CN113989156B - Method, apparatus, medium, device, and program for verifying reliability of desensitization method - Google Patents
Method, apparatus, medium, device, and program for verifying reliability of desensitization methodInfo
- Publication number
- CN113989156B CN113989156B CN202111283508.3A CN202111283508A CN113989156B CN 113989156 B CN113989156 B CN 113989156B CN 202111283508 A CN202111283508 A CN 202111283508A CN 113989156 B CN113989156 B CN 113989156B
- Authority
- CN
- China
- Prior art keywords
- image
- desensitized
- probability value
- desensitization
- evaluation information
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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 OR CALCULATING; 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
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
Abstract
A method, a device, a storage medium, equipment and a computer program for verifying reliability of a desensitization method are disclosed, wherein the method comprises the steps of carrying out desensitization processing on a first original image through a preset desensitization method to obtain a first desensitized image, carrying out image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image, estimating the first image to be a first probability value of a real image, and determining evaluation information based on the first probability value, wherein the evaluation information is used for evaluating the reliability of the preset desensitization method. The reliability of the image data desensitization method is evaluated.
Description
Technical Field
The present disclosure relates to the field of image processing technology, and in particular, to a method, an apparatus, a storage medium, an electronic device, and a computer program for verifying reliability of a desensitization method.
Background
In practice, for an image containing sensitive information, a desensitization process is required to prevent leakage of the sensitive information, and typically, a desensitization process (such as coding) is performed on a position where the sensitive information is located in the image, so as to obtain a desensitized image.
For desensitized images, the image may be restored by anti-desensitization techniques (e.g., by image restoration or restoration) to obtain sensitive information in the original image. The desensitized images generated by different data desensitization processing methods also have different performances when facing the anti-desensitization technology, namely, the reliability of the different data desensitization processing methods is different.
There is no method for verifying the reliability of the image data desensitization method in the related art.
Disclosure of Invention
The present disclosure has been made in order to solve the above technical problems. Embodiments of the present disclosure provide a method, apparatus, storage medium, electronic device, and computer program for reliability verification of a desensitization method.
According to one aspect of the embodiment of the disclosure, a method for verifying reliability of a desensitization method is provided, which comprises the steps of performing desensitization processing on a first original image through a preset desensitization method to obtain a first desensitized image, performing image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image, estimating a first probability value that the first image is a real image, and determining evaluation information based on the first probability value, wherein the evaluation information is used for evaluating the reliability of the preset desensitization method.
According to still another aspect of the embodiment of the present disclosure, there is provided an apparatus for verifying reliability of a desensitization method, including an image desensitization unit configured to perform desensitization processing on a first original image by a preset desensitization method to obtain a first desensitized image, an image restoration module configured to perform image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image, a probability prediction unit configured to estimate a first probability value that the first image is a true image, and an information generation unit configured to determine evaluation information for evaluating reliability of the preset desensitization method based on the first probability value.
According to still another aspect of the embodiments of the present disclosure, there is provided a computer-readable storage medium storing a computer program for performing the reliability verification of the desensitization method in the above embodiments.
According to yet another aspect of the disclosed embodiments, there is provided an electronic device comprising a processor, a memory for storing processor-executable instructions, and a processor for performing the method of reliability verification of the desensitization method of the above embodiments.
According to a further aspect of the disclosed embodiments, there is provided a computer program product comprising a computer program/instruction, characterized in that the computer program/instruction, when executed by a processor, implements a method of reliability verification of the desensitization method of the above embodiments.
According to the method, the device, the storage medium and the electronic equipment for verifying the reliability of the desensitization method, which are provided by the embodiment of the disclosure, the first original image is subjected to desensitization processing through a preset desensitization method to obtain a first desensitized image, then the first desensitized image is subjected to restoration processing to obtain a restored first image, the first image is estimated to be a first probability value of a real image, and evaluation information is generated based on the first probability value and used for evaluating the reliability of the preset desensitization method. The reliability of the image data desensitization method is evaluated.
The technical scheme of the present disclosure is described in further detail below through the accompanying drawings and examples.
Drawings
The above and other objects, features and advantages of the present disclosure will become more apparent by describing embodiments thereof in more detail with reference to the accompanying drawings. The accompanying drawings are included to provide a further understanding of embodiments of the disclosure and are incorporated in and constitute a part of this specification, illustrate embodiments of the disclosure and together with the description serve to explain the disclosure, not to limit the disclosure. In the drawings, like reference numerals generally refer to like parts or steps.
FIG. 1 is a schematic illustration of a scenario in which the present disclosure is applicable;
FIG. 2 is a flow chart of one embodiment of a method of reliability verification of the desensitization method of the present disclosure;
FIG. 3 is a flow chart of generating a first post-desensitized image in one embodiment of a method of reliability verification of the desensitization method of the present disclosure;
FIG. 4 is a flow chart of generating a first post-desensitized image in yet another embodiment of a method of reliability verification of the desensitization method of the present disclosure;
FIG. 5 is a flow chart of yet another embodiment of a method of reliability verification of the desensitization method of the present disclosure;
FIG. 6 is a flow chart of generating assessment information in one embodiment of a method of reliability verification of a desensitization method of the present disclosure;
FIG. 7 is a schematic structural view of one embodiment of an apparatus for reliability verification of the desensitization method of the present disclosure;
FIG. 8 is a schematic diagram of the structure of a probability prediction unit in one embodiment of an apparatus for reliability verification of the desensitization method of the present disclosure;
FIG. 9 is a schematic diagram of the structure of an image desensitizing unit in one embodiment of an apparatus for reliability verification of the desensitizing method of the present disclosure;
FIG. 10 is a schematic structural view of an image desensitizing unit in still another embodiment of an apparatus for reliability verification of the desensitizing method of the present disclosure;
FIG. 11 is a schematic diagram of the structure of an information generating unit in one embodiment of an apparatus for reliability verification of the desensitization method of the present disclosure;
FIG. 12 is a schematic diagram of the structure of an information generation module in one embodiment of an apparatus for reliability verification of the desensitization method of the present disclosure;
fig. 13 is a block diagram of an electronic device according to an exemplary embodiment of the present application.
Detailed Description
Example embodiments according to the present disclosure will be described in detail below with reference to the accompanying drawings. It should be apparent that the described embodiments are only some of the embodiments of the present disclosure and not all of the embodiments of the present disclosure, and that the present disclosure is not limited by the example embodiments described herein.
It should be noted that the relative arrangement of the components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present disclosure unless it is specifically stated otherwise.
It will be appreciated by those of skill in the art that the terms "first," "second," etc. in embodiments of the present disclosure are used merely to distinguish between different steps, devices or modules, etc., and do not represent any particular technical meaning nor necessarily logical order between them.
It should also be understood that in embodiments of the present disclosure, "plurality" may refer to two or more, and "at least one" may refer to one, two or more.
It should also be appreciated that any component, data, or structure referred to in the presently disclosed embodiments may be generally understood as one or more without explicit limitation or the contrary in the context.
In addition, the term "and/or" in the present disclosure is merely an association relationship describing the association object, and indicates that three relationships may exist, such as a and/or B, and may indicate that a exists alone, while a and B exist together, and B exists alone. In addition, the character "/" in the present disclosure generally indicates that the front and rear association objects are an or relationship.
It should also be understood that the description of the various embodiments of the present disclosure emphasizes the differences between the various embodiments, and that the same or similar features may be referred to each other, and for brevity, will not be described in detail.
Meanwhile, it should be understood that the sizes of the respective parts shown in the drawings are not drawn in actual scale for convenience of description.
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses.
Techniques, methods, and apparatus known to one of ordinary skill in the relevant art may not be discussed in detail, but are intended to be part of the specification where appropriate.
It should be noted that like reference numerals and letters refer to like items in the following figures, and thus once an item is defined in one figure, no further discussion thereof is necessary in subsequent figures.
Embodiments of the present disclosure are applicable to electronic devices such as terminal devices, computer systems, servers, etc., which are operational with numerous other general purpose or special purpose computing system environments or configurations. Examples of well known terminal devices, computing systems, environments, and/or configurations that may be suitable for use with an electronic device such as a terminal device, computer system, or server include, but are not limited to, personal computer systems, server computer systems, thin clients, thick clients, hand-held or laptop devices, microprocessor-based systems, set-top boxes, programmable consumer electronics, network personal computers, minicomputers systems, mainframe computer systems, distributed cloud computing environments that include any of the above systems, and the like.
Electronic devices such as terminal devices, computer systems, servers, etc. may be described in the general context of computer system-executable instructions, such as program modules, being executed by a computer system. Generally, program modules may include routines, programs, objects, components, logic, data structures, etc., that perform particular tasks or implement particular abstract data types. The computer system/server may be implemented in a distributed cloud computing environment. In a distributed cloud computing environment, tasks may be performed by remote processing devices that are linked through a communications network. In a distributed cloud computing environment, program modules may be located in both local and remote computing system storage media including memory storage devices.
Exemplary overview
The method comprises the steps of performing desensitization processing on an original image by using a preset desensitization method to be evaluated to obtain a desensitized image, then performing restoration processing on the desensitized image to obtain a first image corresponding to the desensitized image, then estimating a first probability value of the first image being a real image, and determining evaluation information of the preset desensitization method based on the first probability value, wherein the evaluation information is used for evaluating reliability of the preset desensitization method. The method of reliability verification of the desensitization method of the present disclosure is exemplarily described below with reference to fig. 1, and fig. 1 shows one scenario of the method of reliability verification of the desensitization method of the present disclosure.
In the scenario shown in fig. 1, the electronic device 100 is an execution subject of the method for verifying the reliability of the desensitization method of the present disclosure, and the electronic device 100 may be, for example, a terminal computer or a server, on which computer software or code corresponding to a preset desensitization method to be evaluated and computer software or code corresponding to an image restoration algorithm are loaded. The executing body may perform desensitization processing on the first original image 110 by using a preset desensitization method, for example, code coding processing may be performed on an area where sensitive information in the image is located to obtain a first desensitized image 120, and then, restoration processing is performed on the first desensitized image 120 by using a preset image restoration algorithm, for example, a deep neural network for image generation, to obtain a first image 130 corresponding to the first desensitized image 120. Then, the first probability value 140 that the first image 130 is a real image is estimated, for example, the first image 130 may be input into a pre-trained machine learning model or a depth network model for image recognition. Finally, the evaluation information 150 is generated based on the first probability value 140, where the evaluation information 150 may characterize the reliability of the preset desensitization method, for example, the higher the value of the first probability value 140, the closer the first image 130 is to the first original image 110, and the higher the probability that the desensitized image obtained by the preset desensitization method may be restored by the image to obtain the sensitive information, the lower the reliability of the preset desensitization method.
Exemplary method
Fig. 2 is a flow chart of one embodiment of a method of reliability verification of the desensitization method of the present disclosure. As shown in fig. 2, the method comprises the steps of:
and 200, performing desensitization treatment on the first original image by a preset desensitization method to obtain a first desensitized image.
In this example, the preset desensitization method characterizes the desensitization method to be evaluated. The first original image may represent an image which is not subjected to desensitization processing, for example, a raw image (image data in raw format) directly output by a camera sensor (camera sensor), or an RGB image obtained by preprocessing (for example, color interpolation) the raw image. The first desensitized image represents an image obtained by desensitizing an area where sensitive information in the first original image is located by a preset desensitization method.
And 210, performing image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image.
In this embodiment, the first image represents an image obtained by performing an image restoration process on the first desensitized image, and the image restoration process may include one or more image anti-desensitization processing algorithms for the purpose of reproducing sensitive information of a specific area (for example, the first target area or the second target area) in the first desensitized image.
In an alternative example, the executing subject may input the first desensitized image obtained in step 200 to a generator in a pre-trained conditional countermeasure generation network. And taking the first desensitized image as a conditional label, learning an image data distribution rule in the first desensitized image, and then simulating sensitive information in the first desensitized image according to the distribution rule through random noise to obtain the first image.
Step 220, estimating a first probability value that the first image is a real image.
In the present embodiment, the real image represents the first original image which has not been subjected to the desensitization processing. The first probability value represents the similarity degree of the first image and the first original image, and the larger the first probability value is, the more similar the first image and the first original image are, the more sensitive information restored in the first image is close to real sensitive information in the first original image.
As an example, the execution subject may estimate the first probability value of the first image as a real image by means of an image recognition model, and may specifically include the steps of first extracting features related to the image restoration process from the first image, such as the number, the position and the pixel value of pixel points having abnormal pixel values. Thereafter, a first probability value is estimated based on the extracted features.
In some optional implementations of this embodiment, the execution subject may restore the first desensitized image by using multiple image restoration algorithms, to obtain first images, then estimate probability values of each first image being a real image, and determine a mean value or a weighted average value of the multiple probability values as the first probability value.
Step 230, determining evaluation information based on the first probability value.
In this embodiment, the evaluation information is used to evaluate the reliability of the preset desensitization method, and may be in the form of a text description, a numerical value, or a data presentation of an image.
As an example, the execution subject may establish in advance a correspondence between a numerical interval of the first probability value and the evaluation level, for example, when the first probability value is [0.8,1.0], it indicates that the similarity between the first image and the first original image is extremely high, it indicates that the reliability of the preset desensitization method is extremely poor, and at this time, the evaluation level corresponding to the interval may be determined as "extremely poor". When the first probability value is 0,0.3, it indicates that the similarity between the first image and the original image is extremely low, and the reliability of the preset desensitization method is extremely high, and at this time, the evaluation level corresponding to the interval can be determined to be "excellent". Then, the executing body may determine the corresponding evaluation level according to the interval where the first probability value obtained in step 220 is located, so as to obtain the evaluation information of the preset desensitization method.
For another example, different colors may be set for different numerical intervals to color characterize the reliability of the preset desensitization method.
According to the method for verifying the reliability of the desensitization method, the desensitization processing is carried out on the first original image through the preset desensitization method to obtain a first desensitized image, then the restoration processing is carried out on the first desensitized image to obtain a restored first image, the first probability value of the first image being a real image is estimated, and then evaluation information is generated based on the first probability value and used for evaluating the reliability of the preset desensitization method. The reliability of the image data desensitization method is evaluated.
In some alternative implementations of this embodiment, the method may further include inputting the first desensitized image to a generator in a pre-trained antagonistic generation network to obtain a first image corresponding to the first desensitized image.
In the present embodiment, the restoration processing of the first desensitized image can be realized by training the generator in the countermeasure generation network to learn the image restoration processing policy.
Further, the method can estimate a first probability value by inputting the first image into a discriminator in the countermeasure generation network to obtain the confidence of the first image, and determining the first probability value that the first image is a real image based on the confidence of the first image.
In the implementation manner, the discriminators in the countermeasure generation network can learn the image recognition strategy through training, so that the probability that the first image output by the generator is a real image is judged.
In a specific example of the implementation mode, the countermeasure generation network is trained by firstly constructing a sample set, wherein the sample set comprises a first sample image marked as 0 and a second sample image marked as 1, the first sample image is an image generated by a pre-constructed initial countermeasure generation network generator, the second sample image is an image which is not subjected to desensitization, then fixing parameters of the generator, performing primary training on the discriminator, inputting the image in the sample set into the initial countermeasure generation network discriminator, taking the mark of the image as expected output, training the initial countermeasure generation network discriminator to obtain the first trained discriminator, then constructing a sample image pair, wherein the sample image pair consists of a third sample image and a sample label thereof, the sample label is an image which is not subjected to desensitization, the third sample image is an image obtained by the sample label after data desensitization, then fixing parameters of the discriminator, inputting the third image in the sample image pair into the initial countermeasure generation network generator, taking the image corresponding to the initial challenge generation network discriminator, and outputting the initial sample label of the sample image pair as expected output of the initial challenge generation network, and obtaining the initial challenge training of the sample image pair after the sample image pair is obtained.
And connecting the generator after the initial training and the discriminator in series, alternately fixing parameters of the generator and the discriminator, and alternately iterating the parameters until a training termination condition is met, so as to obtain the trained countermeasure generation network, wherein the termination condition can be, for example, the preset iteration times or the confidence coefficient value output by the discriminator. In the process of alternate iteration, parameters of the generator are adjusted through the output result of the discriminator, so that the generator can generate more real images to improve the generating capacity of the generator, and the parameters of the discriminator are adjusted based on the images output by the improved generator, so that the discriminator can accurately identify the images to improve the discriminating capacity of the discriminator. The performance of the two is alternately improved by the opponent game between the generator and the arbiter.
In the implementation manner, the image restoration capability of the generator and the identification capability of the discriminator on the image can be improved by means of cooperative training and game between the generator and the discriminator in the countermeasure generation network, so that the pertinence and the accuracy of the reliability verification method of the desensitization method are improved.
Referring next to fig. 3, as shown in fig. 3, in some alternative implementations of the present embodiment, step 200 may further include the steps of:
And 300, performing demosaicing processing on the first original image, and converting the first original image into a three-channel image.
In this implementation, the first original image may be a native image.
As an example, the execution subject may input a native image into an ISP (IMAGE SIGNAL Processing, image signal processor), and perform demosaicing Processing on the native image through a color restoration module preset in the ISP, to obtain a three-channel image (for example, may be an RGB image) corresponding to the native image.
Step 310, identifying a first target area where sensitive information in the three-channel image is located.
In this implementation manner, the sensitive information may include information of the types of privacy information, portrait information, security information, and the like, and the execution subject may input the three-channel image obtained in the step 300 into a pre-trained image recognition model, for example, a convolutional neural network model, and recognize a first target area where the sensitive information is located from the three-channel image, for example, may mark the outline of the image area where the sensitive information is located through a detection frame.
And 320, adjusting the pixel value of the first target area to obtain a first desensitized image.
As an example, the execution body may set the pixel value of the first target area to 0 (i.e., the values of the three colors of RGB are all 0) or other values, so that each pixel point in the first target area appears black, thereby implementing blanking of the sensitive information, and obtaining the image after the first desensitization.
In one example, the execution subject may further input the image marked out of the first target area into an ISP, and the adjustment of the pixel value of the first target area is implemented in a desensitization module in the ISP, resulting in a first desensitized image.
As can be seen from fig. 3, in the implementation manner shown in fig. 3, the execution subject may first convert the first original image into a three-channel image, then identify a first target area where the sensitive information is located from the three-channel image, and perform desensitization processing on the first target area, so as to obtain a first desensitized image of the three channels.
Referring next to fig. 4, as shown in fig. 4, in other alternative implementations of the present embodiment, step 200 may further employ the following procedure:
step 400, identifying a second target area where the sensitive information in the first original image is located.
In this implementation, the first raw image represents a raw image that has not been desensitized. As an example, the execution subject may input the native image into a pre-built image recognition model that characterizes a correspondence of the native image to the second target region, to recognize the second target region from the first original image.
Step 410, adjusting the pixel value of the second target area to obtain the first desensitized image.
In this implementation manner, the execution body may directly adjust the pixel value of the native image, for example, may adjust the brightness value of each pixel point in the second target area to the minimum, so as to obtain the first desensitized image, so as to implement blanking of sensitive information in the native image.
As can be seen from fig. 4, in the implementation manner shown in fig. 4, the execution subject may directly perform the identifying and desensitizing process on the first original image, so as to obtain the type of the first desensitized image as the native image.
Referring next to fig. 5, fig. 5 shows a flowchart of yet another embodiment of a method of reliability verification of the desensitization method of the present disclosure, as shown in fig. 5, the flowchart comprising the steps of:
And 500, performing desensitization treatment on the first original image by a preset desensitization method to obtain a first desensitized image.
And 510, performing image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image.
Step 520, estimating a first probability value for the first image as a real image.
In the embodiment, the steps 500 to 520 correspond to the steps 200 to 220, respectively, and are not repeated here.
And 530, respectively desensitizing at least one second original image by a preset desensitization method to obtain at least one second desensitized image.
In this embodiment, the at least one second original image is an image different from the first original image. For example, the at least one second original image may include one image or may include a plurality of different images.
And 540, performing image restoration processing on at least one second desensitized image to obtain at least one second image corresponding to the desensitized image.
Step 550, determining probability values of at least one second image corresponding to each of the second desensitized images as a real image, so as to obtain at least one second probability value.
In this embodiment, the process of desensitizing, restoring and estimating the first probability value of the first original image corresponds to the process of desensitizing, restoring and estimating the second probability value of at least one second original image, which is not described herein.
Step 560, determining the assessment information based on the first probability value and the at least one second probability value.
As an example, the executing body may first determine a mean value of the first probability value and the at least one second probability value, and then determine an evaluation level according to a numerical interval in which the mean value is located, to obtain evaluation information of the preset desensitization method.
As can be seen from fig. 5, the embodiment shown in fig. 5 embodies that evaluation information is determined based on probability values of a plurality of images obtained by a preset desensitization method, respectively, being real images, as compared with the embodiment shown in fig. 2. The overall performance of the preset desensitization method can be more accurately depicted through a plurality of images, so that the accuracy of reliability verification of the preset desensitization method can be improved.
Referring next to fig. 6, as shown in fig. 6, in some alternative implementations of the present embodiment, step 560 may further include the steps of:
Step 600, determining a first weight coefficient of the first desensitized image in determining the evaluation information.
In this implementation, the first weight coefficient characterizes a degree of importance of the first desensitized image to the evaluation result.
Step 610, determining respective second weight coefficients of the at least one second desensitized image in determining the evaluation information.
In this implementation, the second weight coefficient characterizes a degree of importance of the at least one second desensitized image to the evaluation result.
Step 620, weighting the first probability value and the at least one second probability value based on the first weight coefficient and the second weight coefficient corresponding to each other, and determining the evaluation information.
As an example, a mapping relationship between a weighted sum or weighted average value and an evaluation level may be previously constructed, and then the execution subject may determine a weighted sum or weighted average of the first probability value and at least one second probability value, and map the value and the evaluation level, determine an evaluation level of a preset desensitization method, and obtain evaluation information.
In the implementation manner, the importance degree of the first desensitized image in the evaluation result is represented by the first weight coefficient, the importance degree of the second desensitized image in the evaluation result is represented by the second weight coefficient, and the evaluation information of the preset desensitization method is determined based on the first probability value and the at least one second probability value weighting result, so that the reliability of the preset desensitization method can be evaluated more accurately.
Exemplary apparatus
Fig. 7 is a schematic structural view of an embodiment of an apparatus for reliability verification of the desensitization method of the present disclosure. The apparatus of this embodiment may be used to implement the corresponding method embodiments of the present disclosure. The apparatus shown in fig. 7 includes an image desensitizing unit 710 configured to desensitize a first original image by a preset desensitizing method to obtain a first desensitized image, an image restoration module 720 configured to perform image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image, a probability predicting unit 730 configured to estimate a first probability value that the first image is a real image, and an information generating unit 740 configured to determine evaluation information for evaluating reliability of the preset desensitizing method based on the first probability value.
In this embodiment, the image desensitizing unit 710 is further configured to input the first desensitized image into a generator in a pre-trained antagonistic generation network to obtain a first image corresponding to the first desensitized image.
As shown in fig. 8, in the present embodiment, the probability prediction unit 730 further includes a prediction module 731 configured to input the first image into a discriminator in the countermeasure generation network to obtain a confidence of the first image, and a determination module 732 configured to determine a probability value that the first image is a true image based on the confidence of the first image.
As shown in fig. 9, in the present embodiment, the image desensitizing unit 710 further includes an image converting module 711 configured to perform demosaicing processing on the first original image to convert the first original image into a three-channel image, a first identifying module 712 configured to identify a first target area where sensitive information in the three-channel image is located, and a first desensitizing module 713 configured to adjust pixel values of the first target area to obtain a first desensitized image.
As shown in fig. 10, in this embodiment, the image desensitizing unit 710 further includes a second identifying module 714 configured to identify a second target area where the sensitive information in the first original image is located, and a second desensitizing module 715 configured to adjust pixel values of the second target area to obtain the first desensitized image.
As shown in fig. 11, in the present embodiment, the information generating unit 740 further includes a third desensitizing module 741 configured to determine that at least one second original image is obtained by desensitizing at least one second original image by a preset desensitizing method, respectively, the at least one second original image being an image different from the first original image, an image restoring module 742 configured to perform an image restoring process on the at least one second desensitized image to obtain a second image corresponding to each of the at least one second desensitized image, a probability predicting module 743 configured to determine probability values of each of the at least one second desensitized image being a real image to obtain at least one second probability value, and an information generating module 744 configured to determine evaluation information based on the first probability value and the at least one second probability value.
As shown in fig. 12, in the present embodiment, the information generating module 744 further includes a first weighting sub-module 7441 configured to determine first weighting coefficients of the first desensitized image in determining the evaluation information, a second weighting sub-module 7442 configured to determine respective second weighting coefficients of the at least one second desensitized image in determining the evaluation information, and a weighting sub-module 7443 configured to weight the first probability value and the at least one second probability value based on the first weighting coefficients and the respective second weighting coefficients, to determine the evaluation information.
Exemplary electronic device
An electronic device according to an embodiment of the present disclosure is described below with reference to fig. 13. Fig. 13 shows a block diagram of an electronic device according to an embodiment of the disclosure. As shown in fig. 13, the electronic device 1300 includes one or more processors 1310 and memory 1320.
The processor 1310 may be a Central Processing Unit (CPU) or other form of processing unit having data processing and/or instruction execution capabilities, and may control other components in the electronic device 1300 to perform desired functions.
Memory 1320 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, random Access Memory (RAM) and/or cache memory (cache), among others. The nonvolatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, and the like. One or more computer program instructions may be stored on the computer readable storage medium that can be executed by the processor 11 to perform the functions of the method of reliability verification of the desensitization method of the various embodiments of the disclosure described above. Various contents such as an input signal, a signal component, a noise component, and the like may also be stored in the computer-readable storage medium.
In one example, electronic device 1300 may also include an input 1330 and an output 1340, among other components, interconnected by a bus system and/or other forms of connection mechanisms (not shown). In addition, the input device 1330 may also include, for example, a keyboard, a mouse, and the like. The output device 1340 can output various information to the outside. The output device 1340 may include, for example, a display, speakers, a printer, and a communication network and remote output devices connected thereto, etc.
Of course, only some of the components of the electronic device 1300 that are relevant to the present disclosure are shown in fig. 13 for simplicity, components such as buses, input/output interfaces, etc. are omitted. In addition, the electronic device 1300 may include any other suitable components depending on the particular application.
Exemplary computer program product and computer readable storage Medium
In addition to the methods and apparatus described above, embodiments of the present disclosure may also be a computer program product comprising computer program instructions which, when executed by a processor, cause the processor to perform steps in a method of reliability verification of a desensitization method according to various embodiments of the present disclosure described in the above "exemplary methods" section of the present description.
The computer program product may write program code for performing the operations of embodiments of the present disclosure in any combination of one or more programming languages, including an object oriented programming language such as Java, C++ or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device, partly on a remote computing device, or entirely on the remote computing device or server.
Furthermore, embodiments of the present disclosure may also be a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, cause the processor to perform steps in a training method of a language model or a method of predicting occurrence probability of words based on a language model according to various embodiments of the present disclosure described in the above "exemplary method" section of the present description.
The computer readable storage medium may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may include, for example, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of a readable storage medium may include an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
The basic principles of the present disclosure have been described above in connection with specific embodiments, but it should be noted that the advantages, benefits, effects, etc. mentioned in the present disclosure are merely examples and not limiting, and these advantages, benefits, effects, etc. are not to be considered as necessarily possessed by the various embodiments of the present disclosure. Furthermore, the specific details disclosed herein are for purposes of illustration and understanding only, and are not intended to be limiting, since the disclosure is not necessarily limited to practice with the specific details described.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different manner from other embodiments, so that the same or similar parts between the embodiments are mutually referred to. For system embodiments, the description is relatively simple as it essentially corresponds to method embodiments, and reference should be made to the description of method embodiments for relevant points.
The block diagrams of the devices, apparatuses, devices, systems referred to in this disclosure are merely illustrative examples and are not intended to require or imply that the connections, arrangements, configurations must be made in the manner shown in the block diagrams. As will be appreciated by one of skill in the art, the devices, apparatus, devices, and systems may be connected, arranged, configured in any manner. Words such as "including," "comprising," "having," and the like are words of openness and mean "including but not limited to," and are used interchangeably therewith. The terms "or" and "as used herein refer to and are used interchangeably with the term" and/or "unless the context clearly indicates otherwise. The term "such as" as used herein refers to, and is used interchangeably with, the phrase "such as, but not limited to.
The methods and apparatus of the present disclosure may be implemented in a number of ways. For example, the methods and apparatus of the present disclosure may be implemented by software, hardware, firmware, or any combination of software, hardware, firmware. The above-described sequence of steps for the method is for illustration only, and the steps of the method of the present disclosure are not limited to the sequence specifically described above unless specifically stated otherwise. Furthermore, in some embodiments, the present disclosure may also be implemented as programs recorded in a recording medium, the programs including machine-readable instructions for implementing the methods according to the present disclosure. Thus, the present disclosure also covers a recording medium storing a program for executing the method according to the present disclosure.
It is also noted that in the apparatus, devices and methods of the present disclosure, components or steps may be disassembled and/or assembled. Such decomposition and/or recombination should be considered equivalent to the present disclosure.
The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present disclosure. Various modifications to these aspects, and the like, will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the disclosure. Thus, the present disclosure is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
The foregoing description has been presented for purposes of illustration and description. Furthermore, this description is not intended to limit embodiments of the disclosure to the form disclosed herein. Although a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, changes, additions, and sub-combinations thereof.
Claims (10)
1. A method of reliability verification of a desensitization method, comprising:
Desensitizing the first original image by a preset desensitizing method to obtain a first desensitized image;
performing image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image;
The method comprises the steps of inputting a first image into a discriminant in a pre-trained countermeasure generation network to obtain the confidence coefficient of the first image, determining the first image as a first probability value of the real image based on the confidence coefficient of the first image, wherein the real image is the first original image, the first probability value represents the similarity degree of the first image and the first original image, and the first probability value is used for representing the probability that the first desensitized image is restored through the image to obtain sensitive information;
Based on the first probability value, evaluation information is determined, wherein the evaluation information is used for evaluating the reliability of the preset desensitization method.
2. The method according to claim 1, wherein performing image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image includes:
inputting the first desensitized image into a generator in the countermeasure generation network to obtain a first image corresponding to the first desensitized image.
3. The method according to claim 1, wherein the desensitizing the first original image by a preset desensitizing method to obtain a first desensitized image comprises:
demosaicing is carried out on the first original image, and the first original image is converted into a three-channel image;
identifying a first target area in which sensitive information in the three-channel image is located;
And adjusting the pixel value of the first target area to obtain the first desensitized image.
4. The method according to claim 1, wherein the desensitizing the first original image by a preset desensitizing method to obtain a first desensitized image comprises:
identifying a second target area in which the sensitive information in the first original image is located;
and adjusting the pixel value of the second target area to obtain the first desensitized image.
5. The method of one of claims 1 to 4, wherein the determining evaluation information based on the first probability value comprises:
Respectively desensitizing at least one second original image by the preset desensitization method to obtain at least one second desensitized image, wherein the at least one second original image is an image different from the first original image;
Performing image restoration processing on the at least one second desensitized image to obtain second images corresponding to the at least one desensitized image respectively;
determining probability values of the second images corresponding to the at least one second desensitized image as real images respectively to obtain at least one second probability value;
The evaluation information is determined based on the first probability value and the at least one second probability value.
6. The method of claim 5, wherein determining evaluation information based on the first probability value and the at least one second probability value comprises:
Determining a first weight coefficient of the first desensitized image in the determination of the evaluation information;
Determining respective second weight coefficients of the at least one second desensitized image in determining the evaluation information;
And determining evaluation information by weighting the first probability value and the at least one second probability value based on the first weight coefficient and the respective corresponding second weight coefficient.
7. An apparatus for reliability verification of a desensitization method, comprising:
the image desensitization unit is configured to desensitize the first original image by a preset desensitization method to obtain a first desensitized image;
The image restoration module is configured to perform image restoration processing on the first desensitized image to obtain a first image corresponding to the first desensitized image;
The probability prediction unit is configured to estimate a first probability value of the first image as a real image, and particularly configured to input the first image into a discriminator in a pre-trained countermeasure generation network to obtain the confidence of the first image; determining a first probability value of the first image as a real image based on the confidence coefficient of the first image, wherein the real image is the first original image, and the first probability value characterizes the similarity degree of the first image and the first original image;
An information generating unit configured to determine evaluation information for evaluating reliability of the preset desensitization method based on the first probability value.
8. A computer readable storage medium storing a computer program for performing the method of any one of the preceding claims 1-6.
9. An electronic device, the electronic device comprising:
A processor;
A memory for storing the processor-executable instructions;
The processor being adapted to perform the method of any of the preceding claims 1-6.
10. A computer program product comprising computer programs/instructions which, when executed by a processor, implement the method of any of the preceding claims 1-6.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111283508.3A CN113989156B (en) | 2021-11-01 | 2021-11-01 | Method, apparatus, medium, device, and program for verifying reliability of desensitization method |
| PCT/CN2022/118401 WO2023071563A1 (en) | 2021-11-01 | 2022-09-13 | Reliability verification method and apparatus for desensitization method, medium, device, and program |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111283508.3A CN113989156B (en) | 2021-11-01 | 2021-11-01 | Method, apparatus, medium, device, and program for verifying reliability of desensitization method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113989156A CN113989156A (en) | 2022-01-28 |
| CN113989156B true CN113989156B (en) | 2025-08-15 |
Family
ID=79745385
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111283508.3A Active CN113989156B (en) | 2021-11-01 | 2021-11-01 | Method, apparatus, medium, device, and program for verifying reliability of desensitization method |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN113989156B (en) |
| WO (1) | WO2023071563A1 (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113989156B (en) * | 2021-11-01 | 2025-08-15 | 北京地平线信息技术有限公司 | Method, apparatus, medium, device, and program for verifying reliability of desensitization method |
| CN117455751B (en) * | 2023-12-22 | 2024-03-26 | 新华三网络信息安全软件有限公司 | Road section image processing system and method |
| CN117892349B (en) * | 2024-01-16 | 2025-03-18 | 唐山启奥科技股份有限公司 | Image desensitization testing method, device and computer-readable storage medium |
| CN117892358B (en) * | 2024-03-18 | 2024-07-05 | 北方健康医疗大数据科技有限公司 | Verification method and system for limited data desensitization method |
| CN118861997B (en) * | 2024-09-24 | 2025-02-11 | 深圳航天信息有限公司 | Multi-source heterogeneous urban big data fusion method, device, equipment and storage medium |
| CN119783157B (en) * | 2024-11-21 | 2025-08-01 | 中国民用航空总局第二研究所 | Data processing method, device, terminal and medium based on data desensitization |
| CN119358605B (en) * | 2024-12-25 | 2025-06-03 | 蚂蚁智信(杭州)信息技术有限公司 | Method, apparatus, storage medium and electronic device for generating challenge-resistant samples |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109872282A (en) * | 2019-01-16 | 2019-06-11 | 众安信息技术服务有限公司 | A computer vision-based image desensitization method and system |
| CN110634103A (en) * | 2019-08-28 | 2019-12-31 | 天津大学 | Image demosaicing method based on generative adversarial network |
Family Cites Families (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110728626A (en) * | 2018-07-16 | 2020-01-24 | 宁波舜宇光电信息有限公司 | Image deblurring method and apparatus and training thereof |
| CN110163218A (en) * | 2019-04-10 | 2019-08-23 | 阿里巴巴集团控股有限公司 | Desensitization process method and device based on image recognition |
| CN110135193A (en) * | 2019-05-15 | 2019-08-16 | 广东工业大学 | A data desensitization method, device, equipment and computer-readable storage medium |
| CN111737750B (en) * | 2020-06-30 | 2023-12-26 | 绿盟科技集团股份有限公司 | A data processing method, device, electronic equipment and storage medium |
| CN111563275B (en) * | 2020-07-14 | 2020-10-20 | 中国人民解放军国防科技大学 | A Data Desensitization Method Based on Generative Adversarial Networks |
| CN112200132B (en) * | 2020-10-28 | 2022-06-03 | 支付宝(杭州)信息技术有限公司 | Data processing method, device and equipment based on privacy protection |
| CN112508929B (en) * | 2020-12-16 | 2024-12-17 | 奥比中光科技集团股份有限公司 | Method and device for training and generating countermeasure network |
| CN112634163B (en) * | 2020-12-29 | 2024-10-15 | 南京大学 | Image motion blur removal method based on improved cyclic generative adversarial network |
| CN113012049B (en) * | 2021-04-15 | 2022-08-02 | 山东新一代信息产业技术研究院有限公司 | Remote sensing data privacy protection method based on GAN network |
| CN113379036B (en) * | 2021-06-18 | 2023-04-07 | 西安石油大学 | Oil-gas image desensitization method based on context encoder |
| CN113379592B (en) * | 2021-06-23 | 2023-09-01 | 北京百度网讯科技有限公司 | Processing method and device for sensitive area in picture and electronic equipment |
| CN113989156B (en) * | 2021-11-01 | 2025-08-15 | 北京地平线信息技术有限公司 | Method, apparatus, medium, device, and program for verifying reliability of desensitization method |
-
2021
- 2021-11-01 CN CN202111283508.3A patent/CN113989156B/en active Active
-
2022
- 2022-09-13 WO PCT/CN2022/118401 patent/WO2023071563A1/en not_active Ceased
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109872282A (en) * | 2019-01-16 | 2019-06-11 | 众安信息技术服务有限公司 | A computer vision-based image desensitization method and system |
| CN110634103A (en) * | 2019-08-28 | 2019-12-31 | 天津大学 | Image demosaicing method based on generative adversarial network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113989156A (en) | 2022-01-28 |
| WO2023071563A1 (en) | 2023-05-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113989156B (en) | Method, apparatus, medium, device, and program for verifying reliability of desensitization method | |
| US11995155B2 (en) | Adversarial image generation method, computer device, and computer-readable storage medium | |
| CN112673381B (en) | Method and related device for identifying confrontation sample | |
| CN110851835A (en) | Image model detection method and device, electronic equipment and storage medium | |
| US20250054324A1 (en) | Video detection method and apparatus, device, storage medium, and product | |
| JP2020508522A (en) | Periodic hostile generation networks for unsupervised cross-domain image generation | |
| CN110647916B (en) | Pornographic picture identification method and device based on convolutional neural network | |
| EP4607475A1 (en) | Model determination method and related apparatus | |
| CN108399379A (en) | The method, apparatus and electronic equipment at facial age for identification | |
| CN113569611A (en) | Image processing method, image processing device, computer equipment and storage medium | |
| CN113222480A (en) | Training method and device for confrontation sample generation model | |
| CN112200380A (en) | Method and device for optimizing risk detection model | |
| CN119341803B (en) | Network security policy generation method and system based on AI large model | |
| CN114186039A (en) | Visual question answering method and device and electronic equipment | |
| CN114187483A (en) | Methods for generating adversarial samples, training methods for detectors, and related equipment | |
| CN115797735A (en) | Target detection method, device, equipment and storage medium | |
| CN112199978B (en) | Video object detection method and device, storage medium and electronic equipment | |
| CN108256633B (en) | Method for testing stability of deep neural network | |
| CN110705603B (en) | Method and system for dynamically judging similarity of user request data | |
| CN116993839B (en) | Encoding method screening method, device, electronic equipment and storage medium | |
| CN112200804A (en) | Image detection method and device, computer readable storage medium and electronic equipment | |
| CN108229320A (en) | Select frame method and device, electronic equipment, program and medium | |
| CN119865263A (en) | Method, device, equipment, storage medium and product for processing interference component | |
| CN117951554A (en) | Method and equipment for predicting default probability of user based on multidimensional data | |
| CN119006931B (en) | Small sample learning image classification method and device and electronic equipment |
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 |