Makeup effect image data processing system based on AI technology

Abstract

The invention belongs to the technical field of makeup effect image data processing, and relates to a makeup effect image data processing system based on an AI technology. By analyzing the similarity between the initial face image of each specified user and each reference initial face image, individuation and customization of a makeup scheme can be ensured, the satisfaction degree of the user and the naturalness degree of the makeup effect are improved, the selection process is simplified, and the user experience is improved. By analyzing the makeup evaluation index of each recommended makeup simulation face image of each specified user, objective and accurate makeup effect evaluation can be provided, the accuracy of a recommended makeup scheme is verified, a personalized makeup recommendation algorithm can be optimized, the recommendation accuracy is improved, and the user experience is improved. By analyzing the makeup favorite degree evaluation index of each recommended makeup simulation face image of each specified user, the preference degree and expectation of the user on the makeup product are accurately known, and the satisfaction degree and loyalty of the user are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of makeup effect image data processing, and relates to a makeup effect image data processing system based on AI technology. Background Art

[0002] In recent years, artificial intelligence technology has made significant progress, especially in the fields of image recognition, deep learning, computer vision, etc. These technological breakthroughs provide strong technical support for the processing of makeup effect image data. With the popularization of the Internet and the rise of e-commerce, the beauty industry has gradually shifted from offline to online. Online shopping platforms, social media, live e-commerce and other channels have become important positions for the sales of beauty products. In order to improve user experience and shopping efficiency, the beauty industry urgently needs a technology that can quickly and accurately display makeup effects. Therefore, the makeup effect image data analysis and processing method based on AI technology is of great significance and role.

[0003] The existing makeup effect image data analysis and processing methods can basically meet the usage requirements, but they still have certain shortcomings: on the one hand, the existing makeup effect image data analysis and processing methods can automatically simulate makeup on the user's face, but the existing makeup effect image data analysis and processing methods lack the formulation and recommendation of makeup plans based on the user's nude makeup facial images, and thus cannot quickly recommend suitable makeup plans to users and cannot simplify the selection process.

[0004] On the other hand, existing methods for analyzing and processing makeup effect image data ignore the analysis of the makeup on the user's face from multiple dimensions such as the makeup effect and the user's personal evaluation, and thus cannot provide an objective and accurate evaluation of the makeup effect, nor is it conducive to users understanding the actual effect of the makeup, and is not conducive to improving user satisfaction and loyalty. Summary of the invention

[0005] In view of this, in order to solve the problems raised in the above background technology, a makeup effect image data processing system based on AI technology is proposed.

[0006] The objective of the present invention can be achieved through the following technical scheme: The present invention provides a makeup effect image data processing system based on AI technology, including: a facial image acquisition module, a facial detection and analysis module, a makeup simulation module, a makeup evaluation feedback module, a user evaluation feedback module, a user preferred makeup acquisition module, an evaluation feedback module and a data storage library.

[0007] The facial image acquisition module is used to record the users as designated users and obtain the initial facial images of the designated users.

[0008] The facial detection and analysis module is used to analyze the similarity between the initial facial image of each designated user and each reference initial facial image, thereby obtaining the reference initial facial image of each designated user, and accordingly obtaining each recommended makeup solution for each designated user.

[0009] The makeup simulation module is used to simulate makeup application for each designated user based on each recommended makeup scheme for each designated user, and obtain a simulated facial image corresponding to each recommended makeup scheme for each designated user.

[0010] The makeup evaluation feedback module is used to compare the simulated facial image corresponding to each recommended makeup solution of each designated user with its corresponding initial facial image, and analyze the makeup evaluation index of the simulated facial image corresponding to each recommended makeup solution of each designated user.

[0011] The user evaluation feedback module is used to obtain user evaluation data of simulated facial images corresponding to each recommended makeup scheme of each designated user, and analyze the makeup likability evaluation index of the simulated facial images corresponding to each recommended makeup scheme of each designated user.

[0012] The user preferred makeup acquisition module is used to acquire each preferred makeup scheme of each designated user based on the makeup evaluation index and makeup likability evaluation index of the simulated facial image corresponding to each recommended makeup scheme of each designated user.

[0013] The evaluation feedback module is used to analyze the user's cosmetics praise rate and provide feedback.

[0014] The data storage library is used to store each reference initial facial image, store each recommended makeup scheme for each reference initial facial image, and store the reference facial contrast deviation, reference facial stereoscopic deviation and reference facial proportion coordination deviation of the simulated facial image corresponding to the makeup scheme and its corresponding initial facial image.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. The present invention obtains each recommended makeup scheme for each designated user by analyzing the similarity between the initial facial image of each designated user and each reference initial facial image, which helps to ensure the personalization and customization of the makeup scheme, helps to improve user satisfaction and the naturalness of the makeup effect, and also quickly recommends suitable makeup schemes to users, simplifying the selection process.

[0016] 2. The present invention helps to provide an objective and accurate evaluation of makeup effects by analyzing the makeup evaluation indicators of simulated facial images corresponding to each recommended makeup scheme of each designated user, which is beneficial for users to understand the actual effects of makeup, verifies the accuracy of the recommended makeup scheme, and further helps to optimize the personalized makeup recommendation algorithm and improve the accuracy of the recommendation.

[0017] 3. The present invention obtains user evaluation data of simulated facial images corresponding to each recommended makeup scheme of each designated user, and analyzes the makeup liking evaluation index of the simulated facial images corresponding to each recommended makeup scheme of each designated user, which helps to accurately understand the user's preference and expectations for makeup products, and also helps to improve user satisfaction and loyalty, which is beneficial to optimize product design and formula, and enhance the market competitiveness of products.

[0018] 4. The present invention helps to provide users with more personalized preferred makeup plans by analyzing the preferred makeup plans of each designated user and the user's makeup praise rate and providing feedback. It is also beneficial to accurately position products, optimize product portfolios, enhance market competitiveness, and promote technological innovation in the makeup industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0020] Figure 1 It is a schematic diagram of system module connection of the present invention.

[0021] Figure 2 It is a schematic diagram of the feature vector acquisition steps of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] See also Figure 1As shown, the present invention provides a makeup effect image data processing system based on AI technology, and the specific modules are distributed as follows: facial image acquisition module, facial detection and analysis module, makeup simulation module, makeup evaluation feedback module, user evaluation feedback module, user preferred makeup acquisition module, evaluation feedback module and data storage library. Among them, the connection mode between the modules is: the facial image acquisition module is connected to the facial detection and analysis module, the facial detection and analysis module is connected to the makeup simulation module, the makeup simulation module is connected to the makeup evaluation feedback module, the makeup evaluation feedback module is connected to the user evaluation feedback module, the user evaluation feedback module is connected to the user preferred makeup acquisition module, the user preferred makeup acquisition module is connected to the evaluation feedback module, and the data storage library is connected to the facial detection and analysis module and the makeup evaluation feedback module respectively.

[0024] The facial image acquisition module is used to record the users as designated users and obtain the initial facial images of the designated users.

[0025] The facial detection and analysis module is used to analyze the similarity between the initial facial image of each designated user and each reference initial facial image, thereby obtaining the reference initial facial image of each designated user, and accordingly obtaining each recommended makeup solution for each designated user.

[0026] As a preferred feasibility example, the similarity between the initial facial image of each designated user and each reference initial facial image is analyzed in a specific manner including: obtaining the feature vector of the initial facial image of each designated user respectively And the feature vectors of each reference initial facial image Wherein a=1,2,...,A, a is the number of each designated user, A is the number of designated users, b=1,2,...,B, b is the number of each reference initial facial image, B is the number of reference initial facial images.

[0027] According to the calculation formula Get the similarity Cos between each reference initial facial image and each specified user's initial facial image ba ,in is the dot product of the feature vector of the bth reference initial facial image and the feature vector of the ath specified user’s initial facial image, are respectively the modulus of the feature vector of the a-th specified user's initial facial image and the modulus of the feature vector of the b-th reference initial facial image.

[0028] As a preferred feasibility example, the feature vectors of the initial facial images of each designated user and the feature vectors of each reference initial facial image are specifically obtained in the following manner: S1, image preprocessing: grayscale and normalize the initial facial images of each designated user and each reference initial facial image respectively to obtain grayscale images of the initial facial images of each designated user and grayscale images of each reference initial facial image.

[0029] S2. Calculate gradient: Calculate the gradient magnitude and direction of each pixel in the grayscale image of each designated user's initial facial image and the grayscale image of each reference initial facial image.

[0030] S3. Divide into cell units: divide the grayscale image of each designated user's initial facial image and the grayscale image of each reference initial facial image into a plurality of small cell units.

[0031] S4. Statistical gradient histogram: a histogram of the statistical gradient direction in each cell unit in the grayscale image of each designated user's initial facial image and the grayscale image of each reference initial facial image.

[0032] S5. Combining into a feature vector: combining the histogram of each cell unit in the grayscale image of each designated user's initial facial image and the grayscale image of each reference initial facial image to form a feature vector of each designated user's initial facial image and a feature vector of each reference initial facial image.

[0033] What needs to be further explained is the schematic diagram of the feature vector acquisition steps as shown in Figure 2 shown.

[0034] As a preferred feasibility example, the reference initial facial images of each designated user are specifically obtained in the following manner: sorting the similarities between each reference initial facial image and the initial facial image of each designated user from large to small, and selecting the reference initial facial images that are ranked first in similarity to the initial facial image of each designated user, and recording them as the reference initial facial images of each designated user.

[0035] As a preferred feasibility example, the specific method of obtaining each recommended makeup scheme for each designated user includes: matching the reference initial facial image of each designated user with the recommended makeup schemes of each reference initial facial image stored in the data repository to obtain the recommended makeup schemes for each designated user.

[0036] The present invention obtains each recommended makeup scheme for each designated user by analyzing the similarity between the initial facial image of each designated user and each reference initial facial image, which helps to ensure the personalization and customization of the makeup scheme, helps to improve user satisfaction and the naturalness of the makeup effect, and also quickly recommends suitable makeup schemes to users, simplifying the selection process.

[0037] The makeup simulation module is used to simulate makeup application for each designated user based on each recommended makeup scheme for each designated user, and obtain a simulated facial image corresponding to each recommended makeup scheme for each designated user.

[0038] The makeup evaluation feedback module is used to compare the simulated facial image corresponding to each recommended makeup solution of each designated user with its corresponding initial facial image, and analyze the makeup evaluation index of the simulated facial image corresponding to each recommended makeup solution of each designated user.

[0039] As a preferred feasibility example, the makeup evaluation index of the simulated facial image corresponding to each recommended makeup scheme of each designated user is specifically analyzed in the following manner: comparing the simulated facial image corresponding to each recommended makeup scheme of each designated user with its corresponding initial facial image, thereby obtaining the facial contrast deviation, facial stereoscopic deviation and facial proportion coordination deviation between the simulated facial image corresponding to each recommended makeup scheme of each designated user and its corresponding initial facial image, which are respectively recorded as FaceCR ad 、FaceSD ad 、FacePC ad , where a=1,2,...,A, a is the number of each designated user, A is the number of designated users, d=1,2,...,D, d is the number of each recommended makeup scheme, and D is the number of recommended makeup schemes.

[0040] Analyze the makeup evaluation index of each recommended makeup scheme for each specified user corresponding to the simulated facial image

[0041] Among them, FaceCR0, FaceSD0, and FacePC0 are the reference facial contrast deviation, reference facial stereo deviation, and reference facial proportion coordination deviation of the simulated facial image corresponding to the makeup scheme extracted from the information reserve and its corresponding initial facial image, respectively; ΔFaceCR0, ΔFaceSD0, and ΔFacePC0 are the allowable differences between the set facial contrast deviation and the reference facial contrast deviation, the allowable differences between the facial stereo deviation and the reference facial stereo deviation, and the allowable differences between the facial proportion coordination deviation and the reference facial proportion coordination deviation, respectively.

[0042] It should be further explained that the set permissible difference between the facial contrast deviation and the reference facial contrast deviation, the permissible difference between the facial stereoscopic deviation and the reference facial stereoscopic deviation, and the permissible difference between the facial proportion coordination deviation and the reference facial proportion coordination deviation can be set to 0.01, 0.01, and 0.01, respectively. This means that the facial contrast deviation and the reference facial contrast deviation, the facial stereoscopic deviation and the reference facial stereoscopic deviation, and the facial proportion coordination deviation and the reference facial proportion coordination deviation can all allow a deviation of 1%.

[0043] The present invention helps to provide an objective and accurate evaluation of makeup effects by analyzing the makeup evaluation indicators of simulated facial images corresponding to each recommended makeup scheme of each designated user, which is beneficial for users to understand the actual effects of makeup, verifies the accuracy of the recommended makeup scheme, and further helps to optimize the personalized makeup recommendation algorithm and improve the accuracy of the recommendation.

[0044] The user evaluation feedback module is used to obtain user evaluation data of simulated facial images corresponding to each recommended makeup scheme of each designated user, and analyze the makeup likability evaluation index of the simulated facial images corresponding to each recommended makeup scheme of each designated user.

[0045] As a preferred feasibility example, the user evaluation data of the simulated facial image corresponding to each recommended makeup scheme of each designated user includes a preference value score and an expected value score.

[0046] It needs to be further explained that the specific method of obtaining the preference value score and the expected value score of the simulated facial image corresponding to each recommended makeup scheme of each designated user is: designing a questionnaire to let each designated user score the preference value and the expected value of the simulated facial image corresponding to each recommended makeup scheme, and then obtaining the preference value score and the expected value score of the simulated facial image corresponding to each recommended makeup scheme of each designated user.

[0047] As a preferred feasibility example, the makeup liking evaluation index of each recommended makeup scheme of each designated user corresponding to the simulated facial image is specifically analyzed in the following manner: extracting the liking value score and expected value score of each recommended makeup scheme of each designated user corresponding to the simulated facial image, which are recorded as Analyze the makeup likability evaluation index of each recommended makeup scheme for each specified user corresponding to the simulated facial image in are the preference value scoring threshold and expected value scoring threshold of the set simulated facial images respectively.

[0048] In a specific example, the preference value score and the expected value score may be 10 points and 10 points respectively, and the preference value score threshold and the expected value score threshold of the simulated facial image may be set to 7 points and 7 points respectively.

[0049] The present invention obtains user evaluation data of simulated facial images corresponding to each recommended makeup scheme of each designated user, and analyzes the makeup liking evaluation index of the simulated facial images corresponding to each recommended makeup scheme of each designated user, which helps to accurately understand the user's preference and expectations for makeup products, and also helps to improve user satisfaction and loyalty, which is beneficial to optimize product design and formula, and enhance the market competitiveness of products.

[0050] The user preferred makeup acquisition module is used to acquire each preferred makeup scheme of each designated user based on the makeup evaluation index and makeup likability evaluation index of the simulated facial image corresponding to each recommended makeup scheme of each designated user.

[0051] As a preferred feasibility example, the specific method of obtaining each preferred makeup scheme of each designated user includes: according to the analysis formula Get the comprehensive evaluation coefficient ξ of each recommended makeup solution for each specified user ad , where β1 and β2 are the weight factors of the comprehensive evaluation coefficients corresponding to the set makeup evaluation index and makeup liking evaluation index, respectively.

[0052] It should be further explained that the weight factors of the comprehensive evaluation coefficients corresponding to the set makeup evaluation index and makeup liking evaluation index can be set to 0.6 and 0.4 respectively.

[0053] Makeup effect evaluation indicators are usually based on professional image analysis technology and aesthetic standards, and can objectively measure the performance of makeup in terms of color matching, contour shaping, clarity, etc. These indicators have a certain degree of scientificity and accuracy, and can provide a relatively reliable evaluation of the quality and effect of makeup. Indicators such as contrast, three-dimensionality, proportion coordination and clarity in the makeup effect evaluation indicators can be determined through specific numerical calculations and analysis, and are not affected by personal subjective preferences. Therefore, assigning a weight factor of 0.6 to the comprehensive evaluation coefficient corresponding to the makeup evaluation index is conducive to ensuring that the comprehensive evaluation results have a certain degree of professionalism and objectivity.

[0054] The makeup likability evaluation index reflects consumers' demand and satisfaction with makeup. Although the makeup effect evaluation index can objectively measure the quality and effect of makeup, the ultimate goal is to meet consumers' needs and preferences. Therefore, the weight factor of the makeup likability evaluation index corresponding to the comprehensive evaluation coefficient is assigned to 0.4.

[0055] The comprehensive evaluation coefficients of the recommended makeup schemes of each designated user are compared with the set comprehensive evaluation coefficient thresholds, and the recommended makeup schemes of each designated user whose comprehensive evaluation coefficients are greater than or equal to the comprehensive evaluation coefficient thresholds are screened out and recorded as the preferred makeup schemes of each designated user.

[0056] The evaluation feedback module is used to analyze the user's cosmetics praise rate and provide feedback.

[0057] As a preferred feasibility example, the specific analysis method of the user's makeup favorable rating includes: counting the preferred makeup schemes of each designated user to obtain the number of preferred makeup schemes of each designated user.

[0058] The makeup liking evaluation index of the simulated facial image corresponding to each recommended makeup scheme of each designated user is extracted, and compared with the set makeup liking evaluation index threshold value respectively; if the makeup liking evaluation index of the simulated facial image corresponding to a recommended makeup scheme of a designated user is greater than or equal to the makeup liking evaluation index threshold value, then the recommended makeup scheme of the designated user is recorded as the designated user's favorite recommended makeup scheme, and then the favorite recommended makeup schemes of each designated user are obtained, and further statistics are performed to obtain the number of favorite recommended makeup schemes of each designated user.

[0059] The number of preferred makeup solutions of each designated user is added to the number of recommended makeup solutions that the user likes, and the sum is recorded as the number of recommended makeup solutions m for each designated user. a ;

[0060] Analyze users’ favorable reviews of cosmetics And provide feedback.

[0061] The data storage library is used to store each reference initial facial image, store each recommended makeup scheme for each reference initial facial image, and store a reference facial contrast deviation, a reference facial stereoscopic deviation, and a reference facial proportion coordination deviation between a simulated facial image corresponding to the makeup scheme and its corresponding initial facial image.

[0062] The present invention helps to provide users with more personalized preferred makeup plans by analyzing the preferred makeup plans of each designated user and the user's makeup praise rate and providing feedback. It is also beneficial to accurately position products, optimize product portfolios, enhance market competitiveness, and promote technological innovation in the makeup industry.

[0063] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined by the present invention, they shall all fall within the protection scope of the present invention.

Claims

1. A makeup effect image data processing system based on AI technology, characterized by: include: A facial image acquisition module is used to record the used users as designated users and obtain the initial facial images of the designated users; A facial detection and analysis module is used to analyze the similarity between the initial facial image of each designated user and each reference initial facial image, thereby obtaining the reference initial facial image of each designated user, and accordingly obtaining each recommended makeup solution for each designated user; A makeup simulation module, used to simulate makeup application for each designated user based on each recommended makeup scheme for each designated user, and obtain a simulated facial image corresponding to each recommended makeup scheme for each designated user; A makeup evaluation feedback module is used to compare the simulated facial image corresponding to each recommended makeup solution of each designated user with its corresponding initial facial image, and analyze the makeup evaluation index of the simulated facial image corresponding to each recommended makeup solution of each designated user; A user evaluation feedback module is used to obtain user evaluation data of each recommended makeup scheme corresponding to the simulated facial image of each designated user, and analyze the makeup likability evaluation index of each recommended makeup scheme corresponding to the simulated facial image of each designated user; A user preferred makeup acquisition module, for acquiring each preferred makeup scheme of each designated user based on the makeup evaluation index and makeup likability evaluation index of each recommended makeup scheme of each designated user corresponding to the simulated facial image; Evaluation and feedback module, used to analyze the user's cosmetics praise rate and provide feedback; The data storage library is used to store each reference initial facial image, store each recommended makeup scheme for each reference initial facial image, and store the reference facial contrast deviation, reference facial stereoscopic deviation and reference facial proportion coordination deviation of the simulated facial image corresponding to the makeup scheme and its corresponding initial facial image.

2. The makeup effect image data processing system based on AI technology according to claim 1, characterized in that: The similarity between the initial facial image of each designated user and each reference initial facial image is analyzed in the following specific manner: Get the feature vector of the initial facial image of each specified user respectively And the feature vectors of each reference initial facial image Wherein a=1,2,...,A, a is the number of each designated user, A is the number of designated users, b=1,2,...,B, b is the number of each reference initial facial image, B is the number of reference initial facial images; According to the calculation formula Get the similarity Cos between each reference initial facial image and each specified user's initial facial image ba ,in is the dot product of the feature vector of the bth reference initial facial image and the feature vector of the ath specified user’s initial facial image, are respectively the modulus of the feature vector of the a-th specified user's initial facial image and the modulus of the feature vector of the b-th reference initial facial image.

3. The makeup effect image data processing system based on AI technology according to claim 2, characterized in that: The specific method of acquiring the feature vector of each designated user's initial facial image and the feature vector of each reference initial facial image includes: S1, image preprocessing: grayscale and normalize the initial facial image of each designated user and each reference initial facial image to obtain a grayscale image of the initial facial image of each designated user and a grayscale image of each reference initial facial image; S2, calculating gradient: calculating the gradient amplitude and direction of each pixel in the grayscale image of the initial facial image of each designated user and the grayscale image of each reference initial facial image; S3, dividing into cell units: dividing the grayscale image of each designated user's initial facial image and the grayscale image of each reference initial facial image into a plurality of small cell units; S4, statistical gradient histogram: a histogram of the statistical gradient direction in each cell unit in the grayscale image of each designated user's initial facial image and the grayscale image of each reference initial facial image; S5. Combining into a feature vector: combining the histogram of each cell unit in the grayscale image of each designated user's initial facial image and the grayscale image of each reference initial facial image to form a feature vector of each designated user's initial facial image and a feature vector of each reference initial facial image.

4. The makeup effect image data processing system based on AI technology according to claim 2, characterized in that: The specific method of obtaining the reference initial facial image of each designated user includes: The similarities between the reference initial facial images and the initial facial images of the designated users are sorted from large to small, and the reference initial facial images with the highest similarity to the initial facial images of the designated users are selected and recorded as the reference initial facial images of the designated users.

5. The makeup effect image data processing system based on AI technology according to claim 4, characterized in that: The specific method of obtaining the recommended makeup solutions of each designated user includes: The reference initial facial images of each designated user are matched with the recommended makeup schemes of each reference initial facial image stored in the data storage library to obtain the recommended makeup schemes of each designated user.

6. The makeup effect image data processing system based on AI technology according to claim 1, characterized in that: The makeup evaluation index of the simulated facial image corresponding to each recommended makeup scheme of each designated user is specifically analyzed in the following manner: The simulated facial image corresponding to each recommended makeup scheme of each designated user is compared with its corresponding initial facial image, so as to obtain the facial contrast deviation, facial stereoscopic deviation and facial proportion coordination deviation of the simulated facial image corresponding to each recommended makeup scheme of each designated user and its corresponding initial facial image, which are respectively denoted as FaceCR ad 、FaceSD ad 、FacePC ad , where a=1,2,...,A, a is the number of each designated user, A is the number of designated users, d=1,2,...,D, d is the number of each recommended makeup solution, D is the number of recommended makeup solutions; Analyze the makeup evaluation index of each recommended makeup scheme for each specified user corresponding to the simulated facial image Among them, FaceCR0, FaceSD0, and FacePC0 are the reference facial contrast deviation, reference facial stereo deviation, and reference facial proportion coordination deviation of the simulated facial image corresponding to the makeup scheme extracted from the information reserve and its corresponding initial facial image, respectively; ΔFaceCR0, ΔFaceSD0, and ΔFacePC0 are the allowable differences between the set facial contrast deviation and the reference facial contrast deviation, the allowable differences between the facial stereo deviation and the reference facial stereo deviation, and the allowable differences between the facial proportion coordination deviation and the reference facial proportion coordination deviation, respectively.

7. The makeup effect image data processing system based on AI technology according to claim 6, characterized in that: The user evaluation data of the simulated facial image corresponding to each recommended makeup scheme of each designated user includes a preference value score and an expected value score.

8. The makeup effect image data processing system based on AI technology according to claim 7, characterized in that: The makeup likability evaluation index of the simulated facial image corresponding to each recommended makeup scheme of each designated user is specifically analyzed in the following manner: Extract the preference value score and expected value score of each recommended makeup scheme corresponding to the simulated facial image of each specified user, which are recorded as Analyze the makeup likability evaluation index of each recommended makeup scheme for each specified user corresponding to the simulated facial image in are the preference value scoring threshold and expected value scoring threshold of the set simulated facial images respectively.

9. The makeup effect image data processing system based on AI technology according to claim 8, characterized in that: The specific method of obtaining the preferred makeup schemes of the designated users includes: According to the analytical formula Get the comprehensive evaluation coefficient ξ of each recommended makeup solution for each specified user ad , where β1 and β2 are the weight factors of the comprehensive evaluation coefficients corresponding to the set makeup evaluation index and makeup likability evaluation index respectively; The comprehensive evaluation coefficients of the recommended makeup schemes of each designated user are compared with the set comprehensive evaluation coefficient thresholds, and the recommended makeup schemes of each designated user whose comprehensive evaluation coefficients are greater than or equal to the comprehensive evaluation coefficient thresholds are screened out and recorded as the preferred makeup schemes of each designated user.

10. The makeup effect image data processing system based on AI technology according to claim 9, characterized in that: The specific analysis method of the user's cosmetics favorable comment rate includes: Counting the preferred makeup plans of each designated user to obtain the number of preferred makeup plans of each designated user; Extracting makeup likeness evaluation indexes of simulated facial images corresponding to each recommended makeup scheme of each designated user, and comparing them with set makeup likeness evaluation index thresholds respectively; if the makeup likeness evaluation index of a simulated facial image corresponding to a recommended makeup scheme of a designated user is greater than or equal to the makeup likeness evaluation index threshold, then recording the recommended makeup scheme of the designated user as the designated user's favorite recommended makeup scheme, thereby obtaining each favorite recommended makeup scheme of each designated user, and further performing statistics on them to obtain the number of favorite recommended makeup schemes of each designated user; The number of preferred makeup solutions of each designated user is added to the number of recommended makeup solutions that the user likes, and the sum is recorded as the number of recommended makeup solutions m for each designated user. a ; Analyze users’ favorable reviews of cosmetics And provide feedback.

Images