CN119205221B

CN119205221B - Advertisement push method and system based on age and gender recognition

Info

Publication number: CN119205221B
Application number: CN202411295911.1A
Authority: CN
Inventors: 张政
Original assignee: Zhejiang Tianhai Technology Co ltd
Current assignee: Zhejiang Tianhai Technology Co ltd
Priority date: 2024-09-14
Filing date: 2024-09-14
Publication date: 2025-03-18
Anticipated expiration: 2044-09-14
Also published as: CN119205221A

Abstract

The present invention discloses an advertisement push method and system based on age and gender recognition, and relates to the field of intelligent advertisement push technology, including: using a multimodal fusion network to perform multimodal fusion on the extracted user features, and comprehensively analyzing the user's age and gender information; identifying the advertisement material that best matches the user's age and gender from the advertisement library according to the user's age and gender information; based on the selected advertisement material, using NLP to generate personalized advertisement copy, and using GANs to generate personalized advertisement pictures and videos; combining personalized advertisement copy with personalized advertisement pictures and videos to generate personalized advertisement content. Through multimodal data collection and fusion, the present invention improves the accuracy of age and gender recognition, accurately matches advertisement content, and generates personalized copy and vision, thereby enhancing advertisement appeal and user experience.

Description

Advertisement pushing method and system based on age and gender identification

Technical Field

The invention relates to the technical field of intelligent advertisement pushing, in particular to an advertisement pushing method and system based on age and gender identification.

Background

In recent years, with rapid development of big data and artificial intelligence technology, personalized advertisement pushing has become an important means for improving advertisement effect and user experience. The traditional advertisement pushing system mainly relies on basic information and historical behavior data of users, and targeted advertisement delivery is achieved through simple statistical analysis. However, with diversification of user demands and enrichment of advertisement contents, the conventional single-mode data processing method cannot meet the requirement of accurate pushing. Under the background, the multi-mode information fusion technology gradually becomes a research hot spot, and the characteristics of a user can be more comprehensively analyzed by integrating various data sources such as facial images, voice signals and gait information of the user, so that the accuracy and individuation level of advertisement pushing are improved.

Although some advertisement recommendation systems based on user behaviors and preferences exist in the market at present, the systems have the defects that firstly, most of the systems depend on data of a single mode (such as historical browsing records or shopping behaviors of users), other potential useful information sources are ignored, so that the accuracy of advertisement pushing is limited, secondly, most of the existing age and sex identification technologies are based on facial identification, are easily influenced by light conditions, shielding objects and individual differences, the identification accuracy is low, and the requirement of accurate advertisement pushing is difficult to meet. In addition, the existing advertisement pushing system lacks an effective feedback mechanism, advertisement content and pushing strategies cannot be timely adjusted according to real-time feedback of users, so that user experience is poor, and advertisement conversion rate is low.

Disclosure of Invention

The present invention has been made in view of the above-described problems occurring in the prior art.

Therefore, the invention provides an advertisement pushing method and system based on age and gender identification, which solve the problems of insufficient multi-mode information fusion, inaccurate age and gender identification and low advertisement pushing accuracy.

In order to solve the technical problems, the invention provides the following technical scheme:

The embodiment of the invention provides an advertisement pushing method based on age and gender identification, which comprises the steps of collecting user information, carrying out feature extraction on the collected user information, carrying out polymorphic fusion on the extracted user features by using a multi-modal fusion network, comprehensively analyzing age and gender information of a user, identifying advertisement materials which are most matched with the age and the gender of the user from an advertisement library according to the age and the gender information of the user, generating personalized advertisement patterns by using NLP (non-line language) based on the selected advertisement materials, generating personalized advertisement pictures and videos by using G (non-line language), combining the personalized advertisement patterns with the personalized advertisement pictures and videos to generate personalized advertisement contents, pushing the personalized advertisement contents to the user, collecting user feedback in real time, and optimizing the advertisement contents and pushing strategies according to the real-time feedback of the user.

The invention relates to an advertisement pushing method based on age and sex identification, which comprises the following steps of collecting user information, extracting characteristics of the collected user information,

Capturing facial images of a user through a high-resolution camera installed in a public place, recording the voice of the user through an integrated microphone array, and detecting gait information of the user through a thermal imager and an infrared sensor;

detecting facial regions from the image using a face detection algorithm, extracting facial features including eye, nose and mouth features;

noise reduction processing is carried out on the audio signal, FFT is used for carrying out frequency spectrum analysis, and voice characteristics including fundamental frequency and formant frequency are extracted;

The gait information is filtered and smoothed, noise is removed by using a Kalman filter, and gait characteristics including step size, stride and pace are extracted.

As a preferable scheme of the advertisement pushing method based on age and gender identification, the invention adopts a multi-mode fusion network to carry out multi-mode fusion on the extracted user characteristics, comprehensively analyzes age and gender information of the user, comprises the following specific steps,

The extracted facial features, language features and gait features are fused by using a multi-modal fusion network, and the expression is as follows:

F=[F_f;F_v;F_g];

Wherein F is a comprehensive feature vector, F _f is a facial feature vector, F _v is a speech feature vector, and F _g is a gait feature vector;

Based on the integrated comprehensive feature vector F, comprehensively analyzing age and gender information of the user by utilizing a comprehensive perception model MLP;

Age analysis of the user, expressed as:

Wherein A is a user age prediction value, F _i is an ith feature in the integrated feature vector F, alpha _i is an amplitude coefficient of the ith feature, d _j is a bias term of the jth layer, m is the number of hidden layers, n is the dimension of the integrated feature vector F, sigma is an activation function, e is a global bias term, tanh is a nonlinear activation function, and Y _i is a phase offset of the ith feature vector;

The gender analysis of the user is expressed as:

wherein β _i is the frequency coefficient of the ith feature, G represents the sex bias;

Defining a distinguishing threshold T _G based on the comprehensive feature vector F, and judging the gender of the user;

When G > T _G, then determining as female;

when G < T _G, then a male is determined.

As an optimal scheme of the advertisement pushing method based on age and gender identification, the advertisement material which is most matched with the age and the gender of the user is identified from an advertisement library according to the age and the gender information of the user, and the specific steps are as follows,

Defining an advertisement library comprising a plurality of advertisement materials based on the product information and the creative content, each advertisement material having a target age range and a target gender bias rate associated therewith;

According to the age and sex of the user, defining a matching function M, and identifying the matching degree of each advertisement material and the age and sex information of the user, wherein the expression is as follows:

Wherein q _k represents the kth advertisement material, mu _k represents the average value of the target age range of the advertisement material q _k, & _k represents the standard deviation of the target age range of the advertisement material q _k, ρ _k represents the central value of the target gender deviation rate of the advertisement material q _k, T _k represents the standard deviation of the target gender deviation rate of the advertisement material q _k, and M (q _k, A, G) is the matching degree of the advertisement material q _k with the user age predicted value A and the gender deviation rate G;

defining a matching degree standard threshold T _M based on service requirements and historical data, and identifying high-matching degree advertisement materials;

When M (q _k,A,G)≥T_M, consider the current advertisement material to be a high-matching-degree material;

When M (q _k,A,G)＜T_M), the current advertisement material is not considered to belong to the high-matching-degree material.

The invention relates to an advertisement pushing method based on age and sex identification, which is characterized in that based on selected advertisement materials, an NLP is used for generating personalized advertisement texts, GANs is used for generating personalized advertisement pictures and videos, the specific steps are as follows,

Extracting key information from the selected advertisement materials, including product names, functional characteristics and target audience, and forming a keyword set K;

According to the type of the advertisement material and the target audience generation document template, filling the related information in the keyword set K into the placeholder position in the document template to generate a preliminary personalized advertisement document;

Performing text color rendering on the generated preliminary text by using NLP, including synonym replacement and sentence structure adjustment, and generating a personalized advertisement text;

according to the content and style of the selected advertisement materials, relevant reference pictures and video clips are collected, and a GANs model is trained;

and generating personalized pictures and videos consistent with the style of the advertisement materials by using the trained GANs model.

As an optimal scheme of the advertisement pushing method based on age and gender identification, the invention combines the personalized advertisement file with the personalized advertisement picture and video to generate personalized advertisement content, and comprises the following specific steps,

Matching and combining the generated personalized advertisement document and the personalized advertisement picture by coordinating the consistency of the document and the picture content, so as to enhance visual appeal;

The generated personalized advertisement document is embedded into the personalized video, and the document content is presented in the forms of subtitles and side notes, so that the video attraction is improved;

after the combining is completed, the final personalized advertising content is derived using Adobe Premiere Pro.

As a preferred scheme of the advertisement pushing method based on age and gender identification, the invention comprises the steps of pushing personalized advertisement content to a user, collecting user feedback in real time, optimizing advertisement content and pushing strategy according to the real-time feedback of the user, specifically comprising the following steps,

Pushing the generated personalized advertisement content to a target user through social media according to the user activity period;

Through a built-in feedback collection mechanism, the response of the user to the advertisement, including click rate, stay time and sharing times, is collected in real time, and subjective feedback comments of the user are directly obtained through questionnaire investigation and online evaluation;

Optimizing pushing time, frequency and advertisement content according to feedback content of a user, and improving effective contact rate of advertisements;

for the user with positive feedback, more personalized advertisement content is provided, and for the user with negative feedback, the pushing strategy is optimized, so that the interference to the life of the user is reduced.

The invention provides an advertisement pushing system based on age and gender identification, which comprises a data acquisition and feature extraction module, a multi-mode fusion and analysis module, an advertisement material matching module, an advertisement content generation module and an advertisement pushing and feedback module, wherein the advertisement material matching module is used for generating personalized advertisement texts by using NLP, generating personalized advertisement pictures and videos by using GANs, the data acquisition and feature extraction module is used for acquiring user information and carrying out feature extraction on the acquired user information, the multi-mode fusion and analysis module is used for carrying out multi-mode fusion on the extracted user features by using a multi-mode fusion network and comprehensively analyzing the age and gender information of a user, the advertisement material matching module is used for identifying advertisement materials which are matched with the age and the gender of the user from an advertisement library according to the age and gender information of the user, the advertisement content generating module is used for generating personalized advertisement texts by using NLP, generating personalized advertisement pictures and videos by using GANs, generating personalized advertisement contents by combining the personalized advertisement texts and the personalized advertisement pictures and videos, and the advertisement pushing and feedback module is used for pushing the personalized advertisement contents to the user, and optimizing advertisement pushing strategies of the user in real time.

In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, where the computer program, when executed by the processor, implements any step of the advertisement pushing method based on age and gender identification according to the first aspect of the present invention.

In a fourth aspect, embodiments of the present invention provide a computer readable storage medium having a computer program stored thereon, wherein the computer program when executed by a processor implements any of the steps of the age and gender identification based advertisement pushing method according to the first aspect of the present invention.

The method has the advantages that user characteristics are comprehensively obtained from facial images, voice signals and gait information through multi-mode data acquisition and feature extraction, the diversity and accuracy of data are enhanced, multi-source heterogeneous data are effectively integrated through multi-mode fusion and comprehensive analysis, the accuracy of age and gender identification is improved, advertisement contents most matched with the user characteristics are accurately screened out through selection of advertisement materials with high matching degree, the relevance and effectiveness of advertisements are improved, personalized advertisement content generation is performed, highly customized text and visual contents are generated through NLP and GANs technology, the attractiveness and interactivity of advertisements are enhanced, the accuracy and individuation of advertisement pushing are finally realized, and user experience and advertisement conversion rate are greatly improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a flowchart of an advertisement pushing method based on age and gender identification in embodiment 1.

Fig. 2 is a flowchart of an advertisement push system based on age and gender identification in embodiment 1.

Detailed Description

In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, but the present invention may be practiced in other ways other than those described herein, and persons skilled in the art will readily appreciate that the present invention is not limited to the specific embodiments disclosed below.

Further, reference herein to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic can be included in at least one implementation of the invention. The appearances of the phrase "in one embodiment" in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.

Embodiment 1, referring to fig. 1 and 2, provides an advertisement pushing method based on age and gender identification, which comprises the following steps:

S1, collecting user information, and extracting features of the collected user information.

Further, capturing facial images of the user through a high-resolution camera installed in a public place, recording the voice of the user through an integrated microphone array, and detecting gait information of the user through a thermal imager and an infrared sensor;

Facial features, including eye, nose, and mouth features, are extracted from the image using a face detection algorithm, for example, after the high resolution camera captures an image of the user's face, the system locates the facial region using a face detection algorithm (e.g., MTCNN based on deep learning) and further identifies the specific locations of the eyes, nose, and mouth. By analyzing the morphology and position of these key parts, the system can extract the size, spacing, shape and position of the nose, and contour of the mouth. These facial feature data will be used for subsequent age and gender identification analysis.

The system performs noise reduction processing on the audio signal, performs spectrum analysis by using FFT, extracts voice characteristics including fundamental frequency and formant frequency, and removes environmental noise after the integrated microphone array records the voice of the user. Next, the time domain signal is converted into a frequency domain signal using a Fast Fourier Transform (FFT), from which the fundamental frequency of the speech (i.e., the fundamental frequency of the vocal cord vibration) and the formant frequencies (i.e., the resonance frequencies of the different vowels in the sound) are extracted. These voice features will help the system more accurately determine the age and gender of the user.

The gait information is filtered and smoothed, noise is removed by using a Kalman filter, and gait characteristics including step size, stride and pace are extracted. For example, when the thermal imager and infrared sensor detect gait information of a user, the system first filters and smoothes the raw data to eliminate environmental interference and noise introduced by sensor errors. Then, the data is further optimized by using a Kalman filter, so that the stability and the accuracy of the data are improved. By analyzing the processed data, the system is able to extract the user's step size (length of each stride), stride (total length of two consecutive strides), and pace (walking speed). These gait features will provide additional support information for age and gender identification.

S2, carrying out polymorphic fusion on the extracted user characteristics by using a multi-modal fusion network, and comprehensively analyzing age and gender information of the user.

Furthermore, the extracted facial features, language features and gait features are fused by using a multi-modal fusion network, and the expression is as follows:

F=[F_f;F_v;F_g];

It should be noted that the system first represents the extracted facial features (such as the morphology of eyes, nose and mouth), language features (such as fundamental frequency and formant frequency), and gait features (such as step size, stride and pace) as feature vectors Ff, F _v and F _g, respectively, and then fuses the three feature vectors into a comprehensive feature vector f= [ F _f;F_v;F_g ] using a multi-modal fusion network, and processes the fused feature vectors through a multi-layer perceptron (MLP) of the neural network to finally generate a comprehensive user feature representation for subsequent age and gender identification analysis. The process effectively integrates multi-source heterogeneous data through the deep learning model, and improves accuracy and robustness of age and gender identification.

Age analysis of the user, expressed as:

Wherein A is a user age predictor, F _i is the ith feature in the integrated feature vector F, alpha _i is the amplitude coefficient of the ith feature, d _j is the bias term of the jth layer, m is the number of hidden layers, n is the dimension of the integrated feature vector F, sigma is an activation function (Sigmoid function), e is a global bias term, tanh is a nonlinear activation function, and Y _i is the phase offset of the ith feature vector;

The gender analysis of the user is expressed as:

When G > T _G, then determining as female;

when G < T _G, then a male is determined.

It should also be noted that the activation function σ maps the weighted input value for each node in the neural network into a nonlinear interval so that the model can capture complex patterns in the data. In this process, the activation function serves to transform the internal calculation result to more conform to the actual probability distribution, thereby improving accuracy and robustness of age and gender identification. For example, in an age analysis, the activation function converts the internal calculation into a value between 0 and 1, which can be interpreted as a probability of a user's age prediction, and in a gender analysis, the activation function also converts the internal calculation into a probability value representing a probability of a user's gender bias. In this way, the activation function helps the system better understand and identify the age and gender characteristics of the user.

And S3, identifying advertisement materials which are most matched with the age and the sex of the user from an advertisement library according to the age and the sex information of the user.

Further, an advertisement library is defined based on product information and creative content, wherein each advertisement material has a target age range and a target gender bias rate associated therewith, and the specific process is to determine the target age range and the target gender bias rate of each advertisement material by analyzing basic information (such as product name, functional characteristics, brand image, etc.) and creative content (such as text, pictures, and videos) of the product, thereby constructing a systematic advertisement library, and ensuring that each advertisement material can precisely match a specific audience group.

According to the comparison result, the system can screen out the advertisement material with high matching degree which is most suitable for the target user, thereby improving the accuracy and effect of advertisement pushing.

S4, based on the selected advertisement materials, generating personalized advertisement texts by using NLP, and generating personalized advertisement pictures and videos by using GANs;

Further, key information including product name, functional features and target audience are extracted from the selected advertisement materials to form a keyword set K, for example, product name "Smart watch", functional features "health monitor, motion tracking, smart Notification" and target audience "young staff" are extracted from a selected advertisement material by the system. These key information are summarized into a set of keywords K, including "smart watch", "health monitor", "motion tracking", "smart notification" and "young staff". This process ensures that the extracted key information accurately reflects the core content and target audience characteristics of the advertising material.

According to the type of the advertisement material and the target audience, a document template is generated, the related information in the keyword set K is filled in the placeholder position in the document template to generate a preliminary personalized advertisement document, for example, according to the type of the advertisement material and the target audience, a document template is generated, and the template comprises the placeholder. Assuming the advertising material is about a smart watch, the target audience is a young office, and the keyword set K includes "smart watch", "health monitor", "motion tracking", "smart notification" and "young office".

For example, the intelligent watch specially designed for young staff is provided with health monitoring, motion tracking and intelligent notification, so that your life is more convenient and healthy, the intelligent watch specially designed for busy young staff is provided with health monitoring, motion tracking and intelligent notification functions, and people can manage life easily and enjoy healthier each day. Through synonym replacement and sentence structure adjustment, the text is more vivid and attractive.

The method comprises the steps of collecting relevant reference pictures and video clips according to the content and style of the selected advertisement materials, training GANs the model, specifically comprising the steps of collecting a series of relevant reference pictures and video clips according to the content and style of the selected advertisement materials, and training and generating an countermeasure network (GANs) model by using the reference materials to ensure that the generated personalized pictures and videos are consistent with the style of the advertisement materials.

It should also be noted that, through the trained GANs model, corresponding personalized pictures and videos are generated according to the content and style of the advertisement materials, so that the generated visual content is ensured to be consistent with the style of the advertisement materials, and the attraction and the relevance of the advertisement are enhanced.

S5, combining the personalized advertisement file with the personalized advertisement picture and video to generate personalized advertisement content;

Furthermore, the generated personalized advertisement document and the personalized advertisement picture are matched and combined by coordinating the consistency of the document and the picture content, so that the visual appeal is enhanced;

For example, the intelligent watch which is specially designed for busy young staff and provided with the personalized advertisement file has the health monitoring, motion tracking and intelligent notification functions, helps you to easily manage life, enjoys healthier every day, is displayed at the key moment of the video in the form of subtitles, and ensures that the file content is coordinated with a video picture by matching with corresponding bystandings, thereby improving the attraction of the video.

It should also be noted that, by coordinating the consistency of the text and the picture content and embedding the personalized advertisement text into the video, the text is presented in the form of subtitles and side notes, so as to ensure that the text is coordinated with the video picture, thereby improving the attraction of the video. And finally, adobe Premiere Pro is used for deriving high-quality personalized advertisement content, so that the overall consistency and visual appeal of advertisements are enhanced, and the user experience and advertisement effect are improved.

And S6, pushing the personalized advertisement content to the user, collecting user feedback in real time, and optimizing the advertisement content and the pushing strategy according to the real-time feedback of the user.

Further, according to the user activity period, the generated personalized advertisement content is pushed to the target user through social media;

It should be further noted that, through a built-in feedback collection mechanism, the instant response of the user to the pushed advertisement is automatically tracked and recorded, such as click rate, page residence time, sharing behavior and other objective data, and the subjective feedback opinion of the user is collected by means of questionnaire investigation, online evaluation and other means, then, according to the collected feedback information, the system can dynamically adjust the pushing time, pushing frequency and content of the advertisement, so as to improve the effective contact rate of the advertisement, for those users with positive feedback, the system further provides personalized advertisement content more in line with the interests of the users, and for users with negative feedback, the pushing strategy is adjusted, the pushing frequency is reduced or the pushing mode is changed, so that the disturbance to the user is reduced, and the user experience is improved. The mechanism not only improves the user experience, but also enhances the self-adaptive capacity of the system, so that the system can continuously improve and optimize the advertisement pushing strategy, thereby improving the overall advertisement effect and the user satisfaction.

The embodiment also provides an advertisement pushing system based on age and gender identification, which comprises a data acquisition and feature extraction module, a multi-mode fusion and analysis module, an advertisement material matching module, an advertisement content generation module and an advertisement pushing and feedback module, wherein the advertisement material matching module is used for generating personalized advertisement texts by using NLP, generating personalized advertisement pictures and videos by using GANs, the data acquisition and feature extraction module is used for acquiring user information and carrying out feature extraction on the acquired user information, the multi-mode fusion and analysis module is used for carrying out multi-mode fusion on the extracted user features by using a multi-mode fusion network and comprehensively analyzing the age and gender information of a user, the advertisement material matching module is used for identifying advertisement materials which are most matched with the age and the gender of the user from an advertisement library according to the age and gender information of the user, the advertisement content generating module is used for generating personalized advertisement texts by using NLP, generating personalized advertisement pictures and videos by using GANs, and the advertisement content generation module is used for combining the personalized advertisement texts and the personalized advertisement pictures and videos to generate personalized advertisement contents, and the pushing and feedback module is used for pushing the personalized advertisement contents to the user, and optimizing advertisement pushing contents according to the real-time feedback of the user and the user feedback.

The embodiment also provides computer equipment, which is suitable for the situation of the advertisement pushing method based on age and gender identification and comprises a memory and a processor, wherein the memory is used for storing computer executable instructions, and the processor is used for executing the computer executable instructions to realize the advertisement pushing method based on age and gender identification, which is provided by the embodiment.

The computer device may be a terminal comprising a processor, a memory, a communication interface, a display screen and input means connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The communication interface of the computer device is used for carrying out wired or wireless communication with an external terminal, and the wireless mode can be realized through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer equipment can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer equipment can be a touch layer covered on the display screen, can also be keys, a track ball or a touch pad arranged on the shell of the computer equipment, and can also be an external keyboard, a touch pad or a mouse and the like.

The present embodiment also provides a storage medium having a computer program stored thereon, which when executed by a processor implements the advertisement pushing method based on age and gender identification as proposed in the above embodiments, and the storage medium may be implemented by any type of volatile or non-volatile storage device or combination thereof, such as a static random access Memory (Static Random Access Memory, SRAM for short), an electrically erasable Programmable Read-Only Memory (ELECTRICALLY ERASABLE PROGRAMMABLE READ-Only Memory, EEPROM for short), an erasable Programmable Read-Only Memory (Erasable Programmable Read Only Memory, EPROM for short), a Programmable Read-Only Memory (ROM for short), a magnetic Memory, a flash Memory, a magnetic disk, or an optical disk.

In summary, the invention realizes the comprehensive acquisition of user characteristics from facial images, voice signals and gait information through multi-mode data acquisition and characteristic extraction, enhances the diversity and accuracy of data, effectively integrates multi-source heterogeneous data through multi-mode fusion and comprehensive analysis, improves the accuracy of age and gender identification, precisely screens out the advertisement content most conforming to the user characteristics through the selection of advertisement materials with high matching degree, improves the relevance and effectiveness of advertisements, generates highly customized text and visual content through personalized advertisement content generation by utilizing NLP and GANs technology, enhances the attraction and interactivity of advertisements, finally realizes the precision and individuation of advertisement pushing, and greatly improves the user experience and advertisement conversion rate.

Embodiment 2, referring to table 1, for the second embodiment of the present invention, for further verifying the technical solution of the present invention, experimental simulation data of an advertisement pushing method based on age and gender identification is provided.

In order to verify the effectiveness of the present invention, the present embodiment provides several information acquisition points within one shopping mall. The acquisition point is equipped with a high resolution camera, a microphone array, a thermal imager, and an infrared sensor. These devices are used to capture facial images of a user, record sounds of the user, and detect gait information of the user.

Firstly, an information acquisition point is arranged in a shopping center in an experiment, facial images, sounds and gait information of a user are acquired by using a high-resolution camera, a microphone array, a thermal imager and an infrared sensor, and facial, voice and gait characteristics are extracted through an algorithm.

And secondly, the extracted multi-mode features are comprehensively processed through a multi-mode fusion network, so that comprehensive feature vectors are generated and are used for accurately identifying the age and the gender of the user.

And then, the system screens out advertisement materials with high matching degree from the advertisement library according to age and gender information of the user, and generates personalized advertisement texts and picture videos by using NLP and GANs technology.

And finally, the personalized advertisement content is pushed to the user through the social media platform, user feedback is collected in real time, and the advertisement pushing strategy is optimized according to the feedback, so that the user experience and the advertisement effect are improved.

The details are shown in table 1 below:

TABLE 1 statistics of user information and advertisement effectiveness

The experimental data show that the multi-mode user information acquisition and analysis system provided by the invention has remarkable accuracy in the aspects of age prediction and gender identification. Taking user 1 as an example, the age prediction value is 28 years old, and the gender bias rate is 0.54 (approaching 0.5 means that the gender bias is not obvious), which indicates that the system can more accurately identify the age and gender characteristics of the user. Further observing the advertisement material match, user 1's match is 0.82, which means that the system can effectively match the advertisement material with the user's age and gender information.

According to the advertisement pushing method based on the age and the gender, the accuracy and the effectiveness of advertisements can be remarkably improved, and the user experience and the satisfaction can be remarkably improved. Particularly, the method and the device embody the remarkable advantages of the invention in the aspects of improving the matching degree of advertisements and user interaction through multi-mode feature fusion and personalized content generation.

It should be noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention may be modified or substituted without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered in the scope of the claims of the present invention.

Claims

1. An advertisement push method based on age and gender recognition, characterized in that: it includes:

Collect user information and perform feature extraction on the collected user information;

Use a multimodal fusion network to perform multimodal fusion on the extracted user features and comprehensively analyze the user's age and gender information;

According to the user's age and gender information, identify the advertising creative that best matches the user's age and gender from the advertising library;

Based on the selected ad creatives, NLP is used to generate personalized ad copy, and GANs is used to generate personalized ad images and videos.

Combine personalized advertising copy with personalized advertising pictures and videos to generate personalized advertising content;

Push personalized advertising content to users and collect user feedback in real time. Optimize advertising content and push strategies based on real-time user feedback.

The multimodal fusion network is used to perform multimodal fusion on the extracted user features and comprehensively analyze the user's age and gender information. The specific steps are as follows:

The extracted facial features, language features and gait features are fused using a multimodal fusion network. The expression is:

in, is the comprehensive feature vector, is the facial feature vector, is the speech feature vector, is the gait feature vector;

Based on the fused comprehensive feature vector F, the comprehensive perception model MLP is used to comprehensively analyze the user's age and gender information;

User age analysis, the expression is:

Among them, A is the predicted value of user age, is the i-th feature in the comprehensive feature vector F, is the amplitude coefficient of the ith feature, is the bias term of the jth layer, is the number of hidden layers, is the dimension of the comprehensive feature vector F, is the activation function, is the global bias term, is a nonlinear activation function, is the phase shift of the i-th eigenvector;

User gender analysis, the expression is:

in, is the frequency coefficient of the ith feature, and G represents the gender bias rate;

The specific steps of identifying the advertising material that best matches the user's age and gender from the advertising library according to the user's age and gender information are as follows:

Based on the product information and creative content, define an ad library containing multiple creatives, each with a target age range and target gender skew associated with it;

Based on the user's age and gender, define a matching function to identify the matching degree between each ad creative and the user's age and gender information. The expression is:

in, represents the kth creative, Represents a creative The average of the target age range, Represents a creative The standard deviation of the target age range, Represents a creative The central value of the target gender bias ratio, Represents a creative The standard deviation of the target gender bias rate, is to measure the creative The degree of match with the user's predicted age value A and gender bias rate G.

2. The method for pushing advertisements based on age and gender recognition according to claim 1, characterized in that: the collecting of user information and feature extraction of the collected user information are performed in the following specific steps:

The user's facial image is captured by a high-resolution camera installed in a public place, the user's voice is recorded by an integrated microphone array, and the user's gait information is detected using a thermal imager and an infrared sensor;

Use facial detection algorithms to detect facial regions from images and extract facial features, including eyes, nose, and mouth features;

Perform noise reduction on the audio signal, use FFT to perform spectrum analysis, and extract speech features, including fundamental frequency and formant frequency;

The gait information is filtered and smoothed, and the Kalman filter is used to remove noise and extract gait features, including step length, stride length, and speed.

3. The method for advertising based on age and gender recognition as claimed in claim 2, characterized in that:

Based on the comprehensive feature vector F, a gender discrimination threshold T _G is defined to determine the user's gender;

When G>T _G , it is judged as female;

When G＜ _TG , the person is judged to be male.

4. The method for advertising based on age and gender recognition as claimed in claim 3, characterized in that:

Based on business needs and historical data, define a matching standard threshold _TM to identify highly matching advertising materials;

when When ≥T _M , the current advertising material is considered to be a high-matching material;

when ＜ _TM , the current advertising material is considered not to be a high-matching material.

5. The method for advertising push based on age and gender recognition as claimed in claim 4 is characterized in that: based on the selected advertising material, NLP is used to generate personalized advertising copy, and GANs is used to generate personalized advertising pictures and videos. The specific steps are as follows:

Extract key information from the selected advertising materials, including product name, features and target audience, to form a set of keywords K;

Generate a copy template according to the type of advertising material and the target audience, fill the relevant information in the keyword set K into the placeholder position in the copy template, and generate preliminary personalized advertising copy;

Use NLP to polish the generated preliminary copy, including synonym replacement and sentence structure adjustment, to generate personalized advertising copy;

According to the content and style of the selected creatives, collect relevant reference images and video clips to train the GANs model;

Use the trained GANs model to generate personalized images and videos that are consistent with the style of the ad creatives.

6. The method for pushing advertisements based on age and gender recognition according to claim 5, characterized in that: the personalized advertisement copy is combined with personalized advertisement pictures and videos to generate personalized advertisement content, and the specific steps are as follows:

By coordinating the consistency between the copy and the image content, the generated personalized advertising copy is matched and combined with the personalized advertising image to enhance the visual appeal;

By embedding the generated personalized advertising copy into personalized videos and presenting the copy content in the form of subtitles and narration, the video's appeal can be improved;

After the combination is completed, use Adobe Premiere Pro to export the final personalized advertising content.

7. The method for pushing advertisements based on age and gender recognition as claimed in claim 6 is characterized in that: the personalized advertisement content is pushed to the user, and user feedback is collected in real time, and the advertisement content and push strategy are optimized according to the user's real-time feedback. The specific steps are as follows:

Based on the user's active time period, the generated personalized advertising content will be pushed to the target users through social media;

Through the built-in feedback collection mechanism, users' responses to ads are collected in real time, including click-through rate, dwell time and number of shares, and subjective feedback from users is directly obtained through questionnaires and online evaluations;

Optimize push timing, frequency and advertising content based on user feedback to increase the effective reach of advertising;

For users with positive feedback, provide more personalized advertising content; for users with negative feedback, optimize push strategies to reduce interference with users' lives.

8. An advertisement push system based on age and gender recognition, based on the advertisement push method based on age and gender recognition according to any one of claims 1 to 7, characterized in that it comprises a data collection and feature extraction module, a multimodal fusion and analysis module, an advertisement material matching module, an advertisement content generation module based on selected advertisement materials, and an advertisement push and feedback module;

The data collection and feature extraction module is used to collect user information and perform feature extraction on the collected user information;

The multimodal fusion and analysis module is used to use the multimodal fusion network to perform multimodal fusion on the extracted user features and comprehensively analyze the user's age and gender information;

An advertisement creative matching module, for identifying, from an advertisement library, advertisement creatives that best match the age and gender of the user based on the age and gender information of the user;

Based on the selected ad creatives: used to generate personalized ad copy using NLP, and use GANs to generate personalized ad images and videos;

An advertising content generation module, used to combine personalized advertising copy with personalized advertising pictures and videos to generate personalized advertising content;

The advertising push and feedback module is used to push personalized advertising content to users, collect user feedback in real time, and optimize advertising content and push strategies based on users' real-time feedback.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein the processor implements the steps of the method for pushing advertisements based on age and gender recognition as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: when the computer program is executed by a processor, the steps of the method for pushing advertisements based on age and gender recognition according to any one of claims 1 to 7 are implemented.