CN113726979A - Picture encryption method, decryption method, encryption system and related devices - Google Patents
Picture encryption method, decryption method, encryption system and related devices Download PDFInfo
- Publication number
- CN113726979A CN113726979A CN202110876887.0A CN202110876887A CN113726979A CN 113726979 A CN113726979 A CN 113726979A CN 202110876887 A CN202110876887 A CN 202110876887A CN 113726979 A CN113726979 A CN 113726979A
- Authority
- CN
- China
- Prior art keywords
- picture
- encrypted
- layer
- metadata
- encrypted data
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 65
- 238000013145 classification model Methods 0.000 claims abstract description 42
- 238000004364 calculation method Methods 0.000 claims abstract description 13
- 238000000605 extraction Methods 0.000 claims description 56
- 230000006870 function Effects 0.000 claims description 16
- 238000004422 calculation algorithm Methods 0.000 claims description 11
- 238000009795 derivation Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 125000004122 cyclic group Chemical group 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 16
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000009471 action Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 235000019738 Limestone Nutrition 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 239000006028 limestone Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/44—Secrecy systems
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Signal Processing (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Storage Device Security (AREA)
Abstract
The application provides a picture encryption method, which comprises the following steps: obtaining a picture to be encrypted, and performing forward calculation on the picture to be encrypted by utilizing a DNN classification model; the DNN classification model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a plurality of intermediate layers; recording metadata of the picture to be encrypted when the picture passes through each layer; and taking the metadata as the encrypted data of the picture to be encrypted. The encryption process is controllable, the encryption process is difficult to crack, and the obtained metadata can be selected and extracted according to the information quantity to serve as the encrypted data, so that the obtained encrypted data is far smaller than the original information. The application also provides a picture decryption method, a picture encryption system, a computer readable storage medium and an electronic device, which have the beneficial effects.
Description
Technical Field
The present application relates to the field of data encryption, and in particular, to a picture encryption method, a picture decryption method, a picture encryption system, and a related apparatus.
Background
At present, the encryption algorithm is divided into symmetric encryption and asymmetric encryption, wherein the encryption and decryption keys of the symmetric encryption algorithm are the same, the encryption key and decryption key of the asymmetric encryption algorithm are different, and besides, a hash algorithm without key is provided.
However, the key management of the symmetric algorithm is difficult, is not suitable for the internet and is generally used for an internal system; the safety is a middle grade; the encryption speed is high, the software encryption and decryption speed is at least 100 times faster, and the method is suitable for encryption and decryption processing of large data volume. The asymmetric algorithm has the advantages of easy key management, strong encryption property and high safety, but has a slow encryption speed, is suitable for encryption and decryption or data signature of small data volume, and has a slow speed when processing large data volume. However, whether symmetric encryption or asymmetric encryption is adopted, compared with original data, the data after encryption is increased by 2-3 times, and great difficulty is brought to data storage after encryption.
Disclosure of Invention
An object of the present application is to provide a picture encryption method, a picture decryption method, a picture encryption system, a computer-readable storage medium, and an electronic device, which can reduce the data size of encrypted information.
In order to solve the technical problem, the present application provides a picture encryption method, which has the following specific technical scheme:
obtaining a picture to be encrypted, and performing forward calculation on the picture to be encrypted by utilizing a DNN classification model; the DNN classification model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a plurality of intermediate layers;
recording metadata of the picture to be encrypted when the picture passes through each layer;
and taking the metadata as the encrypted data of the picture to be encrypted.
Optionally, the method further includes:
acquiring a pre-training model of the DNN classification model from a PyTorch library;
and retraining the last two layers of the pre-training model to obtain the DNN classification model.
The application also provides a picture decryption method, based on the above scheme, the encrypted data includes:
acquiring encrypted data, and taking a noise image or a pure color image as an initial input image of a DNN classification model;
taking the encrypted data as supervision data, and determining a metadata extraction strategy of the encrypted data;
extracting the total loss corresponding to the metadata extraction strategy; the metadata extraction strategy comprises an extraction layer number and a target extraction layer;
and carrying out reverse derivation on the total loss by using a BP algorithm to obtain a decrypted picture corresponding to the encrypted data.
Optionally, extracting the total loss corresponding to the metadata extraction policy includes:
determining a target extraction layer in the metadata extraction strategy;
determining a weight of each of the target extraction layers;
determining the loss of each target extraction layer according to a preset loss function and the weight;
and calculating the total loss of the encrypted data according to the loss of each target extraction layer.
Optionally, after obtaining the decrypted picture corresponding to the encrypted data, the method further includes:
and taking the decrypted picture as the initial input image, performing iterative updating by using an optimizer, stopping iteration when the iteration times meet a preset value, and outputting a cyclic decrypted picture.
Optionally, the performing iterative update by using the optimizer includes:
and reversely calculating the total loss, and iteratively updating the reverse calculation by utilizing the optimizer.
Optionally, the preset loss function is a euclidean distance or a KL divergence.
The present application further provides a picture encryption system, including:
the image acquisition module is used for acquiring an image to be encrypted and carrying out forward calculation on the image to be encrypted by utilizing a DNN classification model; the DNN classification model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a plurality of intermediate layers;
the data recording module is used for recording metadata when the picture to be encrypted passes through each layer;
and the encrypted data module is used for taking the metadata as the encrypted data of the picture to be encrypted.
The present application also provides a computer-readable storage medium having stored thereon a computer program which, when being executed by a processor, carries out the steps of the method as set forth above.
The present application further provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method described above when calling the computer program in the memory.
The application provides a picture encryption method, which comprises the following steps: obtaining a picture to be encrypted, and performing forward calculation on the picture to be encrypted by utilizing a DNN classification model; the DNN classification model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a plurality of intermediate layers; recording metadata of the picture to be encrypted when the picture passes through each layer; and taking the metadata as the encrypted data of the picture to be encrypted.
According to the method and the device, the DNN classification model is adopted to encrypt the picture to be encrypted, the picture to be encrypted is input into the DNN classification model, forward reasoning is carried out, metadata when the picture to be encrypted passes through each layer of the DNN model is obtained and serves as encrypted data, the encryption process is controllable due to the fact that the encryption process is related to the adopted DNN classification model, model parameters and a metadata extraction mode, the encryption process is difficult to crack, and partial metadata can be selected and extracted according to the information quantity to serve as encrypted data, so that the obtained encrypted data are far smaller than original information.
The application also provides an image decryption method, an image encryption system, a computer readable storage medium and an electronic device, which have the beneficial effects and are not repeated herein.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only embodiments of the present application, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
Fig. 1 is a flowchart of a picture encryption method according to an embodiment of the present application;
fig. 2 is a flowchart of a picture decryption method according to an embodiment of the present application;
fig. 3 is a schematic structural diagram of a picture encryption system according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Referring to fig. 1, fig. 1 is a flowchart of a picture encryption method according to an embodiment of the present application, and the present application provides a picture encryption method, which includes the following specific technical solutions:
s101: obtaining a picture to be encrypted, and performing forward calculation on the picture to be encrypted by utilizing a DNN classification model;
s102: recording metadata of the picture to be encrypted when the picture passes through each layer;
s103: and taking the metadata as the encrypted data of the picture to be encrypted.
The present embodiment defaults to obtaining or acquiring the DNN classification model before executing step S101. How to obtain the DNN classification model is not specifically limited, the DNN classification model may be directly obtained from a PyTorch library (a deep learning library), or a pre-trained model of the DNN classification model may be obtained from the PyTorch library, and then the last two layers of the pre-trained model are retrained to obtain the DNN classification model. The ImageNet pre-training model can be directly used, then the model is subjected to fine tuning, the required calculated amount is small, only one operation is needed, and the subsequent process can be repeatedly used.
It should be noted that the DNN classification model includes several layers, but the first layer is an input layer, the last layer is an output layer, and the middle includes several layers of hidden layers. Inputting the picture to be encrypted into the DNN classification model, firstly carrying out forward reasoning, recording metadata reaching each layer of the DNN classification model when the picture to be encrypted passes through the layer, and storing the metadata as encryption data. The metadata extraction may have various policies such as any layer, any number of layers, etc. The encryptor transmits the metadata to the decryptor. It is understood that, except for the input layer, the metadata of each layer may be recorded, but the encryption level may be freely set by those skilled in the art, so as to determine the layer corresponding to the metadata to be extracted and the number of layers, in other words, only one layer of metadata of the hidden layer needs to be included in the encryption process.
In addition, in step S102, metadata of the picture to be encrypted passing through each layer needs to be recorded, but in step S103, all metadata are not necessarily used as the encrypted data, and a person skilled in the art can select metadata of at least one layer of the hidden layers as the encrypted data by his/her own setting.
When a picture to be encrypted passes through each layer in the DNN classification model, an input weighted sum is usually calculated, and an activation function is called to calculate the output of the layer as metadata.
According to the method and the device, the picture to be encrypted is encrypted by adopting the DNN classification model, the picture to be encrypted is input into the DNN classification model, forward reasoning is carried out, metadata when the picture to be encrypted passes through each layer of the DNN model is obtained and serves as encrypted data, the encryption process is controllable and is difficult to crack due to the fact that the encryption process is related to the adopted DNN classification model, model parameters and a metadata extraction mode, and partial metadata can be selected and extracted according to the information quantity to serve as the encrypted data, so that the obtained encrypted data are far smaller than original information.
Referring to fig. 2, fig. 2 is a flowchart of an image decryption method provided in an embodiment of the present application, and the present application further provides an image decryption method, which decrypts encrypted data according to the above scheme, where the scheme includes:
s201: acquiring encrypted data, and taking a noise image or a pure color image as an initial input image of a DNN classification model;
the encrypted data obtained in this step refers to the encrypted data output in the previous embodiment, in this step, the noise map or the pure color map is first input into the DNN classification model as an initial input image, and after metadata is subsequently extracted and loss is calculated, the original image to be encrypted is gradually restored on the basis of the noise map or the pure color map.
S202: taking the encrypted data as supervision data, and determining a metadata extraction strategy of the encrypted data;
in this step, the encrypted data is required to be used as supervision data, and a metadata extraction policy of the encrypted data is determined. The metadata extraction strategy comprises an extraction layer number and a target extraction layer. It should be noted that the metadata extraction layer is slightly for data present in the encrypted data. For example, if the hidden layer includes five layers and the encrypted data is actually data of the first three layers of the hidden layer in the encryption process, the determined metadata extraction policy should also be within the range of the first three layers. In other words, the number of extraction layers in the metadata extraction policy should be no greater than the number of layers contained in the encrypted data. Of course, as the metadata extraction strategy is different, that is, the extracted metadata layers are different, the reduction degree of the obtained decrypted picture is also different accordingly.
S203: extracting the total loss corresponding to the metadata extraction strategy;
after the metadata extraction strategy is determined, the total loss can be calculated, and the following steps can be specifically adopted:
s2031: determining a target extraction layer in the metadata extraction strategy;
s2032: determining a weight of each of the target extraction layers;
s2033: determining the loss of each target extraction layer according to a preset loss function and the weight;
s2034: and calculating the total loss of the encrypted data according to the loss of each target extraction layer.
The method comprises the steps of firstly determining a target extraction layer, determining which layer of metadata needs to be extracted, determining corresponding weight of the layer of metadata, and then determining loss of the target extraction layer according to a preset loss function and the weight. The preset loss function may be a euclidean distance or a KL divergence, or may be other functions, and may be any function that can sufficiently describe a difference between two sets of data, which is not limited herein. Meanwhile, because the depths of all layers in the model are different, the information extracted by different layers is different, and the information extraction capability is also different, a certain weight can be set for the loss of each target extraction layer so as to realize decryption reproduction.
In other words, the purpose of this step is to obtain the total loss in the forward calculation according to the difference between the mean, variance and metadata obtained by the initial input image reaching each layer, and combining with the preset loss function.
S204: and carrying out reverse derivation on the total loss by using a BP algorithm to obtain a decrypted picture corresponding to the encrypted data.
After the total loss is obtained, a Back Propagation (error Back Propagation) algorithm can be used for carrying out reverse derivation on the total loss, then the total loss is subjected to reverse derivation, the reverse derivation is iteratively updated by using an optimizer, and the initial input image is updated based on the derivative, so that one recurrence of the initial input image is realized, and the input image can be continuously and iteratively updated in some details in a single recurrence manner.
For example, after a decrypted picture corresponding to the encrypted data is obtained, the decrypted picture may be used as the initial input image, an optimizer is used to perform iterative update, iteration is stopped when the iteration number meets a preset value, and a circular decrypted picture is output. Therefore, the recurrence is from point to surface, and the whole loop iteration process can observe that the weights of all layers in the DNN classification model can be controlled or the total loop times can be controlled, so that the recurrence steps are controllable and the recurrence can be explained. And the current AI-based model analysis theory cannot be cracked, so that the data security is ensured. The optimization function employed by the optimizer is not particularly limited herein. On the basis of the embodiment, different loss functions or optimization functions can be adopted, and the scheme can be realized.
According to the decryption process, the limestone discloses the image decryption method, the input image is reproduced by utilizing BP reverse derivation, controllable, interpretable, high-compression-ratio and non-interpretable information encryption and decryption are combined, the security of data encryption and decryption is ensured, meanwhile, the decryption process can be observed in a full flow, and the decryption state can be conveniently checked in real time.
In the following, a picture encryption system provided by an embodiment of the present application is introduced, and the picture encryption system described below and the picture encryption method described above may be referred to correspondingly.
Referring to fig. 3, fig. 3 is a schematic structural diagram of an image encryption system provided in an embodiment of the present application, and the present application further provides an image encryption system, including:
the image acquisition module is used for acquiring an image to be encrypted and carrying out forward calculation on the image to be encrypted by utilizing a DNN classification model; the DNN classification model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a plurality of intermediate layers;
the data recording module is used for recording metadata when the picture to be encrypted passes through each layer;
and the encrypted data module is used for taking the metadata as the encrypted data of the picture to be encrypted.
Based on the above embodiment, as a preferred embodiment, the method further includes:
the model obtaining module is used for obtaining a pre-training model of the DNN classification model from the PyTorch library; and retraining the last two layers of the pre-training model to obtain the DNN classification model.
In the following, a picture decryption system provided by an embodiment of the present application is introduced, and the picture decryption system described below and the picture decryption method described above may be referred to correspondingly. The present application further provides a picture decryption system, including:
the data acquisition module is used for acquiring encrypted data and taking the noise image or the pure color image as an initial input image of the DNN classification model;
the strategy determining module is used for determining a metadata extraction strategy of the encrypted data by taking the encrypted data as supervision data;
the loss calculation module is used for extracting the total loss corresponding to the metadata extraction strategy; the metadata extraction strategy comprises an extraction layer number and a target extraction layer;
and the decryption module is used for carrying out reverse derivation on the total loss by utilizing a BP algorithm to obtain a decrypted picture corresponding to the encrypted data.
Based on the above embodiment, as a preferred embodiment, the loss calculation module includes:
the layer number determining unit is used for determining a target extraction layer in the metadata extraction strategy;
a weight determination unit for determining a weight of each of the target extraction layers;
the layer loss determining unit is used for determining the loss of each target extraction layer according to a preset loss function and the weight;
and a total loss determination unit for calculating a total loss of the encrypted data based on the loss of each of the target extraction layers.
Based on the above embodiment, as a preferred embodiment, the method further includes:
and the iteration updating module is used for taking the decrypted picture as the initial input image, performing iteration updating by using an optimizer, stopping iteration when the iteration times meet a preset value, and outputting a cyclic decrypted picture.
Based on the above embodiment, as a preferred embodiment, the iterative update module includes:
and the optimizer unit is used for reversely deriving the total loss and iteratively updating the reverse derivation by utilizing the optimizer.
Based on the above embodiment, as a preferred embodiment, the preset loss function is a euclidean distance or a KL divergence.
The present application also provides a computer readable storage medium having stored thereon a computer program which, when executed, may implement the steps provided by the above-described embodiments. The storage medium may include: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The application further provides an electronic device, which may include a memory and a processor, where the memory stores a computer program, and the processor may implement the steps provided by the foregoing embodiments when calling the computer program in the memory. Of course, the electronic device may also include various network interfaces, power supplies, and the like.
The embodiments are described in a progressive manner in the specification, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. For the system provided by the embodiment, the description is relatively simple because the system corresponds to the method provided by the embodiment, and the relevant points can be referred to the method part for description.
The principles and embodiments of the present application are explained herein using specific examples, which are provided only to help understand the method and the core idea of the present application. It should be noted that, for those skilled in the art, it is possible to make several improvements and modifications to the present application without departing from the principle of the present application, and such improvements and modifications also fall within the scope of the claims of the present application.
It is further noted that, in the present specification, relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
Claims (10)
1. A picture encryption method, comprising:
obtaining a picture to be encrypted, and performing forward calculation on the picture to be encrypted by utilizing a DNN classification model; the DNN classification model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a plurality of intermediate layers;
recording metadata of the picture to be encrypted when the picture passes through each layer;
and taking the metadata as the encrypted data of the picture to be encrypted.
2. The picture encryption method according to claim 1, further comprising:
acquiring a pre-training model of the DNN classification model from a PyTorch library;
and retraining the last two layers of the pre-training model to obtain the DNN classification model.
3. A picture decryption method based on the encrypted data according to claim 1 or 2, comprising:
acquiring encrypted data, and taking a noise image or a pure color image as an initial input image of a DNN classification model;
taking the encrypted data as supervision data, and determining a metadata extraction strategy of the encrypted data;
extracting the total loss corresponding to the metadata extraction strategy; the metadata extraction strategy comprises an extraction layer number and a target extraction layer;
and carrying out reverse derivation on the total loss by using a BP algorithm to obtain a decrypted picture corresponding to the encrypted data.
4. The picture decryption method of claim 3, wherein extracting the total loss corresponding to the metadata extraction policy comprises:
determining a target extraction layer in the metadata extraction strategy;
determining a weight of each of the target extraction layers;
determining the loss of each target extraction layer according to a preset loss function and the weight;
and calculating the total loss of the encrypted data according to the loss of each target extraction layer.
5. The picture decryption method according to claim 4, wherein after obtaining the decrypted picture corresponding to the encrypted data, the method further comprises:
and taking the decrypted picture as the initial input image, performing iterative updating by using an optimizer, stopping iteration when the iteration times meet a preset value, and outputting a cyclic decrypted picture.
6. The picture decryption method of claim 5, wherein the iterative updating with the optimizer comprises:
and reversely calculating the total loss, and iteratively updating the reverse calculation by utilizing the optimizer.
7. The picture decryption method according to claim 4, wherein the predetermined loss function is Euclidean distance or KL divergence.
8. A picture encryption system, comprising:
the image acquisition module is used for acquiring an image to be encrypted and carrying out forward calculation on the image to be encrypted by utilizing a DNN classification model; the DNN classification model comprises an input layer, a hidden layer and an output layer, wherein the hidden layer comprises a plurality of intermediate layers;
the data recording module is used for recording metadata when the picture to be encrypted passes through each layer;
and the encrypted data module is used for taking the metadata as the encrypted data of the picture to be encrypted.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method according to any one of claims 1-2, or 3-7.
10. An electronic device, comprising a memory in which a computer program is stored and a processor which, when called from the memory, implements the steps of the method according to any one of claims 1-2, or 3-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110876887.0A CN113726979B (en) | 2021-07-31 | 2021-07-31 | Picture encryption method, picture decryption method, picture encryption system and related devices |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110876887.0A CN113726979B (en) | 2021-07-31 | 2021-07-31 | Picture encryption method, picture decryption method, picture encryption system and related devices |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113726979A true CN113726979A (en) | 2021-11-30 |
| CN113726979B CN113726979B (en) | 2024-04-26 |
Family
ID=78674491
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110876887.0A Active CN113726979B (en) | 2021-07-31 | 2021-07-31 | Picture encryption method, picture decryption method, picture encryption system and related devices |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113726979B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214193A (en) * | 2017-07-05 | 2019-01-15 | 阿里巴巴集团控股有限公司 | Data encryption, machine learning model training method, device and electronic equipment |
| CN109284313A (en) * | 2018-08-10 | 2019-01-29 | 深圳前海微众银行股份有限公司 | Federal modeling method, equipment and readable storage medium storing program for executing based on semi-supervised learning |
| CN110071798A (en) * | 2019-03-21 | 2019-07-30 | 深圳大学 | A kind of equivalent key acquisition methods, device and computer readable storage medium |
| CN111970519A (en) * | 2020-08-28 | 2020-11-20 | 中国人民解放军国防科技大学 | Airborne video return method |
| CN112926073A (en) * | 2021-03-17 | 2021-06-08 | 深圳前海微众银行股份有限公司 | Federal learning modeling optimization method, apparatus, medium, and computer program product |
| CN113008371A (en) * | 2021-03-05 | 2021-06-22 | 南京大学 | Hyperspectral imaging method for deep learning dispersion-based fuzzy solution |
| CN113051586A (en) * | 2021-03-10 | 2021-06-29 | 北京沃东天骏信息技术有限公司 | Federal modeling system and method, and federal model prediction method, medium, and device |
| CN113067832A (en) * | 2021-03-29 | 2021-07-02 | 郑州铁路职业技术学院 | Communication data encryption method based on block chain and artificial intelligence |
-
2021
- 2021-07-31 CN CN202110876887.0A patent/CN113726979B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109214193A (en) * | 2017-07-05 | 2019-01-15 | 阿里巴巴集团控股有限公司 | Data encryption, machine learning model training method, device and electronic equipment |
| CN109284313A (en) * | 2018-08-10 | 2019-01-29 | 深圳前海微众银行股份有限公司 | Federal modeling method, equipment and readable storage medium storing program for executing based on semi-supervised learning |
| CN110071798A (en) * | 2019-03-21 | 2019-07-30 | 深圳大学 | A kind of equivalent key acquisition methods, device and computer readable storage medium |
| CN111970519A (en) * | 2020-08-28 | 2020-11-20 | 中国人民解放军国防科技大学 | Airborne video return method |
| CN113008371A (en) * | 2021-03-05 | 2021-06-22 | 南京大学 | Hyperspectral imaging method for deep learning dispersion-based fuzzy solution |
| CN113051586A (en) * | 2021-03-10 | 2021-06-29 | 北京沃东天骏信息技术有限公司 | Federal modeling system and method, and federal model prediction method, medium, and device |
| CN112926073A (en) * | 2021-03-17 | 2021-06-08 | 深圳前海微众银行股份有限公司 | Federal learning modeling optimization method, apparatus, medium, and computer program product |
| CN113067832A (en) * | 2021-03-29 | 2021-07-02 | 郑州铁路职业技术学院 | Communication data encryption method based on block chain and artificial intelligence |
Non-Patent Citations (3)
| Title |
|---|
| YI QIN: "Information Encryption in Ghost Imaging With Customized Data Container and XOR Operation", 《 IEEE PHOTONICS JOURNAL》, 31 March 2017 (2017-03-31) * |
| 崔海霞;廖明群;: "基于人工神经网络的数据散列加密算法", 电子工程师, no. 05 * |
| 赵婷婷: "基于神经网络的视频加密与压缩技术的研究", 《中国优秀硕士学位论文全文数据库》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113726979B (en) | 2024-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11556730B2 (en) | Methods and apparatus for distributed use of a machine learning model | |
| US11386343B2 (en) | Real time detection of cyber threats using behavioral analytics | |
| JP6382354B2 (en) | Neural network and neural network training method | |
| US20220006629A1 (en) | Secure Analytics Using Term Generation and Homomorphic Encryption | |
| TWI655587B (en) | Neural network and method of neural network training | |
| US20200202014A1 (en) | Blockchain-based content management system, method, apparatus, and electronic device | |
| CN113239404A (en) | Federal learning method based on differential privacy and chaotic encryption | |
| US20190156057A1 (en) | Efficiently querying databases while providing differential privacy | |
| Kaur et al. | A secure data classification model in cloud computing using machine learning approach | |
| JP2023511514A (en) | Neural flow attestation | |
| CN112613001A (en) | Method for realizing integrity authentication of convolutional neural network through reversible watermark | |
| WO2022241307A1 (en) | Image steganography utilizing adversarial perturbations | |
| Mageshkumar et al. | An improved secure file deduplication avoidance using CKHO based deep learning model in a cloud environment | |
| CN113726979A (en) | Picture encryption method, decryption method, encryption system and related devices | |
| WO2020063349A1 (en) | Data protection method and device, apparatus, computer storage medium | |
| CN112800467B (en) | Online model training method, device and equipment based on data privacy protection | |
| CN114117487A (en) | Plaintext similarity estimation method, device, equipment and medium for encrypted character string | |
| KR102123435B1 (en) | Encryption method for supporting equality query in multi-client environment and apparatus using the same | |
| US9882879B1 (en) | Using steganography to protect cryptographic information on a mobile device | |
| Aritomo et al. | A privacy-preserving similarity search scheme over encrypted word embeddings | |
| Kwik | In Light of the Technical Impracticality of the Right to Erasure, What Answers Can Actor-Network Theories Provide | |
| AlShawi | Applying data mining techniques to improve information security in the cloud: a single cache system approach | |
| US11829239B2 (en) | Managing machine learning model reconstruction | |
| EP3902194A1 (en) | Method and apparatus for training analysis model | |
| WO2022239200A1 (en) | Learning device, inference device, learning method, and computer-readable 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |