An image transmission method, apparatus, device, and computer program product

Abstract

This application provides an image dissemination method, apparatus, device, and computer program product. The method includes: determining the identity information of a target user in a target image; obtaining privacy requirements for disseminating the target image from the target user based on the target user's identity information; performing privacy protection processing on the target image's area containing the target user, matching the privacy requirements; and disseminating the target image after the privacy protection processing has been performed. This application can solve the problem of exposing the privacy of others caused by unauthorized image dissemination.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to an image transmission method, apparatus, device, and computer program product. Background Technology

[0002] Thanks to the rapid development of mobile internet and the global proliferation of smartphones, social networks have entered a golden age of growth. However, with the massive daily dissemination of information, privacy and security issues on social networks have become increasingly prominent. In particular, when users post photos, videos, and other image-based data on social networks, they often intentionally or unintentionally reveal the private information of others, posing significant security risks.

[0003] Therefore, how to prevent users from disseminating images without authorization and exposing the privacy of others is an urgent problem that needs to be solved. Summary of the Invention

[0004] The purpose of this application is to provide an image dissemination method, apparatus, device, and computer program product that can solve the problem of exposing the privacy of others when users disseminate images without authorization.

[0005] To achieve the above objectives, the embodiments of this application are implemented as follows:

[0006] Firstly, an image propagation method is provided, including:

[0007] Determine the identity information of the user in the target image;

[0008] Based on the identity information of the target user, obtain the target user's opinion on privacy requirements for disseminating the target image;

[0009] Perform privacy protection processing on the region of the target user in the target image that matches the privacy requirements.

[0010] The target image after the privacy protection processing has been performed is then propagated.

[0011] Secondly, an image transmission device is provided, comprising:

[0012] The determination module is used to determine the identity information of the user in the target image.

[0013] The acquisition module is used to obtain the privacy requirements for disseminating the target image from the target user based on the target user's identity information;

[0014] An execution module is used to perform privacy protection processing on the region of the target user in the target image that matches the privacy requirements.

[0015] The propagation module is used to send an authorization request to the target user to propagate the target image if the predicted privacy requirement of the target user meets the preset requirements.

[0016] Thirdly, embodiments of this application provide an electronic device, including: a processor; and a memory configured to store computer-executable instructions, which, when executed, cause the processor to perform the method described in the first aspect.

[0017] Fourthly, a computer program product is provided, the computer program product including a computer-readable storage medium storing a computer program operable to cause a computer to perform the method described in the first aspect.

[0018] Before disseminating a target image, this application embodiment first determines the identity information of the target user in the target image. Then, based on the identity information, it actively solicits the target user's opinion on privacy requirements for disseminating the target image. After performing corresponding privacy protection processing on the target image according to the privacy requirements feedback from the target user, it disseminates the image, thereby avoiding the problem of exposing the privacy of others by disseminating images without their consent. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart illustrating the image propagation method according to an embodiment of this application.

[0021] Figure 2 This is a schematic diagram illustrating the process of training a deep model using the image propagation method according to an embodiment of this application.

[0022] Figure 3 This is a schematic diagram of the structure of the image propagation device according to an embodiment of this application.

[0023] Figure 4 This is a schematic diagram of the structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0024] As mentioned earlier, social networks, as information-sharing platforms, inevitably pose privacy and security risks. In particular, when users share image-based data such as photos and videos, they often intentionally or unintentionally reveal others' private information, posing significant security risks to them.

[0025] Statistics show that sharing group photos is a major source of privacy breaches. Group photos involve the portrait privacy of multiple individuals, some of whom are more sensitive about privacy, while others are more casual about it. Directly posting group photos on social networks can provoke resentment from the more sensitive individuals. Therefore, it is necessary to anonymize the individuals in the group photos using photo editing techniques before posting.

[0026] Obviously, retouching images for each user individually is inconvenient. Therefore, this application aims to propose an intelligent image dissemination scheme that can determine the identity information of the users in the image before dissemination and proactively obtain their privacy requirements for image dissemination. The scheme then processes the image according to these privacy requirements before dissemination, thus avoiding the unauthorized exposure of others' privacy.

[0027] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0028] One embodiment of this application proposes an image propagation method. Figure 2 This is a flowchart illustrating the image propagation method, including:

[0029] S102, Determine the identity information of the object user in the target image.

[0030] In this embodiment, the target image is an image containing at least one target user's image that is to be disseminated on a social network. The social network referred to here may be, but is not limited to, chat groups, social media, etc. Furthermore, the dissemination behavior can be one-to-one or one-to-many; this is not specifically limited herein.

[0031] Specifically, this embodiment can obtain the identity information of the target user in the target image from the disseminator of the target image, and / or determine the identity information of the target user in the target image based on facial recognition technology.

[0032] Taking obtaining the identity information of target users from the disseminator as an example, before submitting the target image, the disseminator can proactively annotate the identity information of each target user in the target image. The annotation method could be to manually set electronic tags indicating the identity information in the area of ​​each target user in the target image.

[0033] Taking the determination of the identity information of target users based on facial recognition technology as an example, a facial feature-based identity recognition model can be pre-configured. After the communicator submits the target image, the target image is input into the identity recognition model, which then identifies the corresponding identity information based on the facial features of each target user in the target image.

[0034] It should be noted that there is more than one way to verify identity information, and this article does not make any specific restrictions.

[0035] S104, Based on the identity information of the target user, obtain the target user's opinion on privacy requirements for disseminating the target image.

[0036] In this embodiment, the privacy requirements reflect whether the target user needs privacy protection or not.

[0037] To improve the efficiency of soliciting opinions on privacy requirements, with the support of big data, the historical personal records of the target users can be obtained first based on their identity information. These historical records need to contain information indicating the level of privacy importance the target users place on themselves. Then, based on these historical records, the privacy requirements of the target users can be predicted, resulting in a predicted privacy requirement. If the predicted privacy requirement meets a preset privacy requirement standard (high privacy requirement), then the privacy requirement for disseminating the target image can be solicited from the target users. Conversely, if the predicted privacy requirement does not meet the preset privacy requirement standard (low privacy requirement), then the privacy requirement can be solicited from the target users.

[0038] As an example, historical personal records may include information on dimensions such as information posting behavior, personal profile, privacy protection settings, behavioral preferences, and interactive behaviors.

[0039] Among them, information publishing behavior may include at least one of personal image publishing frequency and personal image access permissions; personal profile may include at least one of age group, gender and occupation; privacy protection settings may include at least one of privacy setting level, privacy setting frequency and password security strength; behavioral preferences may include at least one of image type distribution and privacy information filtering ratio; interactive behavior may include at least one of interaction frequency and interactive information type.

[0040] Taking information posting behavior as an example: if a user posts personal images frequently, it means that they often upload their personal selfies to social networks and do not mind their facial information being made public on social networks, thus indicating a low level of privacy awareness.

[0041] Taking personal profiles as an example: users who are younger, in highly sensitive professions, or deal with highly sensitive topics or genders tend to have a higher level of privacy awareness.

[0042] Taking privacy protection settings as an example: if a user sets a strong password and / or a high level of privacy settings (e.g., non-friends cannot see their Moments, and they reject information from strangers), then the corresponding level of privacy protection is high; conversely, the corresponding level of privacy protection is low.

[0043] Taking behavioral preferences as an example: if the distribution of image types posted by a user indicates that the proportion of them posting personal selfies is much higher than the proportion of them sending personal selfies, it means that they do not like to share personal photos on social networks, and therefore they will mind others uploading their photos to social networks, thus indicating a high level of privacy awareness.

[0044] For example, in terms of interactive behavior: if a user frequently shares photos and chats with other users, their level of privacy awareness is relatively low; conversely, if they do not share photos or chat frequently, their level of privacy awareness is relatively high.

[0045] Clearly, the aforementioned historical personal records can reflect the degree of privacy importance attached to the target users and can serve as a basis for predicting privacy requirements.

[0046] In practical applications, this embodiment can use artificial intelligence technology to predict the privacy requirements of a target user based on the target user's historical personal records.

[0047] As an example: First, features are extracted from historical personal records based on at least two feature dimensions, including information posting behavior, personal profile, privacy protection settings, behavioral preferences, and interaction behavior, to obtain feature data corresponding to at least two feature dimensions for the target user; then, the feature data corresponding to at least two feature dimensions for the target user is fused using a nonlinear transformation to obtain cross-feature data for the target user; finally, the cross-feature data of the target user is input into a deep learning model to predict the privacy requirements of the target user by the deep learning model.

[0048] The ability of the deep learning model to predict privacy requirements is trained based on the cross-feature data of the sample users and the corresponding privacy requirement labels. The privacy requirement labels are used to annotate the privacy requirements of the sample users, and the cross-feature data of the sample users is obtained by non-linearly transforming the feature data of the sample users corresponding to at least two feature dimensions. The training principle is as follows: Figure 2As shown, the cross-feature data of sample users is input into a deep learning model to predict their privacy requirements, yielding the predicted privacy requirements. Then, using the privacy requirements indicated by the privacy requirement labels as the ground truth, the error loss between the predicted privacy requirements and the privacy requirement labels is calculated. The model parameters of the deep learning model are adjusted to reduce this error loss. By iterating through this process until the model parameters converge, the predicted privacy requirements provided by the deep learning model can be made to closely match the ground truth privacy requirements indicated by the privacy requirement labels, thus achieving the ability to predict privacy requirements based on cross-feature data.

[0049] In addition to determining a user's level of privacy awareness through their historical personal records, this embodiment can also perform image semantic analysis on the target image to determine whether the user in the target image requires privacy protection based on the analysis results.

[0050] Image semantics refers to the meaning of image content, which can be expressed through natural language.

[0051] For example, if the target image is a photo taken by the user during an outdoor activity (XXX), then the image semantics could be "photo of the outdoor activity (XXX)".

[0052] For example, if the target image is a photo taken by the target user during a family meal, then the image semantics could be "family meal photo".

[0053] It should be understood that image semantics can reflect the level of privacy of image content to a certain extent. Using the example above, "outdoor activity photos" generally belong to public settings and are mostly used for external sharing; therefore, their privacy level is relatively low, and it can be determined that the target user does not need privacy protection. On the other hand, "family gathering photos" belong to private settings and are mostly used for internal sharing; therefore, their privacy level is relatively high, and it is necessary to proactively ask the target user whether they wish to protect their privacy.

[0054] As a feasible implementation, this embodiment can use a large language model to identify the image semantics of a target image. The method is to set the prompt word of the large language model to "You play the role of image content recognition and determine the image semantics expressed by the input image", and then input the target image into the large language model, which can then identify the image semantics of the target image according to the prompt word.

[0055] It should be noted that recognizing image semantics is an existing capability of large language models. In this embodiment, the image semantic recognition results given by the large language model are used to determine whether the target image is content from a non-public setting. As mentioned above, if the target image is content from a non-public setting, it indicates a high degree of privacy. Therefore, it is necessary to further obtain the privacy requirements for disseminating the target image from the target user based on the identity information of the target user in the target image.

[0056] S106, Perform privacy protection processing on the region of the target user in the target image that matches the privacy requirements.

[0057] Specifically, this embodiment can perform privacy protection processing on the region of the target user in the target image when the privacy request indicates that the target user requires privacy protection. The privacy protection processing can be, but is not limited to, blurring, occlusion, replacement, and cropping, etc., and is not specifically limited herein. Preferably, in order to minimize the impact of image retouching on the target image, an appropriate method for privacy protection processing can be selected based on the position of the target user's region in the target image. For example: if the target user's region is located at the edge of the target image, the region of the target user in the target image is cropped, that is, the size of the target image display area is modified to directly remove the target user; if the target user's region is not at the edge of the target image, cropping is no longer appropriate, and in this case, at least one of occlusion, blurring, and replacement methods can be used to process the region of the target user in the target image.

[0058] Furthermore, if the privacy request indicates that the target user does not require privacy protection, this embodiment may not modify the target user's region in the target image. It should be noted that not modifying the target user's region in the target image is also a form of privacy protection.

[0059] S108, the target image after privacy protection processing is propagated.

[0060] In summary, the method based on the embodiments first determines the identity information of the target users in the target image before disseminating the target image. Then, based on the identity information, it obtains information on the privacy awareness level of the target users to predict their privacy requirements. If the privacy requirements of the target users reach a certain standard, it is necessary to solicit their opinions on privacy requirements for disseminating the target image to understand whether they mind. For target users who mind, their privacy is protected in a specific area of ​​the target image before disseminating the target image, thereby avoiding the problem of unauthorized exposure of others' privacy. In related applications, the method of the embodiments can intelligently analyze target users with a high level of privacy awareness in the image and proactively solicit their privacy requirements for privacy protection processing, which brings great convenience compared to the disseminator performing this process themselves.

[0061] The method flow of this embodiment will be described below in conjunction with actual application scenarios.

[0062] In this application scenario, the disseminator posts the group photo on social media or a chat group. The method in this embodiment is executed by the social media or chat group application, and the corresponding steps include:

[0063] Step 1: Obtain the group photo submitted by the disseminator and ready for publication.

[0064] In this step, the disseminator can select the group photo to be published from their personal terminal's photo album using social media or chat group applications.

[0065] Step two: Determine the identity information of each user in the group photo.

[0066] As mentioned earlier, this embodiment can determine the identity information of each user in the group photo through facial recognition or by the disseminator's own annotation, which will not be elaborated here.

[0067] Specifically, the disseminator can use the application to actively mark the users in the group photo who need to obtain authorization. The application only needs to confirm the identity information of the marked users, and ignores the unmarked users in the group photo, and will not ask for authorization from the unmarked users again.

[0068] In addition, if there are users whose identities cannot be identified, the application can choose to refuse to post the group photo on social media or chat groups, which will not be elaborated on here.

[0069] Step 3: Based on a deep learning model, predict the privacy requirements of the users in the group photo.

[0070] It should be understood that different target users have different privacy requirements, and this application scenario can analyze the historical personal records left by these target users in social media or chat groups.

[0071] Correspondingly, this step extracts features from each target user's historical personal records using at least two feature dimensions, including information publishing behavior, personal profile, privacy protection settings, behavioral preferences, and interactive behavior, to obtain feature data for each target user corresponding to at least two feature dimensions.

[0072] As an example, the description of the feature data can be found in the table below:

[0073]

[0074] Subsequently, nonlinear transformation and fusion are performed on the feature data corresponding to at least two feature dimensions for each target user to obtain the cross-feature data of each target user. The cross-feature data of each target user is then input into a deep learning model to predict the privacy requirements of each target user.

[0075] In this application scenario, the privacy requirements predicted by the deep learning model are represented by a score; a higher score indicates a higher privacy requirement, and a lower score indicates a lower privacy requirement. Correspondingly, the training process of the deep learning model is as follows:

[0076] 1) Prepare the training set for the deep learning model. The training set includes the historical personal records of the sample users and their corresponding ratings. The ratings of the sample users are the privacy requirement tags mentioned in the text.

[0077] 2) Based on at least two feature dimensions among information publishing behavior, personal profile, privacy protection settings, behavioral preferences and interactive behavior, feature extraction is performed on the historical personal records of the sample users to obtain feature data corresponding to at least two feature dimensions of the sample users.

[0078] 3) Perform nonlinear transformation and fusion on the feature data of the sample object users corresponding to at least two feature dimensions to obtain the cross feature data of the sample object users.

[0079] 4) Input the cross-feature data of the sample users into the deep learning model and try to predict the ratings of the sample users.

[0080] 5) Calculate the error loss between the score predicted by the deep learning model and the score labeled with the privacy requirement tag using the loss function.

[0081] 6) Adjust the model parameters of the deep learning model through backpropagation with the aim of reducing error loss.

[0082] In this model, the values ​​of the privacy requirement label and the privacy requirement prediction result are represented by one-hot vectors. One-hot encoding is used to convert discrete classification labels into binary vectors. In one-hot encoding, each categorical variable is assigned a unique binary bit, which is used to represent the value of the variable. If the variable's value is 1, the corresponding binary bit is 1; if the value is 0, the corresponding binary bit is 0. In one-hot encoding, each variable is encoded only once, hence the term "one-bit validity".

[0083] For example, suppose there is a categorical variable containing three categories: A, B, and C. Using One-Hot encoding, a binary bit can be assigned to each of the three categories. If a sample belongs to category A, the binary bit representing A in its One-Hot vector is 1, while the binary bits representing B and C are both 0; if a sample belongs to category B, the binary bit representing B in its One-Hot vector is 1, while the binary bits representing A and C are both 0; and so on.

[0084] As an example:

[0085] Taking a 5-point privacy requirement score as an example, during the training process, one-hot encoding can be used to encode the feature data and privacy requirement labels for each feature dimension.

[0086] In this one-hot vector, each value corresponds to a binary bit. For example, a privacy requirement tag with a score of 3 is encoded as [0,0,1,0,0] using one-hot encoding; and a privacy requirement tag with a score of 4 is encoded as [0,0,0,4,0] using one-hot encoding. The encoding process is represented as follows:

[0087]

[0088] in, This is the encoding result of the privacy requirement label. It is the input privacy requirement score integer value, and one_hot is the encoding function.

[0089] Correspondingly, the prediction process of a deep learning model can be represented as:

[0090]

[0091] in, 'Indicates the output during the training phase of a deep learning model; , and This represents the input, weight matrix, and bias of a deep learning model; Indicates the learning rate; Represents the loss function; Indicates the activation function; .

[0092] Since the scores output by deep learning models are expressed in one-hot encoding, they need to be converted into integer scores, i.e.:

[0093]

[0094] in, Indicates the first The one-hot encoded output corresponding to each training sample; This corresponds to an integer privacy sensitivity score. This is the inverse process of the one-hot encoding function.

[0095] The corresponding loss function uses the mean squared error function:

[0096]

[0097] in, Indicates the number of training samples; Indicates the first The score given by the privacy requirement label for each training sample (i.e., the value of the privacy requirement label). Indicates the deep learning model given to the first The score predicted by each training sample (i.e., the value of the prediction result according to privacy requirements). This indicates the loss due to error.

[0098] Step four: Solicit opinions from users whose scores reach the preset threshold regarding the privacy requirements for the group photo release.

[0099] Taking the 5-point privacy requirement rating as an example, in this step, the application can solicit opinions on the privacy requirements of the photos posted by users who have reached a score of 4.

[0100] Since the target user is also a user of the application, in this step, the application can send the group photo and a text message asking whether privacy protection is required to the target user's account through the system information window, and receive feedback from the target user on privacy requirements.

[0101] Step 5: For users who indicate that they require privacy protection, select any method such as blurring, obscuring, replacing, or cropping to process their area in the group photo for privacy protection.

[0102] For example, if a user whose privacy needs to be protected is located at the edge of a group photo, they will be cropped out of the photo; if a user whose privacy needs to be protected is located in the middle of a group photo, they will be obscured or blurred.

[0103] In summary, in this application scenario, when a disseminator initiates the dissemination of group photos on social media or chat groups, the application identifies the identity information of each user in the photo and searches for relevant historical personal records to predict privacy requirements. For users with high privacy requirements, the application proactively solicits their privacy requests before posting the photo, and then performs privacy protection processing on their image areas in the photo according to their requests before the photo is published on social media or chat groups as instructed by the disseminator. Throughout the entire process, identity recognition, privacy requirement prediction, privacy request solicitation, and privacy protection processing can all be performed by the application. Disseminators cannot bypass the application to arbitrarily publish group photos, thus effectively curbing the problem of disseminators intentionally or unintentionally exposing others' privacy on social media or chat groups.

[0104] Corresponding to Figure 1 The method shown in this application is another embodiment of an image propagation apparatus. Figure 3 This is a schematic diagram of the image transmission device 300, including:

[0105] The determination module 310 is used to determine the identity information of the object user in the target image.

[0106] The acquisition module 320 is used to obtain the privacy requirements for disseminating the target image from the target user based on the target user's identity information.

[0107] The execution module 330 is used to perform privacy protection processing on the region of the object user in the target image that matches the privacy requirements.

[0108] The propagation module 340 is used to send an authorization request to the target user to propagate the target image when the privacy requirement prediction result of the target user meets the preset requirements.

[0109] Before disseminating a target image, the device in this embodiment first determines the identity information of the target user in the target image. Then, based on the identity information, it actively solicits the target user's privacy requirements for disseminating the target image. After performing corresponding privacy protection processing on the target image according to the privacy requirements provided by the target user, it disseminates the image, thereby avoiding the problem of exposing the privacy of others by disseminating images without their consent.

[0110] Optionally, the acquisition module 320 obtains the privacy requirement opinion for disseminating the target image from the target user based on the target user's identity information, including: obtaining the target user's historical personal records based on the identity information; the historical personal records contain information showing the target user's level of privacy importance; predicting the target user's privacy requirements based on the historical personal records to obtain the target user's privacy requirement prediction result; and obtaining the privacy requirement opinion for disseminating the target image from the target user when the target user's privacy requirement prediction result meets the preset privacy requirement standard.

[0111] Optionally, the acquisition module 320 predicts the privacy requirements of the target user based on the historical personal records, including: extracting features from the historical personal records based on at least two feature dimensions among information publishing behavior, personal profile, privacy protection settings, behavioral preferences, and interaction behavior to obtain feature data of the target user corresponding to the at least two feature dimensions; performing nonlinear transformation and fusion on the feature data of the target user corresponding to the at least two feature dimensions to obtain cross-feature data of the target user; and inputting the cross-feature data of the target user into a deep learning model to predict the privacy requirements of the target user.

[0112] Optionally, the image propagation device in this embodiment includes:

[0113] The training module is used to train the deep learning model based on the cross-feature data of the sample users and the corresponding privacy requirement labels before inputting the cross-feature data of the target users into the deep learning model to predict the privacy requirements of the target users; wherein, the cross-feature data of the sample users is obtained by nonlinearly transforming the feature data of the sample users corresponding to at least two feature dimensions, and the privacy requirement labels are used to label the privacy requirements of the sample users.

[0114] Optionally, the training module trains the deep learning model based on the cross-feature data of the sample users and the corresponding privacy requirement labels, including: inputting the cross-feature data of the sample users into the deep learning model to predict the privacy requirements of the sample users and obtain the privacy requirement prediction result of the sample users; calculating the error loss between the privacy requirement prediction result of the sample users and the privacy requirement label, and adjusting the model parameters of the deep learning model with the aim of reducing the error loss.

[0115] Optionally, the formula for calculating the error loss is: ;

[0116] in, Indicates error loss, Indicates the number of training samples. This represents the value of the privacy requirement label for the m-th sample user. This represents the value of the predicted privacy requirements for the m-th sample user.

[0117] Optionally, the training module adjusts the model parameters of the deep learning model to reduce the error loss, including: based on the formula The model parameters of the deep learning model are adjusted.

[0118] in, , This represents the input to the deep learning model. This represents the output of the deep learning model. This represents the model parameters of the deep learning model. Indicates deviation, Indicates the learning rate. Represents the loss function. This represents the activation function.

[0119] Optionally, the values ​​of the privacy requirement label and the privacy requirement prediction result are represented by a one-hot vector, where each value of the one-hot vector corresponds to a binary bit.

[0120] Optionally, the acquisition module 320 obtains the privacy requirements for disseminating the target image from the target user based on the target user's identity information, including: performing image semantic recognition on the target image based on a large language model to obtain the image semantic recognition result corresponding to the target image; determining whether the target image is content from a non-public setting based on the image semantic recognition result; if the target image is content from a non-public setting, obtaining the privacy requirements for disseminating the target image from the target user based on the target user's identity information.

[0121] Optionally, the execution module 330 performs privacy protection processing on the region of the target user in the target image that matches the privacy request opinion, including: when the privacy request opinion indicates that the target user needs privacy protection, performing privacy protection processing on the region of the target user in the target image by blurring, occluding, replacing, and cropping at least one of these methods.

[0122] Optionally, the information publishing behavior includes at least one of: personal image publishing frequency and personal image access permissions;

[0123] The personal profile includes at least one of the following: age group, gender, and occupation;

[0124] The privacy protection settings include at least one of the following: privacy setting level, privacy setting frequency, and password security strength;

[0125] The behavioral preferences include at least one of the distribution of image types published and the proportion of privacy information filtering.

[0126] The interactive behavior includes at least one of the following: interaction frequency and interactive information type.

[0127] It should be noted that the image propagation in this embodiment can be used as... Figure 1 The execution body of the method shown is therefore able to achieve... Figure 1 The steps and functions of the method shown will not be repeated here.

[0128] Figure 4 This is a schematic diagram of the structure of an electronic device provided in one embodiment of this application. Please refer to it. Figure 4 At the hardware level, the electronic device includes a processor, and optionally also includes an internal bus, a network interface, and memory. The memory may include main memory, such as high-speed random-access memory (RAM), or non-volatile memory, such as at least one disk drive. Of course, the electronic device may also include other hardware required for other business operations.

[0129] The processor, network interface, and memory can be interconnected via an internal bus, which can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 The symbol is represented by a single double-headed arrow, but this does not mean that there is only one bus or one type of bus.

[0130] Memory is used to store programs. Specifically, programs may include program code, which includes computer operation instructions. Memory may include main memory and non-volatile memory, and provides instructions and data to the processor.

[0131] The processor reads the corresponding computer program from non-volatile memory into main memory and then executes it, forming the aforementioned image propagation device at the logical level. The processor executes the program stored in memory and specifically performs the following operations:

[0132] Determine the identity information of the user in the target image.

[0133] Based on the identity information of the target user, obtain the target user's opinion on privacy requirements for disseminating the target image.

[0134] Perform privacy protection processing on the region of the target user in the target image that matches the privacy requirements.

[0135] The target image after the privacy protection processing has been performed is then propagated.

[0136] The above is as stated in this application. Figure 1 The methods disclosed in the illustrated embodiments can be applied to or implemented by a processor. The processor may be an integrated circuit chip with signal processing capabilities. During implementation, each step of the above methods can be completed by integrated logic circuits in the processor's hardware or by instructions in software form. The processor can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. It can implement or execute the methods, steps, and logic block diagrams disclosed in one or more embodiments of this application. The general-purpose processor can be a microprocessor or any conventional processor. The steps of the methods disclosed in one or more embodiments of this application can be directly embodied in the execution of a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor. The software module can reside in a mature storage medium in the field, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method.

[0137] Of course, in addition to software implementation, the electronic device of this application does not exclude other implementation methods, such as logic devices or a combination of hardware and software, etc. In other words, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0138] This application also proposes a computer program product, which includes a computer-readable storage medium storing a computer program operable to cause a computer to perform the following operations:

[0139] Determine the identity information of the user in the target image.

[0140] Based on the identity information of the target user, obtain the target user's opinion on privacy requirements for disseminating the target image.

[0141] Perform privacy protection processing on the region of the target user in the target image that matches the privacy requirements.

[0142] The target image after the privacy protection processing has been performed is then propagated.

[0143] In summary, the above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this application should be included within the scope of protection of one or more embodiments of this application.

[0144] The systems, devices, and modules described in the above embodiments can be implemented by computer chips or physical entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, laptop computer, cellular phone, camera phone, smartphone, personal digital assistant, media player, navigation device, email device, game console, tablet computer, wearable device, or any combination of these devices.

[0145] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0146] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0147] The various embodiments in this application are described in a progressive manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.

Claims

1. An image propagation method, characterized in that, include: Determine the identity information of the user in the target image; Based on the identity information of the target user, obtaining the target user's privacy requirements for disseminating the target image includes: obtaining the target user's historical personal records based on the identity information; wherein, the historical personal records contain information showing the target user's level of privacy awareness; predicting the target user's privacy requirements based on the historical personal records to obtain the target user's privacy requirement prediction result; and obtaining the target user's privacy requirement requirements for disseminating the target image if the target user's privacy requirement prediction result meets a preset privacy requirement standard. Perform privacy protection processing on the region of the target user in the target image that matches the privacy requirements. The target image after the privacy protection processing has been performed is then propagated.

2. The method according to claim 1, characterized in that, Predicting the privacy requirements of the target users based on the aforementioned historical personal records includes: Based on at least two feature dimensions among information publishing behavior, personal profile, privacy protection settings, behavioral preferences, and interactive behavior, feature extraction is performed on the historical personal records to obtain feature data of the target user corresponding to the at least two feature dimensions. Nonlinear transformation and fusion are performed on the feature data of the target user corresponding to at least two feature dimensions to obtain the cross feature data of the target user; The cross-feature data of the target users are input into a deep learning model to predict the privacy requirements of the target users.

3. The method according to claim 2, characterized in that, Before inputting the cross-feature data of the target users into a deep learning model to predict the privacy requirements of the target users, the method further includes: The deep learning model is trained based on the cross-feature data of the sample users and the corresponding privacy requirement labels; wherein, the cross-feature data of the sample users is obtained by nonlinearly transforming the feature data of the sample users corresponding to at least two feature dimensions, and the privacy requirement labels are used to label the privacy requirements of the sample users.

4. The method according to claim 3, characterized in that, The deep learning model is trained based on the cross-feature data of the sample users and the corresponding privacy requirement labels, including: The cross-feature data of the sample users are input into the deep learning model to predict the privacy requirements of the sample users, and the prediction result of the privacy requirements of the sample users is obtained. Calculate the error loss between the predicted privacy requirements of the sample users and the privacy requirement labels, and adjust the model parameters of the deep learning model to reduce the error loss.

5. The method according to claim 4, characterized in that, The formula for calculating the error loss is: ; in, Indicates error loss, Indicates the number of training samples. This represents the value of the privacy requirement label for the m-th sample user. This represents the value of the predicted privacy requirements for the m-th sample user.

6. The method according to claim 4, characterized in that, Adjusting the model parameters of the deep learning model to reduce the error loss includes: Based on formula The model parameters of the deep learning model are adjusted. in, , This represents the input to the deep learning model. This represents the output of the deep learning model. This represents the model parameters of the deep learning model. Indicates deviation, Indicates the learning rate. Represents the loss function. This represents the activation function.

7. The method according to claim 6, characterized in that, The values ​​of the privacy requirement label and the privacy requirement prediction result are represented by a one-hot vector, and each value of the one-hot vector corresponds to a binary bit.

8. The method according to any one of claims 1 to 7, characterized in that, Performing privacy protection processing on the region of the target user in the target image that matches the privacy requirements, including: When the privacy request indicates that the target user requires privacy protection, at least one of the following is performed on the target image: blurring, occluding, replacing, and cropping the region of the target user.

9. The method according to claim 8, characterized in that, Blurring, occlusion, replacement, and cropping the region of the object user in the target image at least one of the following: If the region of the target user in the target image is located at the edge of the target image, then the region of the target user in the target image is cropped; otherwise, at least one of occlusion, blurring, and replacement is performed on the region of the target user in the target image.

10. The method according to any one of claims 2 to 7, characterized in that, The information publishing behavior includes at least one of: personal image publishing frequency and personal image access permissions; The personal profile includes at least one of the following: age group, gender, and occupation; The privacy protection settings include at least one of the following: privacy setting level, privacy setting frequency, and password security strength; The behavioral preferences include at least one of the distribution of image types published and the proportion of privacy information filtering. The interactive behavior includes at least one of the following: interaction frequency and interactive information type.

11. The method according to any one of claims 1 to 7, characterized in that, Determine the identity information of the user in the target image, including: Obtain the identity information of the target user in the target image from the disseminator of the target image; And / or, using facial recognition technology to determine the identity information of the user in the target image.

12. The method according to any one of claims 1 to 7, characterized in that, Propagating the target image after the privacy protection processing has been performed includes: The target image, after undergoing the privacy protection process, is published to the designated chat group.

13. An image transmission device, characterized in that, include: The determination module is used to determine the identity information of the user in the target image. The acquisition module is used to obtain the privacy requirements of the target user for disseminating the target image based on the target user's identity information, including: obtaining the target user's historical personal records based on the identity information; wherein, the historical personal records contain information showing the target user's level of privacy awareness; predicting the target user's privacy requirements based on the historical personal records to obtain the target user's privacy requirement prediction result; and obtaining the privacy requirements of the target user for disseminating the target image if the target user's privacy requirement prediction result meets a preset privacy requirement standard. An execution module is used to perform privacy protection processing on the region of the target user in the target image that matches the privacy requirements. The propagation module is used to send an authorization request to the target user to propagate the target image if the predicted privacy requirement of the target user meets the preset requirements.

14. An electronic device, comprising: processor; And a memory arranged to store computer-executable instructions, which, when executed, cause the processor to perform the method as described in any one of claims 1 to 12.

15. A computer program product comprising a computer-readable storage medium storing a computer program operable to cause a computer to perform the method as claimed in any one of claims 1 to 12.

Images