprises public voiceprint data and/or voiceprint data of a preset pilot is collected;
building a convolution neural network for voiceprint recognition, and training by using a voiceprint data set for training to obtain a voiceprint recognition model;
training includes training a universal background model for voiceprints UBM, training an I-vector extractor, and/or training a PLDA scoring model.
As a preferred scheme of the dual recognition authentication method, the voice information of the person to be recognized is preprocessed in the voice fingerprint feature of the speaker in the voice information of the person to be recognized, wherein the preprocessing comprises pre-emphasis, framing and windowing; after preprocessing, extracting 20-dimensional mfcc characteristic information, adding first-order and second-order difference, calculating a cepstrum mean value, and generating 60-dimensional mfcc characteristics; and finally, extracting the voice print feature of the speaker from the 60-dimensional mfcc feature by utilizing a GMM-UBM and i-vector extractor.
As a preferred scheme of the dual identification authentication method, the human face key features in the human face information of the personnel to be identified are compared with the similarity of the human face data of each preset pilot in the pilot human face voiceprint information base; if the similarity is greater than the face authentication threshold, the person to be identified is authenticated as the preset pilot, and preferably, the face authentication threshold is defined as 90%;
performing PLDA scoring on voice print characteristics of a speaker in voice information of a person to be identified and voice print data of each preset pilot in a pilot face voice print information base; and if the PLDA value is larger than 0, the person to be identified is identified as the preset pilot.
The invention also provides a dual identification authentication device, comprising,
the system comprises an acquisition module, a recognition module and a recognition module, wherein the acquisition module is configured to acquire face information and voice information of a person to be recognized;
the first extraction module is configured to extract human face key features in human face information of a person to be recognized;
the second extraction module is configured to extract the voice print characteristics of the speaker in the voice information of the person to be identified;
the authentication module is configured to match the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information with the voiceprint information base of the pilot face, and if the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information point to the same preset pilot in the voiceprint information base of the pilot face, the authentication of the person to be recognized is passed; otherwise, the authentication of the person to be identified is not passed.
The present invention also provides a computer device including a processor and a memory, wherein the memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the operations performed by the dual identification authentication method.
The present invention also provides a storage medium, which stores at least one instruction that is loaded and executed by a processor to implement the operations performed by the dual identification authentication method.
Compared with the prior art, the invention has the beneficial effects that: the identity authentication can be carried out through two biological indications, namely the face and the voiceprint of the person to be identified, the accuracy of the identity authentication is greatly improved, meanwhile, the situation that an unauthorized person identifies fraud through a photo or a recording can be prevented, and the safety of a pilot identity authentication system is guaranteed.
Drawings
FIG. 1 is a flow chart of the present invention.
Fig. 2 is a flow chart of face recognition in the present invention.
Fig. 3 is a flow chart of voiceprint recognition in the present invention.
Detailed Description
The invention will be described in further detail below with reference to specific embodiments and drawings. Here, the description of the embodiments is provided to help understanding of the present invention, but the present invention is not limited thereto. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
Referring to fig. 1, a high-security dual identification authentication method for an onboard device is shown, which includes the following steps:
a, training a face recognition model and a voice print recognition model by respectively using a face data set and a voice data set;
b, acquiring the face and voice data of the pilot to be registered in a face library and a voiceprint library of the pilot;
step C, for the person to be identified, acquiring face information and voice information of the person by using a camera and a microphone respectively;
step D, extracting key characteristic information of the human face by using a trained human face recognition model for the captured human face information;
step E, extracting voice print information of the speaker by using the trained voice print recognition model for the captured voice information;
step F, matching the key feature information of the human face and the voice print information of the speaker with data in a pilot human face library and a voice print library respectively;
and G, if the face recognition result and the voiceprint recognition result are authenticated as the same person, the authentication is passed, the corresponding authority is opened, and if not, the authentication is refused and re-authenticated.
A method for training a face recognition model and a voiceprint recognition model by using a face data set and a voice data set in the step A comprises the following steps:
a. preparing a face data set and a voice data set by utilizing the open source data set and the collected pilot data;
b. and (4) building a convolutional neural network for face recognition, and training on a face data set.
c. On a voice data set, extracting mfcc characteristics, training a general background model GMM-UBM of a voiceprint Gaussian mixture model, training an I-vector extractor and training a PLDA scoring model.
According to the invention, the data volume is enlarged by using the public data set, the model fitting and generalization capability is enhanced, and meanwhile, the acquired pilot data is added, so that the model has higher robustness.
Referring to fig. 2, in the present invention, the pilot face registration and identification method in steps B and D is as follows:
a. for the collected pilot face data, extracting characteristic values of the pilot face data by using a trained neural network model, and storing the characteristic values in a pilot face library;
b. for the face data to be recognized captured by a camera, acquiring depth information and color information of the data by using a trained neural network model, calibrating face region coordinates, and acquiring key feature values of the face;
c. the face characteristic values obtained in the step b are matched with the face characteristic values registered in the voice print library in the step a one by one, and if the similarity is larger than 90%, the matching is judged to be successful;
referring to fig. 3, in the present invention, the method for registering and identifying the voiceprint of the pilot in steps B and E is as follows:
a. preprocessing the collected pilot voice data, extracting the mfcc characteristics, extracting the i-vector by using a trained UBM and i-vector extractor, and storing the i-vector in a pilot voiceprint library.
b. And preprocessing the voice of the person to be recognized captured by the microphone, extracting the mfcc characteristic, and extracting the i-vector by using the trained UBM and the i-vector extractor.
c. And (4) scoring and judging the human voice print features i-vector to be recognized extracted from the step (b) and the i-vector in the voice print library in the step (a) one by using a trained PLDA scoring model, wherein if the score of the PLDA is greater than 0, the voice print matching is successful, and if the score of the PLDA is smaller than 0, the matching is failed.
The invention requires that the face recognition and the voiceprint recognition are successfully matched and the recognition result is the same person, the recognition is judged to be successful, the corresponding authority is opened, the recognition is judged to be failed by any recognition channel matching failure or the two results are different, and the authority is refused to be opened.
In conclusion, the high-safety double-identification authentication method for the airborne equipment greatly improves the accuracy of identity authentication of the pilot to be tested, and the safety of the aircraft permission is improved by the double-authentication mode.
Based on the above-mentioned dual identification authentication method, those skilled in the art can specifically obtain the following:
a dual-recognition authentication device comprises a first recognition module,
the system comprises an acquisition module, a recognition module and a recognition module, wherein the acquisition module is configured to acquire face information and voice information of a person to be recognized;
the first extraction module is configured to extract human face key features in human face information of a person to be recognized;
the second extraction module is configured to extract the voice print characteristics of the speaker in the voice information of the person to be identified;
the authentication module is configured to match the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information with the voiceprint information base of the pilot face, and if the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information point to the same preset pilot in the voiceprint information base of the pilot face, the authentication of the person to be recognized is passed; otherwise, the authentication of the person to be identified is not passed.
A computer device comprising a processor and a memory, said memory having stored therein at least one instruction that is loaded and executed by said processor to perform the operations performed by the above-described dual identification authentication method.
A storage medium having stored therein at least one instruction that is loaded and executed by a processor to perform the operations performed by the above-described dual-recognition authentication method.
While the foregoing is directed to embodiments of the present invention, which are more particularly and specifically described herein, it is not intended to limit the scope of the invention. It should be noted that various changes and modifications can be made by those skilled in the art without departing from the spirit of the invention, and these changes and modifications are all within the scope of the invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims (10)

1. A dual identification authentication method for authenticating the identity of a pilot by an aircraft is characterized by comprising the following steps of,
acquiring face information and voice information of a person to be identified;
extracting face key features in face information of a person to be identified;
extracting the voice print characteristics of a speaker in the voice information of a person to be identified;
matching the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information with a pilot human face voiceprint information base, and if the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information point to the same preset pilot in the pilot human face voiceprint information base, indicating that the identity authentication of the person to be recognized passes; otherwise, the identity authentication of the person to be identified is not passed.
2. The dual recognition authentication method of claim 1, wherein extracting the key features of the face of the person to be recognized is performed by using a face recognition model.
3. The dual recognition authentication method according to claim 2, wherein the method for acquiring the face recognition model comprises,
providing a face data set for training, wherein the face data set for training comprises published face data and/or face data collected from a preset pilot;
and (3) building a convolutional neural network for face recognition, and training by using a face data set for training to obtain a face recognition model.
4. The dual recognition authentication method of claim 1, wherein extracting the voiceprint feature of the speaker in the voice information of the person to be recognized is performed by using a voiceprint recognition model.
5. The dual recognition authentication method of claim 4, wherein the training method of the voiceprint recognition model comprises,
providing a voiceprint data set for training, wherein the voiceprint data set for training comprises public voiceprint data and/or voiceprint data of a preset pilot is collected;
building a convolution neural network for voiceprint recognition, and training by using a voiceprint data set for training to obtain a voiceprint recognition model;
preferably, the training comprises training a universal background model for voiceprints UBM, training an I-vector extractor, and/or training a PLDA scoring model.
6. The dual recognition authentication method according to claim 1, wherein in extracting the voice print feature of the speaker in the voice information of the person to be recognized, the voice information of the person to be recognized is preprocessed, wherein the preprocessing includes pre-emphasis, framing, and windowing; after preprocessing, extracting 20-dimensional mfcc characteristic information, adding first-order and second-order difference, calculating a cepstrum mean value, and generating 60-dimensional mfcc characteristics; and finally, extracting the voice print feature of the speaker from the 60-dimensional mfcc feature by utilizing a GMM-UBM and i-vector extractor.
7. The dual identification authentication method of claim 1,
comparing the similarity between the key human face features in the human face information of the personnel to be identified and the human face data of each preset pilot in the pilot human face voiceprint information base; if the similarity is greater than the face authentication threshold, the person to be identified is authenticated as the preset pilot, and preferably, the face authentication threshold is defined as 90%;
performing PLDA scoring on voice print characteristics of a speaker in voice information of a person to be identified and voice print data of each preset pilot in a pilot face voice print information base; and if the PLDA value is larger than 0, the person to be identified is identified as the preset pilot.
8. The double identification and authentication device is characterized by comprising,
the system comprises an acquisition module, a recognition module and a recognition module, wherein the acquisition module is configured to acquire face information and voice information of a person to be recognized;
the first extraction module is configured to extract human face key features in human face information of a person to be recognized;
the second extraction module is configured to extract the voice print characteristics of the speaker in the voice information of the person to be identified;
the authentication module is configured to match the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information with the voiceprint information base of the pilot face, and if the key features of the human face in the human face information of the person to be recognized and the voiceprint features of the speaker in the voice information point to the same preset pilot in the voiceprint information base of the pilot face, the authentication of the person to be recognized is passed; otherwise, the authentication of the person to be identified is not passed.
9. Computer device, characterized in that it comprises a processor and a memory, in which at least one instruction is stored, which is loaded and executed by the processor to implement the operations performed by the method for dual identification authentication according to any one of claims 1 to 7.
10. Storage medium, characterized in that it has stored therein at least one instruction, which is loaded and executed by a processor to carry out the operations performed by the dual identification authentication method according to any one of claims 1 to 7.
CN202110437704.5A 2021-04-22 2021-04-22 Dual-identification authentication method and device, computer equipment and storage medium Pending CN115238260A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110437704.5A CN115238260A (en) 2021-04-22 2021-04-22 Dual-identification authentication method and device, computer equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110437704.5A CN115238260A (en) 2021-04-22 2021-04-22 Dual-identification authentication method and device, computer equipment and storage medium

Publications (1)

Publication Number Publication Date
CN115238260A true CN115238260A (en) 2022-10-25

Family

ID=83666548

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110437704.5A Pending CN115238260A (en) 2021-04-22 2021-04-22 Dual-identification authentication method and device, computer equipment and storage medium

Country Status (1)

Country Link
CN (1) CN115238260A (en)

Similar Documents

Publication Publication Date Title
WO2017215558A1 (en) Voiceprint recognition method and device
US10476872B2 (en) Joint speaker authentication and key phrase identification
CN103475490B (en) A kind of auth method and device
US9792912B2 (en) Method for verifying the identity of a speaker, system therefore and computer readable medium
US11869513B2 (en) Authenticating a user
CN102737634A (en) Authentication method and device based on voice
CN102543084A (en) Online voiceprint recognition system and implementation method thereof
US20070038460A1 (en) Method and system to improve speaker verification accuracy by detecting repeat imposters
GB2541466A (en) Replay attack detection
CN111611568A (en) Face voiceprint rechecking terminal and identity authentication method thereof
CN107346568B (en) Authentication method and device of access control system
JP2002514318A (en) System and method for detecting recorded speech
CN109448759A (en) A kind of anti-voice authentication spoofing attack detection method based on gas explosion sound
CN110827453A (en) Fingerprint and voiceprint double authentication method and authentication system
CN112382300A (en) Voiceprint identification method, model training method, device, equipment and storage medium
CN112751838A (en) Identity authentication method, device and system
CN111816185A (en) Method and device for identifying speaker in mixed voice
CN110379433A (en) Method, apparatus, computer equipment and the storage medium of authentication
KR20190085731A (en) Method for user authentication
CN112541174A (en) Service data verification method, device, equipment and storage medium
US6161094A (en) Method of comparing utterances for security control
CN109273012A (en) A kind of identity identifying method based on Speaker Identification and spoken digit recognition
KR20140076056A (en) Voice based CAPTCHA method and voice based CAPTCHA apparatus
KR20180049422A (en) Speaker authentication system and method
CN115238260A (en) Dual-identification authentication method and device, computer equipment and storage medium

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