CN113705410A - Face image desensitization processing and verifying method and system - Google Patents
Face image desensitization processing and verifying method and system Download PDFInfo
- Publication number
- CN113705410A CN113705410A CN202110965050.3A CN202110965050A CN113705410A CN 113705410 A CN113705410 A CN 113705410A CN 202110965050 A CN202110965050 A CN 202110965050A CN 113705410 A CN113705410 A CN 113705410A
- Authority
- CN
- China
- Prior art keywords
- desensitization
- face image
- face
- image
- desensitized
- 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.)
- Withdrawn
Links
- 238000000586 desensitisation Methods 0.000 title claims abstract description 170
- 238000000034 method Methods 0.000 title claims abstract description 60
- 238000012545 processing Methods 0.000 title claims abstract description 39
- 238000012549 training Methods 0.000 claims abstract description 29
- 238000013507 mapping Methods 0.000 claims abstract description 20
- 238000012795 verification Methods 0.000 claims description 22
- 230000000694 effects Effects 0.000 claims description 13
- 238000013528 artificial neural network Methods 0.000 claims description 8
- 238000007781 pre-processing Methods 0.000 claims description 8
- 125000004122 cyclic group Chemical group 0.000 claims description 7
- 238000003672 processing method Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 5
- 230000001815 facial effect Effects 0.000 claims description 3
- 238000013135 deep learning Methods 0.000 abstract description 4
- 238000004891 communication Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000008447 perception Effects 0.000 description 2
- 241001391944 Commicarpus scandens Species 0.000 description 1
- 230000003042 antagnostic effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
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
- G06F21/60—Protecting data
- G06F21/62—Protecting access to data via a platform, e.g. using keys or access control rules
- G06F21/6218—Protecting access to data via a platform, e.g. using keys or access control rules to a system of files or objects, e.g. local or distributed file system or database
- G06F21/6245—Protecting personal data, e.g. for financial or medical purposes
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Bioethics (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Computer Hardware Design (AREA)
- Computer Security & Cryptography (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a face image desensitization processing and verifying method and a face image desensitization verifying system. The method comprises the steps of training by using a deep learning circulation confrontation generation method to obtain a face image desensitization model, carrying out face positioning and cutting on a face image to be desensitized, inputting the face image to be desensitized into a preset face image desensitization model, mapping to a desensitization domain, and outputting a desensitization result. The desensitization result of the method realizes two requirements of same-domain matching and cross-domain desensitization, plays a role in protecting the privacy of the face, and is not perceived for a face recognition system.
Description
Technical Field
The invention relates to the technical field of face recognition, in particular to a face image desensitization processing and verifying method and system without face recognition perception.
Background
Nowadays, data has become one of the key elements in people's lives. The face image data is used for identification, authentication and the like, and is privacy data which is sensitive to individuals. At present, protection of sensitive private data, especially desensitization processing of face image data, is receiving more and more attention from all social circles.
In general, desensitization processing of a face image can be realized by establishing a desensitization rule, performing global or local deformation and transformation on the face image, and mapping the face image to a desensitization domain, so as to realize face privacy protection. In this scheme, the face image before desensitization processing is regarded as being in the original domain, and the face image after desensitization processing is regarded as being in the desensitization domain. According to the scheme, the cross-domain mismatch, which is the basic requirement of face desensitization, can be realized, and the specific meaning is as follows: when the face recognition system compares a desensitized face image collected from a person with a corresponding original face image, identity matching cannot be realized, namely face sensitive information is protected. Under the background of public safety, a further requirement for desensitization of the face image, namely same-domain matching, is provided, and the specific meaning is as follows: in the same domain (no matter the original domain or the desensitization domain), two face images (which are mapped to the same desensitization domain) which are acquired from the same person and are subjected to the same desensitization treatment, and the comparison result of the face recognition system can realize identity matching. The face recognition system under the scene is a general system and is not adjusted according to a desensitization method, so that the face image desensitization method provided by the invention is not perceived by the face recognition system. The key point for solving the problem of desensitization of the face image in the public security scene is to find a desensitization mapping rule which can simultaneously realize the two requirements.
One way to find satisfactory desensitization mapping rules is to train a round-robin generation network in two face style domains. The original domain is one of the face style domains, and the other preset face style domain is the desensitization domain. For face image data in real life, it is very difficult to form strictly paired data. The training method based on deep learning loop confrontation generation can perform a face image desensitization model under the condition that training data are not paired. Before training, a plurality of face style domains are selected as candidates of a desensitization domain based on image domain expert knowledge. After the training is finished, according to the desensitization verification indexes of the face images, a plurality of domains with the best desensitization effect are selected from the candidate domains to serve as desensitization domains which are selected finally, and the trained models corresponding to the desensitization domains are the face image desensitization models. The face image desensitization model based on deep learning essentially has the characteristics of a black box system, is low in interpretability and can be used as an advantage in data desensitization requirements, so that desensitization mapping rules are not easy to break through, and the face image desensitization model has high confidentiality.
The above concepts detail the technical background of the present invention, and can be summarized here as follows: firstly, the face image data needs privacy protection, so that face desensitization is required to be realized through a desensitization mapping rule; secondly, original-desensitized face image data pairs cannot be found in real life, so that model training is required to be carried out under the condition of unpaired data by adopting cyclic generation antagonistic learning; desensitization mapping rules need to ensure good confidentiality and safety and avoid being broken, so that a black box system is formed by using a neural network model; and fourthly, performing cross-domain mismatching and intra-domain matching on the face desensitization requirement in the public security scene, and therefore, considering design evaluation indexes to verify the effectiveness of the desensitization domain and selecting a desensitization mapping rule with a good desensitization effect.
Disclosure of Invention
The invention provides a method and a system for desensitizing and verifying a face image, aiming at training a face image desensitization model by using a deep learning loop confrontation generation technology, mapping the face image to a desensitization domain and playing the effects of desensitizing face image information and protecting privacy.
The application is realized by the following technical scheme:
a facial image desensitization processing and verification system comprising:
the face image database module is used for storing and reading the face image data to be desensitized;
the face image preprocessing module is used for positioning and cutting a face image;
the face image desensitization model training module is used for training a face image desensitization model;
the face image desensitization processing module is used for setting a desensitization model and performing desensitization processing on a face image to be desensitized to obtain a desensitization face image;
and the face image desensitization effect verification module is used for verifying the effectiveness of a desensitization processing result.
A human face image desensitization processing method comprises the following steps:
s1, acquiring a face image to be desensitized by using a face image database module;
s2, desensitizing the face image according to a desensitizing rule to obtain a desensitized face image;
and S3, outputting parameters according to a preset image to obtain a final image.
Preferably, before acquiring the face image to be desensitized, the method further comprises: processing an original image containing a human face, and positioning a human face part in the image; utilizing a face image preprocessing module to cut the face part according to the face position coordinates and the length and width information; the situation that a plurality of human faces exist in an image is adaptively processed.
Preferably, before desensitizing the face image, the method further comprises: training a neural network desensitization rule by using a cyclic confrontation generation method by using a face image desensitization model training module; training a plurality of groups of desensitization rules by using a face image desensitization processing module, and uniquely identifying each group of desensitization rules; the desensitization rules used are selected prior to desensitization treatment.
Preferably, during the desensitization treatment of the face image, the method further comprises: coding a face image to be desensitized into face features by using a face feature coder; and inputting the human face features into a neural network desensitization rule, mapping the human face features to a desensitization domain, and outputting a result which is a desensitized human face image.
Preferably, after obtaining the output desensitized face image, the method further comprises: and extracting the structural characteristic information of the final image, and storing the information, the desensitization face image and the unique identifier of the desensitization rule in a database, so that the information is convenient to read and query.
A face image desensitization processing verification method comprises the following steps:
(a) acquiring an original face image A, a face image B with the same identity as the original face image, an image A 'of the A after desensitization, and an image B' of the B after desensitization;
(b) respectively calculating the similarity of A and A ' and the similarity of A ' and B ' by using a face recognition system;
(c) and setting a face matching similarity threshold, and comparing the obtained similarity value with the set threshold to obtain a comparison result.
Preferably, after the step (b) obtains the face matching similarity comparison result, the method further includes: evaluating the effect of the desensitization method according to the comparison result;
preferably, the effective evaluation criterion of the desensitization method is that the original face image a and the desensitized face image a ' cannot be matched, and after the same desensitization mapping, two desensitized face images a ' and B ' with the same identity can be matched.
An electronic device, comprising:
(a) and the memory is used for storing the computer program and data used when the program runs.
(b) And a processor for implementing the steps of the human face image desensitization processing method or the verification method when executing the computer program.
In summary, the present invention comprises a facial image desensitization system that employs cyclic confrontation generation learning to achieve both cross-domain desensitization and intra-domain matching. The algorithm design, system design and verification method of the proposed method are all the contents of the invention. The invention does not relate to a face recognition system, but the face image desensitization verification method provided by the invention needs to be implemented based on the face recognition system, so that certain requirements are made on the technical specification of the face recognition system.
The invention has the beneficial effects that: the invention provides a face image desensitization processing and verification method and a face image desensitization processing and verification system. The human face desensitization processing method provided by the invention is effective, can realize the effects of intra-domain matching and cross-domain mismatching of the human face with the identity, ensures the confidentiality of desensitization mapping by using the black box characteristic of a neural network, and realizes the strict protection of the privacy information of the human face.
Drawings
In order to more clearly illustrate the embodiments and examples of the present invention, the drawings used in the description of the embodiments and examples will be briefly described.
FIG. 1 is a software module flow diagram of the present invention. Mainly expresses the specific method and the flow for implementing the face desensitization operation according to certain steps by using the method of the invention.
FIG. 2 is a schematic diagram of the training of the desensitization model of the face image according to the present invention. The main expression loop confrontation generation training method and process.
Fig. 3 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
The embodiments of the present invention will be described in detail below with reference to the accompanying drawings: the present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the protection scope of the present invention is not limited to the following embodiments.
As shown in FIG. 1, a human face desensitization system without perception for human face recognition comprises a human face image database module, a human face image preprocessing module, a human face image desensitization model training module, a human face image desensitization processing module and a human face image desensitization effect verification module.
When the system is deployed, firstly, a face image database service is started, and the face image preprocessing system can inquire and read data from the face image database. The training system of the face image desensitization model is started in advance and completes the training of the desensitization model, and a series of selectable desensitization models are provided. After the desensitization model is determined, the human face image desensitization processing system carries out desensitization processing on the preprocessed human face image by using the desensitization model to obtain the desensitized human face image. The face image desensitization effect verification system can be used for verifying the effectiveness of desensitization treatment.
As shown in fig. 2, in the aspect of face image desensitization model training, two groups of completely symmetrical model frameworks are used to form cyclic confrontation generation learning. In each group of model architectures, the method comprises the following steps: a discriminant model and a desensitization model (or inverse generative model). In the figure, the original domain is denoted as X and the desensitized domain is denoted as Y. X- > Y refers to the mapping from the original domain to the desensitized domain, and Y- > X refers to the mapping from the desensitized domain to the original domain.
A face image desensitization processing method comprises the following steps:
s1, acquiring a face image to be desensitized by using a face image database module;
s2, desensitizing the face image according to a desensitizing rule to obtain a desensitized face image;
and S3, outputting parameters according to a preset image to obtain a final image.
Before acquiring the face image to be desensitized, the method further comprises the following steps: processing an original image containing a human face, and positioning a human face part in the image; the face image preprocessing module is used for cutting the face part according to the face position coordinates and the length and width information; the situation that a plurality of human faces exist in an image is adaptively processed.
Before desensitizing the face image, the method further comprises the following steps: as shown in fig. 2, a training module of a face image desensitization model is used for training a neural network desensitization rule by using a cyclic confrontation generation method; training a plurality of groups of desensitization rules by using a face image desensitization processing module, and uniquely identifying each group of desensitization rules; the desensitization rules used are selected prior to desensitization treatment.
During desensitization treatment of the face image, the method further comprises the following steps: coding a face image to be desensitized into face features by using a face feature coder; and inputting the human face features into a neural network desensitization rule, mapping the human face features to a desensitization domain, and outputting a result which is a desensitized human face image.
After obtaining the output desensitized face image, the method further comprises: and extracting the structural characteristic information of the final image, and storing the information, the desensitization face image and the unique identifier of the desensitization rule in a database, so that the information is convenient to read and query.
Provided is a human face image desensitization processing system, comprising:
the face image database module is used for storing and reading the face image data to be desensitized;
the face image preprocessing module is used for positioning and cutting a face image;
the face image desensitization model training module is used for training a face image desensitization model;
the face image desensitization processing module is used for setting a desensitization model and performing desensitization processing on a face image to be desensitized to obtain a desensitization face image;
the face image desensitization effect verification module is used for verifying the effectiveness of a desensitization processing result;
a face image desensitization verification method comprises the following steps:
acquiring an original face image A, a face image B with the same identity as the original face image, an image A 'of the A after desensitization, and an image B' of the B after desensitization;
respectively calculating the similarity of A and A ' and the similarity of A ' and B ' by using a face recognition system;
and setting a face matching similarity threshold, and comparing the obtained similarity value with the set threshold to obtain a comparison result.
After obtaining the face matching similarity comparison result, the method further comprises the following steps: evaluating the effect of the desensitization method according to the comparison result; the effective desensitization evaluation standard is that the original face image A and the desensitized face image A ' cannot be matched, and after the same desensitization mapping, two desensitization face images A ' and B ' with the same identity can be matched.
An electronic device is used for building a face image desensitization system for face image desensitization processing and verification. As shown in fig. 3, the electronic device may include: a processor 310, a memory 320, a bus 330, and a communication interface 340. Wherein the processor 310 is used to execute the computer program for desensitization processing and verification of the face image, and the memory 320 is used to store the program and related data. The processor 310, the memory 320 and the communication interface 340 are connected and communicate via a bus 330 to accomplish data communication.
The invention has the beneficial effects that: a face image desensitization processing and verification method and a face image desensitization processing and verification system are provided, wherein a face detection model is used for cutting a face image from an original image, and desensitization processing is carried out on the face image by desensitization mapping obtained by training through a cyclic confrontation generation method. The human face desensitization processing method provided by the invention is effective, can realize the effects of intra-domain matching and cross-domain mismatching of the human face with the identity, ensures the confidentiality of desensitization mapping by using the black box characteristic of a neural network, and realizes the strict protection of the privacy information of the human face.
The foregoing shows and describes the general principles and broad features of the present invention and advantages thereof. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (10)
1. A facial image desensitization processing and verification system, comprising:
the face image database module is used for storing and reading the face image data to be desensitized;
the face image preprocessing module is used for positioning and cutting a face image;
the face image desensitization model training module is used for training a face image desensitization model;
the face image desensitization processing module is used for setting a desensitization model and performing desensitization processing on a face image to be desensitized to obtain a desensitization face image;
and the face image desensitization effect verification module is used for verifying the effectiveness of a desensitization processing result.
2. A human face image desensitization processing method is characterized by comprising the following steps:
s1, acquiring a face image to be desensitized by using a face image database module;
s2, desensitizing the face image according to a desensitizing rule to obtain a desensitized face image;
and S3, outputting parameters according to a preset image to obtain a final image.
3. A method for desensitizing human face image according to claim 2, wherein before the step S1 of acquiring the human face image to be desensitized, the method further comprises: processing an original image containing a human face, and positioning a human face part in the image; utilizing a face image preprocessing module to cut the face part according to the face position coordinates and the length and width information; the situation that a plurality of human faces exist in an image is adaptively processed.
4. A method for desensitizing human face images according to claim 2, wherein before the desensitizing process is performed on the human face images in step S2, the method further comprises: training a neural network desensitization rule by using a cyclic confrontation generation method by using a face image desensitization model training module; training a plurality of groups of desensitization rules by using a face image desensitization processing module, and uniquely identifying each group of desensitization rules; the desensitization rules used are selected prior to desensitization treatment.
5. A method for desensitizing human face images according to claim 2, wherein in the desensitizing process of step S2, the method further comprises: coding a face image to be desensitized into face features by using a face feature coder; and inputting the human face features into a neural network desensitization rule, mapping the human face features to a desensitization domain, and outputting a result which is a desensitized human face image.
6. The method for desensitizing human face image according to claim 2, wherein after obtaining the output desensitized human face image in step S2, the method further comprises: and extracting the structural characteristic information of the final image, and storing the information, the desensitization face image and the unique identifier of the desensitization rule in a database, so that the information is convenient to read and query.
7. A face image desensitization processing verification method is characterized by comprising the following steps:
(a) acquiring an original face image A, a face image B with the same identity as the original face image, an image A 'of the A after desensitization, and an image B' of the B after desensitization;
(b) respectively calculating the similarity of A and A ' and the similarity of A ' and B ' by using a face recognition system;
(c) and setting a face matching similarity threshold, and comparing the obtained similarity value with the set threshold to obtain a comparison result.
8. The method for desensitizing process verification of human face images according to claim 7, wherein after the human face matching similarity comparison result obtained in the step (b), the method further comprises: and evaluating the effect of the desensitization method according to the comparison result.
9. The method for desensitization processing verification according to claim 8, wherein the effective criteria for the desensitization method is that the original face image a and the desensitized face image a ' cannot be matched, and after the same desensitization mapping, two desensitization face images a ' and B ' of the same identity can be matched.
10. An electronic device, comprising:
and the memory is used for storing the computer program and data used when the program runs.
A processor for implementing the steps of the human face image desensitization processing method or authentication method according to any one of claims 1 to 9 when executing a computer program.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110965050.3A CN113705410A (en) | 2021-08-20 | 2021-08-20 | Face image desensitization processing and verifying method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110965050.3A CN113705410A (en) | 2021-08-20 | 2021-08-20 | Face image desensitization processing and verifying method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113705410A true CN113705410A (en) | 2021-11-26 |
Family
ID=78653802
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110965050.3A Withdrawn CN113705410A (en) | 2021-08-20 | 2021-08-20 | Face image desensitization processing and verifying method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113705410A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115408710A (en) * | 2022-08-30 | 2022-11-29 | 腾讯科技(深圳)有限公司 | Image desensitization method and related device |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101621723B1 (en) * | 2015-02-13 | 2016-05-19 | 경일대학교산학협력단 | Apparatus for processing picture adapted to protect privacy based on face recognition, method thereof and computer recordable medium storing the method |
US20190050592A1 (en) * | 2018-09-27 | 2019-02-14 | Intel IP Corporation | Systems and methods for processing and handling privacy-sensitive image data |
CN109872282A (en) * | 2019-01-16 | 2019-06-11 | 众安信息技术服务有限公司 | A kind of image desensitization method based on computer vision and system |
CN111476200A (en) * | 2020-04-27 | 2020-07-31 | 华东师范大学 | Face de-identification generation method based on generation of confrontation network |
CN112528318A (en) * | 2020-11-27 | 2021-03-19 | 国家电网有限公司大数据中心 | Image desensitization method and device and electronic equipment |
KR20210066390A (en) * | 2019-11-28 | 2021-06-07 | 한국전자기술연구원 | Personal Information De-identification Method and System |
CN113010904A (en) * | 2021-03-17 | 2021-06-22 | 腾讯科技(深圳)有限公司 | Data processing method and device and electronic equipment |
CN113222809A (en) * | 2021-05-21 | 2021-08-06 | 支付宝(杭州)信息技术有限公司 | Picture processing method and device for realizing privacy protection |
-
2021
- 2021-08-20 CN CN202110965050.3A patent/CN113705410A/en not_active Withdrawn
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101621723B1 (en) * | 2015-02-13 | 2016-05-19 | 경일대학교산학협력단 | Apparatus for processing picture adapted to protect privacy based on face recognition, method thereof and computer recordable medium storing the method |
US20190050592A1 (en) * | 2018-09-27 | 2019-02-14 | Intel IP Corporation | Systems and methods for processing and handling privacy-sensitive image data |
CN109872282A (en) * | 2019-01-16 | 2019-06-11 | 众安信息技术服务有限公司 | A kind of image desensitization method based on computer vision and system |
KR20210066390A (en) * | 2019-11-28 | 2021-06-07 | 한국전자기술연구원 | Personal Information De-identification Method and System |
CN111476200A (en) * | 2020-04-27 | 2020-07-31 | 华东师范大学 | Face de-identification generation method based on generation of confrontation network |
CN112528318A (en) * | 2020-11-27 | 2021-03-19 | 国家电网有限公司大数据中心 | Image desensitization method and device and electronic equipment |
CN113010904A (en) * | 2021-03-17 | 2021-06-22 | 腾讯科技(深圳)有限公司 | Data processing method and device and electronic equipment |
CN113222809A (en) * | 2021-05-21 | 2021-08-06 | 支付宝(杭州)信息技术有限公司 | Picture processing method and device for realizing privacy protection |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115408710A (en) * | 2022-08-30 | 2022-11-29 | 腾讯科技(深圳)有限公司 | Image desensitization method and related device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Galbally et al. | Iris image reconstruction from binary templates: An efficient probabilistic approach based on genetic algorithms | |
KR102503939B1 (en) | Face image de-identification apparatus and method | |
Galbally et al. | From the iriscode to the iris: A new vulnerability of iris recognition systems | |
Alheeti | Biometric iris recognition based on hybrid technique | |
Kant et al. | A watermarking based approach for protection of templates in multimodal biometric system | |
CN116453196B (en) | Face recognition method and system | |
Maček et al. | Multimodal biometric authentication in IoT: Single camera case study | |
Ahmed et al. | A state of art for survey of combined iris and fingerprint recognition systems | |
CN113705410A (en) | Face image desensitization processing and verifying method and system | |
CN110598522A (en) | Identity comparison method based on face and palm print palm vein recognition | |
CN112041847A (en) | Providing images with privacy tags | |
Zafar et al. | A novel crypt-biometric perception algorithm to protract security in MANET | |
CN111130794B (en) | Identity verification method based on iris and private key certificate chain connection storage structure | |
CN116886315A (en) | Authentication method based on biological characteristics and zero knowledge proof under web3.0 | |
Kekre et al. | Performance comparison of DCT and VQ based techniques for iris recognition | |
Esan et al. | Bimodal biometrics for financial infrastructure security | |
CN112733635B (en) | Object identification method and device and electronic equipment | |
CN108306738A (en) | A kind of method and system of identification identity | |
CN114612991A (en) | Conversion method and device for attacking face picture, electronic equipment and storage medium | |
CN103647769B (en) | A kind of coding encrypting and application process of human ear information | |
CN114373213A (en) | Juvenile identity recognition method and device based on face recognition | |
CN115708135A (en) | Face recognition model processing method, face recognition method and device | |
Vinothkanna et al. | A multimodal biometric approach for the recognition of finger print, palm print and hand vein using fuzzy vault | |
Mowla et al. | Selective fuzzy ensemble learner for cognitive detection of bio-identifiable modality spoofing in MCPS | |
Amirthalingam et al. | Multimodal biometric cryptosystem for face and ear recognition based on fuzzy vault |
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 | ||
WW01 | Invention patent application withdrawn after publication | ||
WW01 | Invention patent application withdrawn after publication |
Application publication date: 20211126 |