CN113435922B - Advertisement data pushing method, device, equipment and storage medium - Google Patents
Advertisement data pushing method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN113435922B CN113435922B CN202110651316.7A CN202110651316A CN113435922B CN 113435922 B CN113435922 B CN 113435922B CN 202110651316 A CN202110651316 A CN 202110651316A CN 113435922 B CN113435922 B CN 113435922B
- Authority
- CN
- China
- Prior art keywords
- user
- target user
- target
- category
- geographic position
- 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
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000001815 facial effect Effects 0.000 claims abstract description 67
- 239000013598 vector Substances 0.000 claims description 42
- 238000012549 training Methods 0.000 claims description 32
- 238000006243 chemical reaction Methods 0.000 claims description 15
- 238000002372 labelling Methods 0.000 claims description 6
- 238000012216 screening Methods 0.000 claims description 5
- 238000012360 testing method Methods 0.000 description 10
- 238000012795 verification Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 9
- 230000008569 process Effects 0.000 description 9
- 230000000694 effects Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 6
- 238000003064 k means clustering Methods 0.000 description 5
- 230000009471 action Effects 0.000 description 4
- 238000004138 cluster model Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 3
- 239000003086 colorant Substances 0.000 description 2
- 230000008921 facial expression Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000007423 decrease Effects 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 235000013410 fast food Nutrition 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 238000000844 transformation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
Landscapes
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Image Processing (AREA)
Abstract
An embodiment of the present specification provides an advertisement data pushing method, apparatus, device and storage medium, where the method includes: acquiring a face image of a target user and geographic position information of the target user; determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user; predicting age information of the target user according to the facial image of the target user through the target prediction model; and pushing advertisement data to the target user according to the predicted age information of the target user. According to the embodiment, the age information of the user can be accurately predicted, and the accuracy of advertisement data pushing is improved.
Description
Technical Field
The present document relates to the field of computer technologies, and in particular, to a method, an apparatus, a device, and a storage medium for pushing advertisement data.
Background
At present, when advertisement data is pushed to a user of a mobile terminal, it is common practice to push matched advertisement data to the user according to information of the user, such as gender, age, occupation, consumption level and the like. When the age of the user is not obtained, the prior art predicts the age of the user mainly through a model.
Prior art in predicting the age of a user by means of a model, a model common to all users is usually trained for prediction. Because different users in the same age group have incomplete facial expression, the ages of all users are predicted by a universal model, and the problem of inaccurate prediction is unavoidable, so that the accuracy of advertisement data pushing is reduced.
Disclosure of Invention
An object of one embodiment of the present disclosure is to provide a method, an apparatus, a device, and a storage medium for pushing advertisement data, so as to accurately predict age information of a user and improve accuracy of advertisement data pushing.
To achieve the above technical object, an embodiment of the present specification is implemented as follows:
In a first aspect, an embodiment of the present disclosure provides an advertisement data pushing method, including:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
and pushing advertisement data to the target user according to the predicted age information of the target user.
In a second aspect, another embodiment of the present disclosure provides an advertisement data pushing apparatus, including:
The data acquisition module is used for acquiring the face image of the target user and the geographic position information of the target user;
the model selection module is used for determining a target prediction model matched with the target user from all pre-trained age information prediction models according to the face image of the target user and the geographic position information of the target user;
the information prediction module is used for predicting age information of the target user according to the facial image of the target user through the target prediction model;
and the advertisement pushing module is used for pushing advertisement data to the target user according to the predicted age information of the target user.
In a third aspect, another embodiment of the present specification provides an advertisement data pushing apparatus, including:
A processor; and
A memory arranged to store computer executable instructions that, when executed, cause the processor to:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
and pushing advertisement data to the target user according to the predicted age information of the target user.
In a fourth aspect, yet another embodiment of the present description provides a storage medium storing computer-executable instructions that, when executed by a processor, perform the method of:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
and pushing advertisement data to the target user according to the predicted age information of the target user.
According to the advertisement data pushing method, device, equipment and storage medium provided by one or more embodiments of the present disclosure, according to the facial image of the target user and the geographic position information of the target user, a matched target prediction model is selected for the target user in each pre-trained age information prediction model to perform age prediction, and according to the age information of the target user obtained by prediction, advertisement data is pushed to the target user.
Drawings
In order to more clearly illustrate the technical solution in one or more embodiments of the present description, the following description will briefly introduce the drawings that are needed in the embodiments or the description of the prior art, it is obvious that the drawings in the following description are only some embodiments described in the present description, and other drawings may be obtained according to these drawings without inventive effort to a person of ordinary skill in the art.
Fig. 1 is a flowchart of an advertisement data pushing method according to an embodiment of the present disclosure;
FIG. 2 is a schematic diagram of a training process of an age information prediction model according to an embodiment of the present disclosure;
fig. 3 is a schematic diagram of module components of an advertisement data pushing device according to an embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of an advertisement data pushing device according to an embodiment of the present disclosure.
Detailed Description
In order to enable a person skilled in the art to better understand the technical solutions in one or more embodiments of the present specification, the technical solutions in one or more embodiments of the present specification will be clearly and completely described below with reference to the drawings in one or more embodiments of the present specification, and it is obvious that the described embodiments are only some embodiments of the present specification, not all embodiments. All other embodiments, which can be made by one or more embodiments of the present disclosure without inventive faculty, are intended to be within the scope of the present disclosure.
Fig. 1 is a flow chart of an advertisement data pushing method according to an embodiment of the present disclosure, where the advertisement pushing method can be applied to a server side and executed by the server, and as shown in fig. 1, the flow chart includes:
Step S102, obtaining a face image of a target user and geographic position information of the target user;
step S104, determining a target prediction model matched with the target user from all pre-trained age information prediction models according to the face image of the target user and the geographic position information of the target user;
step S106, predicting age information of the target user according to the facial image of the target user through the target prediction model;
Step S108, advertisement data are pushed to the target user according to the predicted age information of the target user.
In this embodiment, according to the facial image of the target user and the geographic location information of the target user, a matched target prediction model is selected for the target user in each pre-trained age information prediction model to perform age prediction, and advertisement data is pushed to the target user according to the age information of the target user obtained by prediction.
In step S102, a face image of the target user and geographic position information of the target user are acquired. The target user may be a user who is to push an advertisement, or may be a user who is not pushing an advertisement but who needs to predict age. The face image of the target user may be transmitted to the server by the user terminal of the target user, for example, the target user uses a mobile phone to self-photograph, and the self-photographing image may be transmitted to the server by the mobile phone. The geographic position information of the target user can be longitude and latitude information obtained by positioning a user terminal of the target user, such as longitude and latitude information obtained by positioning a mobile phone through a built-in GPS device, and the longitude and latitude information can represent the current position of the target user. The geographic location information of the target user may also be sent to the server in the native region of the target user, e.g., the target user fills out his native region.
In this embodiment, a plurality of age information prediction models are trained in advance, each pre-trained age information prediction model corresponds to a predetermined user category, and each user category is obtained based on skin color characteristics of a user and geographic position information of the user. Each user category is a predetermined user category. Because the types of the users in different areas are possibly different and the skin color characteristics of the different types of the users are different, the users can be divided based on the geographic position information of the users and the skin color characteristics of the users to obtain various user categories, and an age information prediction model is independently built for each user category, so that the age is predicted by the users in different areas and different types of the users in different areas by using the corresponding age information prediction models, the differentiation of the users in the same age range in different areas in the facial expression is fully considered, the accuracy of age prediction is improved, and the accuracy of advertisement data pushing is further improved.
Specifically, for example, in east asia, south asia, saharan africa and europe, the corresponding ethnicity is yellow ethnicity, brown ethnicity, black ethnicity and white ethnicity in turn, and the skin color characteristics of these ethnicities are different, so that the user may be classified into four user categories of east asia yellow ethnicity, south asia brown ethnicity, africa black ethnicity and european white ethnicity according to the distribution region of the user and the skin color characteristics of the user, and a corresponding age prediction model may be separately established for each user category.
Under certain special conditions, the people in a region have multiple kinds, such as white people, yellow people and brown people in a country A, and when users are classified, the people can be classified into the white people in the country A, the yellow people in the country A and the brown people in the country A, and corresponding age prediction models are independently built for each user category, so that fine granularity division of the region and the people is realized, and fine granularity modeling of the models is realized.
Based on this, in the above step S104, a target prediction model matching the target user is determined from the face image of the target user and the geographical position information of the target user, among the respective pre-trained age information prediction models, specifically:
(a1) Determining a target category matched with the target user in each user category according to the face image of the target user and the geographic position information of the target user;
(a2) And determining a model corresponding to the target category in each pre-trained age information prediction model according to the one-to-one correspondence between each user category and each pre-trained age information prediction model, and taking the model as a target prediction model matched with the target user.
In this embodiment, each user category is a predetermined user category, for example, the predetermined user category includes a country a white race, a country a yellow race, a country a brown race, an east asian yellow race, a south asian brown race, an african black race, and a european white race. In the act (a 1), a target category matching the target user is determined from the respective user categories according to the face image of the target user and the geographic position information of the target user, for example, it is determined that the target category corresponding to the target user is european white race. In the action (a 2), searching a model corresponding to the target category in each pre-trained age information prediction model, and taking the searched model corresponding to the target category as a target prediction model corresponding to the target user. For example, an age information prediction model corresponding to a european white race is used as a target prediction model corresponding to a target user.
In terms of user classification, the clustering algorithm and the clustering model have the advantages of high classification accuracy and easy realization, so in this embodiment, the classification of each user category can be realized through the clustering algorithm, based on this, in the above action (a 1), according to the face image of the target user and the geographic position information of the target user, the target category matched with the target user is determined in each user category, specifically:
(a11) Converting the face image of the target user and the geographic position information of the target user into feature vectors respectively, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
(a12) And determining a target category matched with the target user from all user categories according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
In this embodiment, a clustering model is trained in advance, and the clustering model may be implemented based on a k-means algorithm, where the clustering model may divide users into preset user categories. Therefore, when determining the target category matching with the target user from the face image of the target user and the geographic position information of the target user, the face image of the target user may be converted into a feature vector acceptable by the cluster model, where the feature vector may represent the skin color feature of the target user, and the geographic position information of the target user may be converted into a feature vector acceptable by the cluster model, and then the converted feature vector may be input into the trained cluster model. After the clustering model receives the input feature vectors, the input feature vectors can be divided into preset user categories, so that user classification is achieved. For example, after each user category is represented by a sequence number, the clustering model receives the input feature vector, and then can output the sequence number of the user category, where the sequence number represents the user category to which the target user is divided.
From the above description, each user category is divided by way of clustering, and in one embodiment, each user category is determined by:
(b1) Acquiring facial images of a plurality of sample users, and acquiring geographic position information of the sample users;
(b2) And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users to obtain each user category.
In the action (b 1), face images of a plurality of sample users are acquired, and geographical position information of each sample user is acquired. The geographical position information of the sample user can be a native place through area uploaded by the sample user, or can be longitude and latitude information obtained by positioning a user terminal of the sample user through a positioning element. In one example, facial images of sample users of different ethnicities in a large number of different regions are acquired from a network by means of crawlers, and the native regions of the sample users are determined as geographic location information of the sample users.
In the action (b 2), a k-means clustering algorithm may be adopted, and the sample users are clustered according to skin color features corresponding to facial images of the sample users and geographic position information of the sample users, so as to obtain each user category. In one embodiment, according to skin color features corresponding to face images of sample users and geographic position information of the sample users, clustering the sample users to obtain user categories, specifically: and respectively converting the facial image of the sample user and the geographic position information of the sample user into feature vectors, inputting the feature vectors obtained by conversion into a clustering model based on a preset clustering algorithm, and clustering the sample user according to skin color features corresponding to the facial image of the sample user and the geographic position information of the sample user through the clustering model to obtain each user category.
Specifically, the face image of the sample user is converted into the feature vector, and the feature vector corresponding to the face image of the sample user can reflect the skin color feature of the sample user because the face image of the sample user has the skin color of the sample user. The geographical location information of the sample user is also converted into feature vectors. And inputting the two feature vectors obtained through conversion into a clustering model based on a preset clustering algorithm, such as a clustering model based on a k-means clustering algorithm, and clustering sample users according to the input feature vectors through the clustering model to obtain each user category.
In a specific embodiment, the clustering model based on the k-means clustering algorithm also needs to be trained in advance, and the training process mainly includes the step of specifying the number of user categories obtained by classification. Specifically, after a large number of face images of sample users in different areas and different ethnicities are obtained, feature vectors of the face images and feature vectors corresponding to geographic position information of the sample users are input into a clustering model, and the number of user categories obtained by clustering the clustering model is drawn. Since it is not clear at first to divide a large number of sample users into several kinds of suitable, the number n of user categories can be firstly drawn by experience, and whether the classification condition based on the number accords with the real condition is observed manually, for example, after the sample users are initially divided into n kinds according to the skin color characteristics and the geographic position information of the sample users, whether the users in each kind belong to the users with the similar skin colors, and further the number n of the user categories is debugged according to the result of manual observation until each kind of users divided into after the debugging belong to the users with the similar skin colors, so that the effect of manual satisfaction is achieved, and then the clustering model training based on the k-means clustering algorithm is completed. And classifying the sample users based on the number n of the debugged user categories, and obtaining each user category.
Because the clustering model training of the k-means clustering algorithm is completed and the classification quantity is fixed, after the face image of the target user and the geographic position information of the target user are acquired, the face image of the target user and the geographic position information of the target user are respectively converted into feature vectors, the feature vectors obtained through conversion are input into the clustering model, and the clustering model can output the sequence number of the user category, wherein the sequence number not only represents the target category matched with the target user in each pre-divided user category.
As mentioned previously, each user category is partitioned based on the user's skin tone characteristics and the user's geographic location information. Each pre-trained age information prediction model corresponds to a predetermined user category. In this embodiment, the age information prediction model may be a CNN (Convolutional Neural Networks, convolutional neural network) model, and fig. 2 is a schematic diagram of a training flow of the age information prediction model provided in an embodiment of the present disclosure, as shown in fig. 2, where the age information prediction model is obtained by training in the following manner:
Step S202, obtaining face images of sample users in each user category, and obtaining age information of the sample users in each user category;
Step S204, labeling face images of sample users in each user category according to age information of the sample users in each user category;
Step S206, training an age information prediction model corresponding to each user category by using the labeled face images of the sample users in each user category.
A large number of sample users have been previously classified into respective user categories at the time of user category classification, face images of the sample users in each user category are acquired, and age information of the sample users in each user category is acquired in step S202. In step S204, in each user category, the face image of the sample user is labeled according to the age information of the sample user. In step S206, in each user category, the age information prediction model corresponding to the user category is trained using the face image of the labeled sample user, and therefore, the trained age information prediction model corresponds to the user category one by one.
In a specific embodiment, a plurality of facial images of sample users in different areas and different ethnicities are collected by means of a web crawler or from different websites, the above-mentioned clustering model is trained by using the facial images, and the sample users are classified into multiple classes according to the facial images of the sample users and the geographic position information of the sample users through the trained clustering model. In each type of user, the age of each sample user is obtained, and the facial image of the sample user is labeled according to the age of each sample user. After labeling, an age information prediction model is trained for each type of sample user. Since the training process of each age information prediction model is the same, an age information prediction model will be described later as an example.
For each age information prediction model, the model is set up as a tensorflow deep learning framework, which is trained on a GPU (graphics processing unit, graphics processor). Firstly, facial images of a sample user corresponding to the model are divided into a training set, a verification set and a test set, and the image quantity ratio of the training set, the verification set and the test set is 8:1:1. Then, the age information prediction model is trained by using a training set, taking the user category of the yellow race of the A country as an example, when the age information prediction model is trained by the model, an image in the training set is read by the model, the image enters the model through an input layer of the model, a series of transformations are carried out in the model, an output layer of the model is reached, the output layer outputs a result, the result and a label of the image do a difference value, namely the loss value of the training set, parameters of the model can be updated after the loss value is obtained, namely the whole training process of the image is realized, the model can be input one by one for a large number of sample images, and the process is similar.
After the model is trained once by using the training set, the model parameters are updated for a plurality of times, which means that the model has a certain learning effect, and the learning effect of the model can be verified by using the verification set. After the model is trained once through the whole training set, the model is verified on the verification set to check the training progress and the actual effect of the model. During verification, the loss function value of the training set and the loss function value of the verification set can be calculated, in general, the loss function value of the verification set gradually decreases each time of verification, which means that the model is always being learned, the effect is better and better, and when the loss function value of the verification set is not decreasing, and the loss function value of the verification set is equal to or close to the loss function value of the training set, the model training is determined to be completed. To prevent model training from fitting, means such as early stop, regularization, data enhancement, dropout, etc. may be employed. And finally, testing the model by using a test set, and detecting the generalization capability of the model on new data, wherein the generalization capability is a final index for measuring the performance of the model, namely the prediction capability of the model on never seen images is trained, and only the generalization capability is strong, the model can make reasonable predictions on photos of new users, so that the model really brings value.
And after model training is completed, testing the model by using a test set. And inputting the images of the test set one by one into a trained model, comparing an output result of the model with a label, if the output result is consistent with the label, indicating that the model prediction is correct, otherwise, indicating that the model prediction is incorrect. And counting the number of misprediction and the number of correctly predicted, and calculating the accuracy, namely the accuracy of the model on a test set, wherein the accuracy measures the prediction capability of the model on new data outside a training set. If the accuracy of the test set is high, the model is shown to be good on new data, the generalization capability of the model is strong, and the model is further shown to be a good model. Considering that the age information of the sample user is a specific value, the predicted age information is allowed to have a certain error, so that when the error between the predicted age information and the age information of the tag label is within the allowable range, the prediction can be considered to be correct.
And after the model passes the test, carrying out online deployment. Specifically, using the python web service framework flash, a model service is built on the GPU server. The model service can dynamically expand the capacity and process more user access quantity in time along with the increase of the user access quantity. Thus, a user category corresponds to a corresponding age information prediction model, and the age information of the target user is identified more pertinently.
In this embodiment, classifying users, training an age information prediction model for each user category has the following advantages:
1. the sample size required by training each model can be reduced, the training difficulty of each model is reduced, and the model training time is further shortened;
2. Because each age information prediction model specifically predicts the age information of the user in one user category, the accuracy of age information prediction can be improved, and the accuracy of advertisement pushing can be further improved.
Returning to the flow in fig. 1, after determining the target prediction model to which the target user matches, step S106 is performed, by which age information of the target user is predicted from the face image of the target user. In this step, the face image of the target user is input to the target prediction model, so that age information of the target user is predicted by the target prediction model. The predicted age information of the target user may be a specific age value, such as 25 years old, or may be an age range, such as 25-26 years old.
Next, step S108 is executed to push advertisement data to the target user according to the predicted age information of the target user. In one embodiment, advertisement data is pushed to a target user according to the predicted age information of the target user, specifically: and determining target groups matched with the target users in each preset advertisement audience group according to the age information of the target users obtained through prediction, and pushing advertisement data configured for the target groups to the target users.
In this embodiment, a plurality of advertisement audience groups are predetermined, for example, an advertisement audience group is aged 0-10 years old, an advertisement audience group is aged 10-15, an advertisement audience group is aged 15-20, and advertisement data recommended to different advertisement audience groups is different. In the step, firstly, according to the age information of the target user obtained through prediction, a target group matched with the target user is determined in each preset advertisement audience group.
In a specific embodiment, according to the age information of the target user obtained through prediction, a target group matched with the target user is determined in each preset advertisement audience group, specifically: determining the value range of the age information of each advertisement audience group, and determining the advertisement audience group with the value range matched with the age information of the target user as the target group matched with the target user. For example, if the predicted age of the target user is 28 years old, determining the advertisement audience group with the age information in the range of 25-30 as the target group matched with the target user, or if the predicted age of the target user is 28-29 years old, determining the advertisement audience group with the age information in the range of 25-30 as the target group matched with the target user.
Then, pushing advertisement data configured for a target group to a target user, specifically: and acquiring geographic position information of the target user, screening advertisement data matched with the geographic position information from advertisement data configured for the target group, and pushing the screened advertisement data to the target user.
Specifically, the geographical location information of the target user may be geographical location information of the user terminal of the target user, or may be a local area of the target user. The advertisement data matched with the geographic position information is screened from the advertisement data configured for the target group, which can be the advertisement data of the entity store close to the geographic position information, for example, the advertisement data of the fast food store close to the target user is screened from the advertisement data configured for the target group. The advertisement data configured for the target group may be advertisement data in which the use place of the product is close to the geographical position information, for example, the target user is currently located in the park, and advertisement data of the product which can be used in the park is screened for the user. And finally, pushing the screened advertisement data to the target user.
In summary, by the advertisement data pushing method in this embodiment, accuracy of predicting age information of the target user can be improved, and accuracy of advertisement data pushing is further improved.
Fig. 3 is a schematic block diagram of an advertisement data pushing device according to an embodiment of the present disclosure, where the device is configured to implement the above advertisement data pushing method, as shown in fig. 3, and the device includes:
A data acquisition module 31, configured to acquire a face image of a target user and geographic location information of the target user;
A model selection module 32, configured to determine a target prediction model matched with the target user from all pre-trained age information prediction models according to the face image of the target user and the geographic location information of the target user;
An information prediction module 33 for predicting age information of the target user from the face image of the target user through the target prediction model;
And the advertisement pushing module 34 is configured to push advertisement data to the target user according to the predicted age information of the target user.
Optionally, each pre-trained age information prediction model corresponds to a predetermined user category; each user category is divided based on skin color characteristics of the user and geographic position information of the user.
Optionally, the model selection module 32 is specifically configured to:
Determining a target category matched with the target user in each user category according to the face image of the target user and the geographic position information of the target user;
And determining a model corresponding to the target category in each pre-trained age information prediction model according to the one-to-one correspondence between each user category and each pre-trained age information prediction model, and taking the model as a target prediction model matched with the target user.
Optionally, the model selection module 32 is further specifically configured to:
Converting the facial image of the target user and the geographic position information of the target user into feature vectors respectively, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
And determining a target category matched with the target user in each user category according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
Optionally, the system further comprises a classification module for:
Acquiring facial images of a plurality of sample users, and acquiring geographic position information of the sample users;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users to obtain the user categories.
Optionally, the classification module is specifically configured to:
respectively converting the facial image of the sample user and the geographic position information of the sample user into feature vectors, and inputting the feature vectors obtained by conversion into a clustering model based on a preset clustering algorithm;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users through the clustering model to obtain the user categories.
Optionally, the training module is further included for:
Acquiring face images of the sample users in each user category, and acquiring age information of the sample users in each user category;
labeling facial images of the sample users in each user category according to age information of the sample users in each user category;
And training an age information prediction model corresponding to each user category by using the labeled face images of the sample users in each user category.
Optionally, the advertisement pushing module 34 is specifically configured to:
Determining a target group matched with the target user in each preset advertisement audience group according to the predicted age information of the target user;
And pushing advertisement data configured for the target group to the target user.
Optionally, the advertisement pushing module 34 is further specifically configured to:
determining a value range of age information of each advertisement audience group;
and determining the advertisement audience group with the value range matched with the age information of the target user as a target group matched with the target user.
Optionally, the advertisement pushing module 34 is further specifically configured to:
Obtaining geographic position information of the target user;
screening advertisement data matched with the geographic position information from advertisement data configured for the target group;
and pushing the screened advertisement data to the target user.
In this embodiment, according to the facial image of the target user and the geographic location information of the target user, a matched target prediction model is selected for the target user in each pre-trained age information prediction model to perform age prediction, and advertisement data is pushed to the target user according to the age information of the target user obtained by prediction.
The advertisement data pushing device in this embodiment can implement the foregoing processes of the advertisement data pushing method embodiment, and achieve the same functions and effects, which are not repeated here.
An embodiment of the present disclosure further provides an advertisement data pushing device, and fig. 4 is a schematic structural diagram of the advertisement data pushing device provided in an embodiment of the present disclosure, as shown in fig. 4, where the device includes: a memory 601, a processor 602, a bus 603 and a communication interface 604. The memory 601, processor 602, and communication interface 604 communicate over a bus 603, and the communication interface 604 may include input and output interfaces including, but not limited to, a keyboard, mouse, display, microphone, loudspeaker, and the like.
In one embodiment, an advertisement data pushing apparatus includes: a processor; and a memory arranged to store computer executable instructions that, when executed, cause the processor to implement the following:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
and pushing advertisement data to the target user according to the predicted age information of the target user.
Optionally, the computer-executable instructions, when executed, each pre-trained age information prediction model corresponds to a predetermined user category; each user category is divided based on skin color characteristics of the user and geographic position information of the user.
Optionally, the computer executable instructions, when executed, determine a target prediction model matching the target user from among respective pre-trained age information prediction models based on the face image of the target user and the geographic location information of the target user, comprising:
Determining a target category matched with the target user in each user category according to the face image of the target user and the geographic position information of the target user;
And determining a model corresponding to the target category in each pre-trained age information prediction model according to the one-to-one correspondence between each user category and each pre-trained age information prediction model, and taking the model as a target prediction model matched with the target user.
Optionally, the computer executable instructions, when executed, determine a target category matching the target user from among the user categories based on the face image of the target user and the geographic location information of the target user, comprising:
Converting the facial image of the target user and the geographic position information of the target user into feature vectors respectively, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
And determining a target category matched with the target user in each user category according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
Optionally, the computer executable instructions, when executed, each of the user categories is determined by:
Acquiring facial images of a plurality of sample users, and acquiring geographic position information of the sample users;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users to obtain the user categories.
Optionally, when executed, the computer executable instructions cluster the sample users according to skin color features corresponding to the facial images of the sample users and geographic location information of the sample users, to obtain each user category, including:
respectively converting the facial image of the sample user and the geographic position information of the sample user into feature vectors, and inputting the feature vectors obtained by conversion into a clustering model based on a preset clustering algorithm;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users through the clustering model to obtain the user categories.
Optionally, the computer executable instructions, when executed, train the age information prediction model by:
Acquiring face images of the sample users in each user category, and acquiring age information of the sample users in each user category;
labeling facial images of the sample users in each user category according to age information of the sample users in each user category;
And training an age information prediction model corresponding to each user category by using the labeled face images of the sample users in each user category.
Optionally, the computer executable instructions, when executed, push advertisement data to the target user according to the predicted age information of the target user, including:
Determining a target group matched with the target user in each preset advertisement audience group according to the predicted age information of the target user;
And pushing advertisement data configured for the target group to the target user.
Optionally, the computer executable instructions, when executed, determine a target group of the target users matching from among respective predetermined advertisement audience groups based on the predicted age information of the target users, comprising:
determining a value range of age information of each advertisement audience group;
and determining the advertisement audience group with the value range matched with the age information of the target user as a target group matched with the target user.
Optionally, the computer-executable instructions, when executed, push advertisement data configured for the target group to the target user, comprising:
Obtaining geographic position information of the target user;
screening advertisement data matched with the geographic position information from advertisement data configured for the target group;
and pushing the screened advertisement data to the target user.
In this embodiment, according to the facial image of the target user and the geographic location information of the target user, a matched target prediction model is selected for the target user in each pre-trained age information prediction model to perform age prediction, and advertisement data is pushed to the target user according to the age information of the target user obtained by prediction.
The advertisement data pushing device in this embodiment can implement the foregoing processes of the advertisement data pushing method embodiment, and achieve the same functions and effects, which are not repeated here.
Another embodiment of the present specification also provides a storage medium storing computer-executable instructions that, when executed by a processor, perform the method of:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
and pushing advertisement data to the target user according to the predicted age information of the target user.
Optionally, the computer-executable instructions, when executed by the processor, each pre-trained age information prediction model corresponds to a predetermined user category; each user category is divided based on skin color characteristics of the user and geographic position information of the user.
Optionally, the computer executable instructions, when executed by the processor, determine a target prediction model matching the target user from among respective pre-trained age information prediction models based on the face image of the target user and the geographic location information of the target user, comprising:
Determining a target category matched with the target user in each user category according to the face image of the target user and the geographic position information of the target user;
And determining a model corresponding to the target category in each pre-trained age information prediction model according to the one-to-one correspondence between each user category and each pre-trained age information prediction model, and taking the model as a target prediction model matched with the target user.
Optionally, the computer executable instructions, when executed by the processor, determine a target category matching the target user from among the user categories based on the facial image of the target user and the geographic location information of the target user, comprising:
Converting the facial image of the target user and the geographic position information of the target user into feature vectors respectively, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
And determining a target category matched with the target user in each user category according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
Optionally, the computer executable instructions, when executed by the processor, each of the user categories is determined by:
Acquiring facial images of a plurality of sample users, and acquiring geographic position information of the sample users;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users to obtain the user categories.
Optionally, when executed by the processor, the computer executable instructions cluster the sample user according to skin color features corresponding to the facial image of the sample user and geographic location information of the sample user, to obtain each user category, including:
respectively converting the facial image of the sample user and the geographic position information of the sample user into feature vectors, and inputting the feature vectors obtained by conversion into a clustering model based on a preset clustering algorithm;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users through the clustering model to obtain the user categories.
Optionally, the computer executable instructions, when executed by the processor, train the age information prediction model by:
Acquiring face images of the sample users in each user category, and acquiring age information of the sample users in each user category;
labeling facial images of the sample users in each user category according to age information of the sample users in each user category;
And training an age information prediction model corresponding to each user category by using the labeled face images of the sample users in each user category.
Optionally, the computer executable instructions, when executed by the processor, push advertisement data to the target user according to the predicted age information of the target user, including:
Determining a target group matched with the target user in each preset advertisement audience group according to the predicted age information of the target user;
And pushing advertisement data configured for the target group to the target user.
Optionally, the computer executable instructions, when executed by the processor, determine a target group of the target users matching from among respective predetermined advertisement audience groups based on the predicted age information of the target users, comprising:
determining a value range of age information of each advertisement audience group;
and determining the advertisement audience group with the value range matched with the age information of the target user as a target group matched with the target user.
Optionally, the computer executable instructions, when executed by the processor, push advertisement data configured for the target group to the target user, comprising:
Obtaining geographic position information of the target user;
screening advertisement data matched with the geographic position information from advertisement data configured for the target group;
and pushing the screened advertisement data to the target user.
The storage medium includes Read-Only Memory (ROM), random access Memory (Random Access Memory RAM), magnetic disk or optical disk, etc.
In this embodiment, according to the facial image of the target user and the geographic location information of the target user, a matched target prediction model is selected for the target user in each pre-trained age information prediction model to perform age prediction, and advertisement data is pushed to the target user according to the age information of the target user obtained by prediction.
The computer executable instructions stored in the storage medium in this embodiment, when executed by the processor, can implement the foregoing respective processes of the advertisement data pushing method embodiment, and achieve the same functions and effects, and are not repeated here.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments.
The foregoing is merely exemplary of the present disclosure and is not intended to limit the present disclosure. Various modifications and variations of the embodiments described herein will be apparent to those skilled in the art. Any modifications, equivalent substitutions, improvements, etc. that fall within the spirit and principles of the present document are intended to be included within the scope of the claims of the present document.
Claims (11)
1. The advertisement data pushing method is characterized by comprising the following steps:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user, wherein each pre-trained age information prediction model corresponds to a predetermined user category, and each user category is obtained based on skin color characteristics of the user and geographic position information of the user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
pushing advertisement data to the target user according to the predicted age information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user, wherein the method comprises the following steps:
Converting the facial image of the target user and the geographic position information of the target user into feature vectors respectively, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
And determining a target category matched with the target user in each user category according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
2. The method of claim 1, wherein determining a target prediction model matching the target user from among respective pre-trained age information prediction models based on the face image of the target user and the geographic location information of the target user comprises:
And determining a model corresponding to the target category in each pre-trained age information prediction model according to the one-to-one correspondence between each user category and each pre-trained age information prediction model, and taking the model as a target prediction model matched with the target user.
3. The method of claim 1, wherein each of the user categories is determined by:
Acquiring facial images of a plurality of sample users, and acquiring geographic position information of the sample users;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users to obtain the user categories.
4. The method of claim 3, wherein clustering the sample users according to skin color features corresponding to the facial images of the sample users and geographic location information of the sample users to obtain the user categories comprises:
respectively converting the facial image of the sample user and the geographic position information of the sample user into feature vectors, and inputting the feature vectors obtained by conversion into a clustering model based on a preset clustering algorithm;
And clustering the sample users according to the skin color characteristics corresponding to the facial images of the sample users and the geographic position information of the sample users through the clustering model to obtain the user categories.
5. A method according to claim 3, wherein the age information prediction model is trained by:
Acquiring face images of the sample users in each user category, and acquiring age information of the sample users in each user category;
labeling facial images of the sample users in each user category according to age information of the sample users in each user category;
And training an age information prediction model corresponding to each user category by using the labeled face images of the sample users in each user category.
6. The method of claim 1, wherein pushing advertisement data to the target user based on the predicted age information of the target user comprises:
Determining a target group matched with the target user in each preset advertisement audience group according to the predicted age information of the target user;
And pushing advertisement data configured for the target group to the target user.
7. The method of claim 6, wherein determining the target group to which the target user matches among the respective predetermined advertising audience groups based on the predicted age information of the target user comprises:
determining a value range of age information of each advertisement audience group;
and determining the advertisement audience group with the value range matched with the age information of the target user as a target group matched with the target user.
8. The method of claim 6, wherein pushing advertisement data configured for the target group to the target user comprises:
Obtaining geographic position information of the target user;
screening advertisement data matched with the geographic position information from advertisement data configured for the target group;
and pushing the screened advertisement data to the target user.
9. An advertising data pushing apparatus, comprising:
The data acquisition module is used for acquiring the face image of the target user and the geographic position information of the target user;
The model selection module is used for determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user, wherein each pre-trained age information prediction model corresponds to a predetermined user category, and each user category is obtained based on skin color characteristics of the user and geographic position information of the user in a dividing mode;
the information prediction module is used for predicting age information of the target user according to the facial image of the target user through the target prediction model;
The advertisement pushing module is used for pushing advertisement data to the target user according to the predicted age information of the target user;
the model selection module is used for respectively converting the facial image of the target user and the geographic position information of the target user into feature vectors, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
And determining a target category matched with the target user in each user category according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
10. An advertising data pushing apparatus, comprising:
A processor; and
A memory arranged to store computer executable instructions that, when executed, cause the processor to implement the following:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user, wherein each pre-trained age information prediction model corresponds to a predetermined user category, and each user category is obtained based on skin color characteristics of the user and geographic position information of the user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
pushing advertisement data to the target user according to the predicted age information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user, wherein the method comprises the following steps:
Converting the facial image of the target user and the geographic position information of the target user into feature vectors respectively, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
And determining a target category matched with the target user in each user category according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
11. A storage medium for storing computer-executable instructions which, when executed by a processor, perform the method of:
Acquiring a face image of a target user and geographic position information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user, wherein each pre-trained age information prediction model corresponds to a predetermined user category, and each user category is obtained based on skin color characteristics of the user and geographic position information of the user;
predicting age information of the target user according to the facial image of the target user through the target prediction model;
pushing advertisement data to the target user according to the predicted age information of the target user;
Determining a target prediction model matched with the target user in each pre-trained age information prediction model according to the face image of the target user and the geographic position information of the target user, wherein the method comprises the following steps:
Converting the facial image of the target user and the geographic position information of the target user into feature vectors respectively, and inputting the feature vectors obtained by conversion into a clustering model for dividing each user category;
And determining a target category matched with the target user in each user category according to the skin color characteristics corresponding to the facial image of the target user and the geographic position information of the target user through the clustering model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110651316.7A CN113435922B (en) | 2021-06-10 | 2021-06-10 | Advertisement data pushing method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110651316.7A CN113435922B (en) | 2021-06-10 | 2021-06-10 | Advertisement data pushing method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113435922A CN113435922A (en) | 2021-09-24 |
| CN113435922B true CN113435922B (en) | 2024-04-26 |
Family
ID=77755633
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110651316.7A Active CN113435922B (en) | 2021-06-10 | 2021-06-10 | Advertisement data pushing method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113435922B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102930454A (en) * | 2012-10-07 | 2013-02-13 | 乐配(天津)科技有限公司 | Intelligent 3D (Three Dimensional) advertisement recommendation method based on multiple perception technologies |
| CN104143079A (en) * | 2013-05-10 | 2014-11-12 | 腾讯科技（深圳）有限公司 | Method and system for face attribute recognition |
| CN109299701A (en) * | 2018-10-15 | 2019-02-01 | 南京信息工程大学 | Expand the face age estimation method that more ethnic group features cooperate with selection based on GAN |
| CN109730637A (en) * | 2018-12-29 | 2019-05-10 | 中国科学院半导体研究所 | A kind of face face-image quantified system analysis and method |
| CN109829071A (en) * | 2018-12-14 | 2019-05-31 | 平安科技（深圳）有限公司 | Face image searching method, server, computer equipment and storage medium |
| CN110012060A (en) * | 2019-02-13 | 2019-07-12 | 平安科技（深圳）有限公司 | Information-pushing method, device, storage medium and the server of mobile terminal |
| CN110046571A (en) * | 2019-04-15 | 2019-07-23 | 北京字节跳动网络技术有限公司 | The method and apparatus at age for identification |
| CN110310144A (en) * | 2019-05-22 | 2019-10-08 | 深圳壹账通智能科技有限公司 | Advertisement sending method, device, equipment and storage medium based on the age |
| CN110717401A (en) * | 2019-09-12 | 2020-01-21 | Oppo广东移动通信有限公司 | Age estimation method and device, equipment and storage medium |
-
2021
- 2021-06-10 CN CN202110651316.7A patent/CN113435922B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102930454A (en) * | 2012-10-07 | 2013-02-13 | 乐配(天津)科技有限公司 | Intelligent 3D (Three Dimensional) advertisement recommendation method based on multiple perception technologies |
| CN104143079A (en) * | 2013-05-10 | 2014-11-12 | 腾讯科技（深圳）有限公司 | Method and system for face attribute recognition |
| CN109299701A (en) * | 2018-10-15 | 2019-02-01 | 南京信息工程大学 | Expand the face age estimation method that more ethnic group features cooperate with selection based on GAN |
| CN109829071A (en) * | 2018-12-14 | 2019-05-31 | 平安科技（深圳）有限公司 | Face image searching method, server, computer equipment and storage medium |
| CN109730637A (en) * | 2018-12-29 | 2019-05-10 | 中国科学院半导体研究所 | A kind of face face-image quantified system analysis and method |
| CN110012060A (en) * | 2019-02-13 | 2019-07-12 | 平安科技（深圳）有限公司 | Information-pushing method, device, storage medium and the server of mobile terminal |
| CN110046571A (en) * | 2019-04-15 | 2019-07-23 | 北京字节跳动网络技术有限公司 | The method and apparatus at age for identification |
| CN110310144A (en) * | 2019-05-22 | 2019-10-08 | 深圳壹账通智能科技有限公司 | Advertisement sending method, device, equipment and storage medium based on the age |
| CN110717401A (en) * | 2019-09-12 | 2020-01-21 | Oppo广东移动通信有限公司 | Age estimation method and device, equipment and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| Barathi, J. Jeya ; Kavitha, G ; Imran, M. Mohamed ; .Building a Mobile Personalized Marketing System using Multidimensional Data.International Conference on Smart Technologies and Management for Computing, Communication, Controls, Energy and Materials (ICSTM).2015,全文. * |
| Kim, Yu Hwan ; Nam, Se Hyun ; Park, Kang Ryoung.Enhanced Cycle Generative Adversarial Network for Generating Face Images of Untrained Races and Ages for Age Estimation. IEEE Access .2020,第9卷全文. * |
| 基于肤色模型及神经网络的人脸检测算法;段沛沛;邵浩杰;文煜洁;邱雪燕;尹玉萍;;科技经济导刊(第05期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113435922A (en) | 2021-09-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111444952A (en) | Method and device for generating sample identification model, computer equipment and storage medium | |
| CN111126514A (en) | Image multi-label classification method, device, equipment and medium | |
| CN111950528B (en) | Graph recognition model training method and device | |
| CN106651057A (en) | Mobile terminal user age prediction method based on installation package sequence table | |
| CN111667308A (en) | Advertisement recommendation prediction system and method | |
| CN111191133B (en) | Service search processing method, device and equipment | |
| CN103761254A (en) | Method for matching and recommending service themes in various fields | |
| CN110020022B (en) | Data processing method, device, equipment and readable storage medium | |
| CN117011616B (en) | Image content auditing method and device, storage medium and electronic equipment | |
| CN117332090B (en) | Sensitive information identification method, device, equipment and storage medium | |
| WO2022028147A1 (en) | Image classification model training method and apparatus, computer device, and storage medium | |
| CN113435922B (en) | Advertisement data pushing method, device, equipment and storage medium | |
| KR101028810B1 (en) | Apparatus and method for analyzing advertisement target | |
| CN114692889A (en) | Meta-feature training model for machine learning algorithm | |
| CN111782774B (en) | Method and device for recommending problems | |
| CN116127156B (en) | Charging station recommendation method and device, electronic equipment and readable storage medium | |
| CN116611891A (en) | Content information recommendation method, device, server and storage medium | |
| CN116664250A (en) | Content information recommendation method, device, server and storage medium | |
| CN118035800A (en) | Model training method, device, equipment and storage medium | |
| CN116468907A (en) | Method and device for image processing, image classification and image detection | |
| CN118613737A (en) | Method and device for determining the position of a mobile device | |
| CN115982351A (en) | Test question evaluation method and related device, electronic equipment and storage medium | |
| CN115205606A (en) | Image multi-label classification method and device and related products | |
| CN114241253A (en) | Model training method, system, server and storage medium for illegal content identification | |
| CN114186043B (en) | Pre-training method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |