Integrated costume style recommendation method based on team requirements

Abstract

The invention provides an integrated clothing style recommendation method based on team demands, and belongs to the technical field of intelligent recommendation, and the method comprises the steps: carrying out the analysis of the team demands based on a team theme, and obtaining a global parameter set and a local parameter set surrounding the team theme, collecting a first clothing recommendation set matched with each global parameter in the global parameter set; mapping the local parameter set to the outline framework of each individual member, and obtaining a second clothing recommendation set for each individual member; and according to the first clothing recommendation set and all the second clothing recommendation sets, clothing integrated processing is carried out to obtain a clothing recommendation result, team experience is improved on the basis of effectively meeting requirements, and matching efficiency is effectively improved through a series of intelligent recommendations.

Description

Technical Field

[0001] The present invention relates to the field of intelligent recommendation technology, and in particular to an integrated clothing style recommendation method based on team needs. Background Art

[0002] When a team performs, in addition to their own strong abilities, the overall clothing matching is also an important factor in creating stage effects. In many cases, the clothing matching needs to be planned and implemented by the team stylists themselves, from the fabrics of the clothes to the shapes of the clothes. In addition, the rationality of the matching needs to be repeatedly adjusted manually during the matching process. This will undoubtedly waste a lot of time and reduce the experience and matching efficiency of meeting the needs of the team.

[0003] Therefore, the present invention proposes an integrated clothing style recommendation method based on team needs. Summary of the invention

[0004] The present invention provides an integrated clothing style recommendation method based on team needs, which is used to determine a global parameter set and a local parameter set in combination with the team theme and team needs, providing a data basis for subsequent clothing recommendations, and by obtaining clothing recommendation sets of different parameter sets and integrated processing of the clothing recommendation sets, the team experience is improved on the basis of effectively meeting the needs, and the matching efficiency is effectively improved through a series of intelligent recommendations.

[0005] The present invention provides an integrated clothing style recommendation method based on team needs, comprising: Step 1: Analyze team needs based on team theme to obtain a global parameter set and a local parameter set around the team theme, wherein the global parameter set is related to the overall clothing style of the team, and the local parameter set is related to the clothing style of individuals in the team; Step 2: Collect the first clothing recommendation set matching each global parameter in the global parameter set; Step 3: Mapping the local parameter set to the self-contour framework of each individual member, and obtaining a second clothing recommendation set for each individual member; Step 4: Perform clothing integration processing based on the first clothing recommendation set and all second clothing recommendation sets to obtain clothing recommendation results, wherein the clothing recommendation results include the overall dress of the team and the individual dress of each member.

[0006] Preferably, the team requirements are parsed based on the team theme to obtain a global parameter set and a local parameter set around the team theme, including: Based on the team theme, a demand analysis model is obtained by matching from a theme-model comparison table; Analyze the team requirements according to the requirements analysis model to obtain a plurality of first initial parameters and a plurality of second initial parameters; Based on the required role of each first initial parameter in the parsing process, an initial radius is set to the corresponding first initial parameter; Setting a first weight based on the first initial parameters corresponding to all initial radii; At the same time, an equal weight is assigned to each second initial parameter as a second weight; Counting all first initial parameters set with first weights as a global parameter set; All second initial parameters set with second weights are counted as a local parameter set.

[0007] Preferably, setting the first weight based on the first initial parameters corresponding to all the initial radii includes: Determine the circumference and the circumference area of ​​the circle corresponding to the first initial parameter based on the set initial radius; Performing a first normal distribution process on all the circumferences of the circles to obtain a first radius, and at the same time, performing a second normal distribution process on all the circumference areas to obtain a second radius; Selecting a maximum radius and a minimum radius from all initial radii, and setting a radius threshold based on a difference between the maximum radius and the minimum radius and in combination with the first radius and the second radius; ; in, Indicates the corresponding radius threshold; Respectively represent the average values ​​based on all first radii and all second radii; Respectively represent the maximum radius and the minimum radius; Respectively represent the number of all first radii and the number of all second radii; represents the radius fine-tuning function; Performing intersection processing on the first initial quantity of the long side radius outside the first normal distribution and the first initial quantity of the long side radius outside the second normal distribution to obtain an intersection initial parameter; If the number of the intersection initial parameters is 0, setting a first importance value for each first initial parameter according to a difference between an initial radius of each first initial parameter and a radius threshold; ; in, represents the first important value of the i1th first initial parameter; represents the initial radius of the i1th first initial parameter; If the number of the intersection initial parameters is not 0, then according to the number of long side radii corresponding to each first initial parameter and in combination with the difference between the initial radius of each first initial parameter and the radius threshold, a second important value is set for each first initial parameter; ; in, represents the second important value of the i1th first initial parameter; Indicates that the initial radius corresponding to the i1th first initial parameter belongs to the number of long side radii; Indicates the number of parameters whose initial radius belongs to the long side radius among all the first initial parameters; ln represents the sign of the logarithmic function; Based on the importance values ​​of all first initial parameters, first weights are set to corresponding first initial parameters.

[0008] Preferably, collecting the first clothing recommendation set matching each global parameter in the global parameter set includes: According to the parameter attribute of the global parameter, a first clothing recommendation set matching the parameter attribute is matched from an attribute-clothing database.

[0009] Preferably, mapping the local parameter set to the self-profile framework of each individual member comprises: Determine the action position of each local parameter in the local parameter set, compare the action position with the self-contour framework, and combine the initial setting performance result of each local parameter to obtain the initial local sequence corresponding to the individual member; Obtain the outstanding advantages and convertible advantages of each individual member's own profile structure, and adjust the initial local sequence in combination with the outstanding characteristics of the team theme; The second clothing recommendation set corresponding to the individual member is obtained by adjusting the local sequence.

[0010] Preferably, adjusting the initial local sequence includes: Acquire the transferable and outstanding advantages of each individual member based on their own profile framework; Determine a first imbalance coefficient corresponding to the convertible advantage of the individual member, and at the same time, determine a second imbalance coefficient corresponding to the outstanding advantage of the individual member; ; in, represents the first imbalance coefficient corresponding to the j1th convertible advantage of the individual member; represents the second imbalance coefficient corresponding to the j2th prominent advantage of the individual member; It indicates the number of times the historical costumes corresponding to the j1-th convertible advantage of a single member are recommended and adjusted successfully; represents the total number of historical occurrences consistent with the j1-th convertible advantage of the corresponding individual member; represents the number of failed historical recommended adjustments consistent with the j1th convertible advantage of the corresponding individual member; It represents the advantage prominence coefficient corresponding to the j1th convertible advantage of the individual member; It represents the advantage prominence coefficient corresponding to the j2th prominence of the individual member; Indicates the total number of convertible advantages involved in the corresponding individual member; It represents the total number of convertible advantages involved by all individual members; The variance of the advantage salience coefficient representing the convertible advantages involved by all individual members; Extracting the adaptation factor based on each action position from the first imbalance factors and the second imbalance factors of all individual members, and calculating the overall adaptation value of the corresponding action position; If the overall adaptation value is greater than the preset adaptation value of the corresponding action position, the corresponding sequence is adjusted according to the first imbalance factor and the second imbalance factor involved in the corresponding action position and in combination with the prominent characteristics of the team theme to achieve partial clothing adjustment; Otherwise, the sequence of the corresponding action positions is not adjusted; According to the final result of each action position, an adjusted local sequence is obtained.

[0011] Preferably, obtaining a second clothing recommendation set corresponding to a single member by adjusting the local sequence includes: comparing the adjusted local sequence with a personalized clothing database; According to the comparison result, a second clothing recommendation set corresponding to the individual member is obtained.

[0012] Preferably, performing clothing integration processing according to the first clothing recommendation set and all second clothing recommendation sets to obtain clothing recommendation results includes: Inputting the first clothing recommendation set and each second clothing recommendation set into a team clothing analysis table in sequence, wherein the team clothing analysis table includes overall clothing elements and unique clothing elements of different individual members; Count the coordination between each column's unique clothing element and each overall clothing element; If all coordination is greater than the corresponding preset, the corresponding unique clothing elements will be retained; Otherwise, the elements whose coordination is less than the preset among all the unique clothing elements of each individual member are locked and regarded as the first elements, and the elements to be adjusted and the personnel to be adjusted are screened by counting the first number of the first elements involved in each individual member and counting the second number of the first elements involved in each column of unique clothing elements; Weakening the local parameter set of the personnel to be adjusted according to the global parameter set and in combination with the division of labor roles of the personnel to be adjusted; Performing information standardization processing on the element information of the unique clothing element in the column corresponding to the element to be adjusted according to the clothing set with the maximum recommendation coefficient in the first clothing recommendation set; According to the weakening processing results and the information standardization processing results, the clothing recommendation results are obtained.

[0013] Compared with the prior art, the present invention has the following beneficial effects: Based on the team theme and team needs, the global parameter set and the local parameter set are determined to provide a data basis for subsequent clothing recommendations. By obtaining clothing recommendation sets with different parameter sets and integrating the clothing recommendation sets, the team experience is improved on the basis of effectively meeting the needs, and the matching efficiency is effectively improved through a series of intelligent recommendations.

[0014] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0015] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 The present invention is a flowchart of an integrated clothing style recommendation method based on team needs in an embodiment of the present invention. DETAILED DESCRIPTION

[0017] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0018] The present invention provides an integrated clothing style recommendation method based on team needs, such as Figure 1 As shown, including: Step 1: Analyze team needs based on team theme to obtain a global parameter set and a local parameter set around the team theme, wherein the global parameter set is related to the overall clothing style of the team, and the local parameter set is related to the clothing style of individuals in the team; Step 2: Collect the first clothing recommendation set matching each global parameter in the global parameter set; Step 3: Mapping the local parameter set to the self-contour framework of each individual member, and obtaining a second clothing recommendation set for each individual member; Step 4: Perform clothing integration processing based on the first clothing recommendation set and all second clothing recommendation sets to obtain clothing recommendation results, wherein the clothing recommendation results include the overall dress of the team and the individual dress of each member.

[0019] In this embodiment, the team theme may be a chorus theme, a drama theme, or other related performance themes.

[0020] In this embodiment, team demand refers to the performance effect demand, atmosphere demand, etc. proposed by the organizing teacher or team members for chorus, drama, etc.

[0021] In this embodiment, the global parameter set is constructed based on the overall clothing style of the team, for example, clothing fabric, clothing color, clothing style, etc. of the entire team.

[0022] The local parameter set is constructed based on the personal style of individual members in the team, such as the individual structure of each member's clothing cut, clothing accessories, etc.

[0023] In this embodiment, the first clothing recommendation set includes clothing sample recommendation results related to clothing fabrics, clothing colors, and clothing styles.

[0024] In this embodiment, the self-contour framework refers to the 3D body model of the corresponding member.

[0025] In this embodiment, the local parameter set is mapped to the corresponding self-contour framework for personalized design.

[0026] In this embodiment, the second clothing recommendation set includes personalized clothing recommendations corresponding to a single member, that is, the advantages of the member are highlighted as much as possible.

[0027] In this embodiment, integrated processing means that the advantages of each member can be highlighted under the premise of meeting the overall theme, and the overall clothing style of the team can be not affected in the process of highlighting the advantages.

[0028] In this embodiment, the clothing recommendation results can be clothing-related parameters such as clothing shape and size, which can be input into the clothing model to obtain AI virtual clothing effects for the member, and then subsequent cutting and adjustment can be performed again to obtain targeted clothing for the member, and then generate a nesting diagram.

[0029] The beneficial effects of the above technical solution are: determining the global parameter set and the local parameter set in combination with the team theme and team needs, providing a data basis for subsequent clothing recommendations, and by obtaining clothing recommendation sets of different parameter sets and integrated processing of clothing recommendation sets, improving the team experience on the basis of effectively meeting needs, and effectively improving matching efficiency through a series of intelligent recommendations.

[0030] The present invention provides an integrated clothing style recommendation method based on team needs, which analyzes team needs based on team themes to obtain a global parameter set and a local parameter set around the team theme, including: Based on the team theme, a demand analysis model is obtained by matching from a theme-model comparison table; Analyze the team requirements according to the requirements analysis model to obtain a number of first initial parameters and a number of second initial parameters; Based on the required role of each first initial parameter in the parsing process, an initial radius is set to the corresponding first initial parameter; Setting a first weight based on first initial parameters corresponding to all initial radii; At the same time, an equal weight is assigned to each second initial parameter as a second weight; Counting all first initial parameters set with first weights as a global parameter set; All second initial parameters set with second weights are counted as a local parameter set.

[0031] In this embodiment, the team theme refers to the theme that the team needs to perform, for example, the chorus theme of "Green Tornado", and then the demand analysis model related to the chorus theme is matched from the theme-model comparison table, and the table includes different team themes and the demand analysis models corresponding to the themes, and the demand analysis model is based on the requirements of different team themes for the theme and the required matching parameters for the requirement as samples to train the neural network model.

[0032] In this embodiment, the first initial parameter is for the team as a whole, and the second initial parameter is for individual members of the team.

[0033] In this embodiment, the demand effect refers to the situation where the corresponding initial parameter is applied to team clothing. The more irreplaceable the application situation is, the greater its effect is.

[0034] In this embodiment, the initial radius is obtained based on matching of an action-radius comparison table, and the comparison table includes different required actions and initial radii that match the actions, and the value range of the initial radius is 0 to 1 cm.

[0035] In this embodiment, the global parameter set includes first initial parameters set with first weights.

[0036] In this embodiment, the local parameter set includes second initial parameters set with second weights.

[0037] The beneficial effect of the above technical solution is: by obtaining the model related to the subject from the comparison table to realize the analysis of the demand, and then by obtaining the role of the parameter in the analysis process, the corresponding weight is set for the parameter, and then the global parameter and the local parameter are obtained.

[0038] The present invention provides an integrated clothing style recommendation method based on team needs, which sets a first weight based on first initial parameters corresponding to all initial radii, including: Determine the circumference and the circumference area of ​​the circle corresponding to the first initial parameter based on the set initial radius; Performing a first normal distribution process on all the circumferences of the circles to obtain a first radius, and at the same time, performing a second normal distribution process on all the circumference areas to obtain a second radius; Selecting a maximum radius and a minimum radius from all initial radii, and setting a radius threshold based on a difference between the maximum radius and the minimum radius and in combination with the first radius and the second radius; ; in, Indicates the corresponding radius threshold; Respectively represent the average values ​​based on all first radii and all second radii; Respectively represent the maximum radius and the minimum radius; Respectively represent the number of all first radii and the number of all second radii; represents the radius fine-tuning function; Performing intersection processing on the first initial quantity of the long side radius outside the first normal distribution and the first initial quantity of the long side radius outside the second normal distribution to obtain an intersection initial parameter; If the number of the intersection initial parameters is 0, setting a first importance value for each first initial parameter according to a difference between an initial radius of each first initial parameter and a radius threshold; ; in, represents the first important value of the i1th first initial parameter; represents the initial radius of the i1th first initial parameter; If the number of the intersection initial parameters is not 0, then according to the number of long side radii corresponding to each first initial parameter and in combination with the difference between the initial radius of each first initial parameter and the radius threshold, a second important value is set for each first initial parameter; ; in, represents the second important value of the i1th first initial parameter; Indicates that the initial radius corresponding to the i1th first initial parameter belongs to the number of long side radii; Indicates the number of parameters whose initial radius belongs to the long side radius among all the first initial parameters; ln represents the sign of the logarithmic function; Based on the importance values ​​of all first initial parameters, first weights are set to corresponding first initial parameters.

[0039] In this embodiment, the calculation formula for the circumference of a circle is: , the calculation formula for the circumference area is: .

[0040] In this embodiment, normal distribution processing refers to satisfying the normal distribution probability, wherein the normal distribution probability of the first normal distribution processing is 80%, and the normal distribution probability of the second normal distribution processing is 70%. Therefore, other perimeter and area distribution positions that do not satisfy the normal distribution probability include two parts, one is the position of small radius, that is, the position that is smaller than the minimum radius within the normal distribution probability distribution, and the other is the position of large radius, that is, the position that is larger than the maximum radius within the normal distribution probability distribution.

[0041] In this embodiment, for example, the first initial quantities of the long side radius outside the first normal distribution are: u1, u2, and the first initial quantities of the long side radius outside the second normal distribution are: u1, u2, u3. At this time, the existing intersection initial parameters are: u1, u2.

[0042] In this embodiment, the first weight=importance value corresponding to the first initial parameter / sum of importance values ​​of all first initial parameters.

[0043] The beneficial effects of the above technical solution are: starting from the demand effect to determine the perimeter and area, and then combining the normal distribution processing to set the radius threshold, and determining the important values ​​of the initial parameters under different intersection conditions through the intersection processing results, thereby providing a convenient basis for setting the first weight.

[0044] The present invention provides an integrated clothing style recommendation method based on team needs, collecting a first clothing recommendation set matching each global parameter in the global parameter set, including: According to the parameter attribute of the global parameter, a first clothing recommendation set matching the parameter attribute is matched from an attribute-clothing database.

[0045] In this embodiment, the parameter attribute may be fabric, color, style, etc.

[0046] In this embodiment, the attribute-clothing database includes a combination of different parameter attributes and clothing recommendation results matching the combination, and there is more than one clothing recommendation result, so a first clothing recommendation set is constructed, and the first clothing recommendation set includes at least one clothing recommendation result.

[0047] The beneficial effect of the above technical solution is: by obtaining attributes to compare and analyze with the database, the first clothing recommendation set is directly obtained.

[0048] The present invention provides an integrated clothing style recommendation method based on team needs, mapping the local parameter set to the self-profile framework of each individual member, including: Determine the action position of each local parameter in the local parameter set, compare the action position with the self-contour framework, and combine the initial setting performance result of each local parameter to obtain the initial local sequence corresponding to the individual member; Obtain the outstanding advantages and convertible advantages of each individual member's own profile structure, and adjust the initial local sequence in combination with the outstanding characteristics of the team theme; The second clothing recommendation set corresponding to the individual member is obtained by adjusting the local sequence.

[0049] In this embodiment, the action position refers to the decorative position of the local parameter on the relevant frame. For example, a personal pendant needs to be set at the waist of the frame, or a scarf needs to be worn around the neck, which can be a cotton scarf, a silk scarf, etc.

[0050] In this embodiment, the initial setting of the performance result refers to the use of an ideal effect for matching and displaying first, but whether the ideal effect is used by the person himself still needs to be analyzed again, that is, the initial local sequence is adjusted, wherein the ideal effect is predetermined, generally the ideal effect that can be determined by using a standard framework, which can be preliminarily determined by software.

[0051] In this embodiment, to highlight advantages, such as being tall, these members can be equipped with light strip clothing and head accessories that create an atmosphere to boost the overall atmosphere.

[0052] In this embodiment, the advantage can be converted, for example, if the legs are short, the defect of short legs can be concealed by wearing clothing that covers the legs.

[0053] In this embodiment, the initial local sequence: {the initial setting performance result of each local parameter involved in the corresponding individual member}.

[0054] In this embodiment, adjusting the local sequence: {adjusting each sequence in the initial local sequence to obtain the fitting setting performance result}.

[0055] In this embodiment, the prominent characteristics of the team theme refer to the performance characteristics that the team needs to highlight, for example, the "Green Tornado" chorus is aimed at the characteristics of special forces and unity and cooperation.

[0056] The beneficial effect of the above technical solution is that by comparing and analyzing the local parameters with the framework, the initial local sequence can be effectively obtained, and then the sequence can be adjusted from the three aspects of highlighting advantages, convertible advantages and highlighting characteristics to effectively obtain the second clothing recommendation set.

[0057] The present invention provides an integrated clothing style recommendation method based on team needs, which adjusts the initial local sequence, including: Acquire the transferable and outstanding advantages of each individual member based on their own profile framework; Determine a first imbalance coefficient corresponding to the convertible advantage of the individual member, and at the same time, determine a second imbalance coefficient corresponding to the outstanding advantage of the individual member; ; ; in, represents the first imbalance coefficient corresponding to the j1th convertible advantage of the individual member; represents the second imbalance coefficient corresponding to the j2th prominent advantage of the individual member; It indicates the number of times the historical costumes corresponding to the j1-th convertible advantage of a single member are recommended and adjusted successfully; represents the total number of historical occurrences consistent with the j1-th convertible advantage of the corresponding individual member; represents the number of failed historical recommended adjustments consistent with the j1th convertible advantage of the corresponding individual member; It represents the advantage prominence coefficient corresponding to the j1th convertible advantage of the individual member; It represents the advantage prominence coefficient corresponding to the j2th prominence of the individual member; Indicates the total number of convertible advantages involved in the corresponding individual member; It represents the total number of convertible advantages involved by all individual members; The variance of the advantage salience coefficient representing the convertible advantages involved by all individual members; Extracting the adaptation factor based on each action position from the first imbalance factors and the second imbalance factors of all individual members, and calculating the overall adaptation value of the corresponding action position; If the overall adaptation value is greater than the preset adaptation value of the corresponding action position, the corresponding sequence is adjusted according to the first imbalance factor and the second imbalance factor involved in the corresponding action position and in combination with the prominent characteristics of the team theme to achieve partial clothing adjustment; Otherwise, the sequence of the corresponding action positions is not adjusted; According to the final result of each action position, an adjusted local sequence is obtained.

[0058] In this embodiment, the body model framework and the skull model framework are combined to obtain a self-contour framework.

[0059] In this embodiment, the convertible advantage is, for example, short legs, which can be concealed by dressing up in clothes, and the outstanding advantage refers to the advantages that exist in oneself, which can be used to make the overall performance atmosphere more prominent.

[0060] In this embodiment, each convertible advantage and outstanding advantage is realized based on its own contour frame, that is, there is a certain positional comparison relationship with its own contour frame. At this time, in addition to determining the first imbalance factor and the second imbalance factor, position matching is required to determine the imbalance factor based on the position, and regard it as an adaptation factor to realize the calculation of the overall adaptation value.

[0061] For example, for the first imbalance factor a1 and the second imbalance factor a2 involved in action position 1, at this time, a1 and a2 are the adaptation factors of the corresponding action position, and the overall adaptation value of the position = the position weight of the position × (a1+a2).

[0062] In this embodiment, the preset adaptation value is pre-set and stored in a position-adaptation value comparison table, which contains different action positions and preset adaptation values ​​matching the positions.

[0063] In this embodiment, the adjustment of the corresponding sequence is obtained by inputting the initial setting performance results, the first imbalance factor involved, the second imbalance factor and the prominent characteristics of the team theme into the combination analysis model as a whole for analysis. The combination analysis model contains different setting combinations and the sequence adjustment results corresponding to the combinations as samples for training the neural network model, which can be directly trained.

[0064] In this embodiment, the difference between the final result and the initial setting performance result is that the result is more in line with the overall needs of the team and also more in line with the members themselves.

[0065] The beneficial effect of the above technical solution is: by determining the imbalance coefficient between the convertible advantage and the outstanding advantage, it is convenient to calculate the overall adaptation value, and then by making a size judgment analysis on the overall adaptation configuration, a reasonable adjustment of the sequence can be achieved. While meeting the overall demand effect, it is also more suitable for the person himself, providing a basis for clothing integration.

[0066] The present invention provides an integrated clothing style recommendation method based on team needs, which obtains a second clothing recommendation set corresponding to a single member by adjusting a local sequence, including: comparing the adjusted local sequence with a personalized clothing database; According to the comparison result, a second clothing recommendation set corresponding to the individual member is obtained.

[0067] In this embodiment, the personalized clothing database includes recommended clothing under different combination realization results, and is a pre-set storage database.

[0068] The beneficial effect of the above technical solution is that by comparing the sequence with the database, it is easy to obtain a recommendation set, which facilitates subsequent integration.

[0069] The present invention provides an integrated clothing style recommendation method based on team needs, which performs clothing integration processing according to the first clothing recommendation set and all second clothing recommendation sets to obtain clothing recommendation results, including: Inputting the first clothing recommendation set and each second clothing recommendation set into a team clothing analysis table in sequence, wherein the team clothing analysis table includes overall clothing elements and unique clothing elements of different individual members; Count the coordination between each column's unique clothing element and each overall clothing element; If all coordination is greater than the corresponding preset, the corresponding unique clothing elements will be retained; Otherwise, the elements whose coordination is less than the preset among all the unique clothing elements of each individual member are locked and regarded as the first elements, and the elements to be adjusted and the personnel to be adjusted are screened by counting the first number of the first elements involved in each individual member and counting the second number of the first elements involved in each column of unique clothing elements; Weakening the local parameter set of the personnel to be adjusted according to the global parameter set and in combination with the division of labor roles of the personnel to be adjusted; performing information standardization processing on the element information of the unique clothing element in the column corresponding to the element to be adjusted according to the clothing set with the maximum recommendation coefficient in the first clothing recommendation set; According to the weakening processing results and the information standardization processing results, the clothing recommendation results are obtained.

[0070] In this embodiment, the team clothing analysis table includes descriptions of all clothing elements, and clothing elements are divided into overall clothing elements and unique clothing elements. The clothing recommendation set is input into the clothing element analysis model to obtain clothing elements of corresponding members. The model is obtained by training a neural network model based on different clothing recommendation sets and analysis results of clothing elements in the recommendation sets by clothing stylists as samples. Therefore, clothing elements involved in each member can be directly obtained, wherein overall clothing elements correspond to global parameters, and unique clothing elements correspond to local parameters. In this embodiment, all the results in the corresponding table are standardized to obtain corresponding values ​​to facilitate subsequent calculations.

[0071] In this embodiment, coordination = the average value of the sum of the relative matching values ​​of the corresponding column unique clothing elements and the corresponding overall clothing elements, and the relative matching value is obtained by matching from the unique-overall clothing comparison table, which contains the results of different unique clothing elements and the degree of coordination of the results of the overall clothing elements, that is, whether the clothes are coordinated. The more coordinated, the higher the degree of coordination of the results, and the average value of all the coordination degrees of the results is the coordination.

[0072] In this embodiment, the preset degree is preset. For example, the coordination degree between the scarf and the jumpsuit is 0.6 and the preset degree is 0.4. At this time, the corresponding unique clothing element is retained. Otherwise, the unique element with a coordination degree less than the preset degree is first locked and regarded as the first element. At the same time, the second element with a coordination degree less than the preset degree is locked from each column. If the first quantity of the same member is greater than half of the quantity of all clothing elements, the member is regarded as a person to be adjusted.

[0073] If the second number is greater than half of the number of all individual members, the element is regarded as an element to be adjusted, and the element to be adjusted is only for a unique clothing element.

[0074] In this embodiment, the purpose of the weakening process is to weaken the effect of each local parameter in the corresponding local parameter set on the corresponding person. For example, the person to be adjusted was originally asked to wear old-age clothes, but after weakening, he can wear round-neck clothes.

[0075] In this embodiment, the first recommended clothing in the first clothing recommendation set is sorted from large to small according to the recommendation coefficient, and the first recommended clothing is obtained in the matching process according to the global parameter set. This matching process can directly obtain the corresponding recommendation coefficient, that is, the closer the match, the higher the corresponding recommendation coefficient.

[0076] In this embodiment, the information standardization process refers to standardizing the element information of the unique clothing elements based on the clothing set with the maximum recommendation coefficient, so as to meet the overall clothing effect.

[0077] In this embodiment, the clothing recommendation result is the result of weakening, information standardization and original unchanged.

[0078] For example, the weakening process is to change the turtleneck into a round neck, and the information standardization process is to adjust the 1*1m wide silk scarf of the corresponding element to a 0.5*0.5m wide silk scarf, or to increase the number of cuff buttons from 2 to 5, etc.

[0079] The beneficial effects of the above technical solution are: through the team clothing analysis table, the coordination analysis of unique clothing elements and overall clothing elements is carried out, and then the number of different elements is determined by size comparison to preliminarily determine the elements to be adjusted and the personnel to be adjusted, and combined with weakening processing and information standardization processing to ensure the integrity and integration of the results, so as to ensure the overall performance atmosphere and improve the efficiency of clothing recommendations.

[0080] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An integrated clothing style recommendation method based on team needs, characterized in that: include: Step 1: Analyze team needs based on team theme to obtain a global parameter set and a local parameter set around the team theme, wherein the global parameter set is related to the overall clothing style of the team, and the local parameter set is related to the clothing style of individuals in the team; Step 2: Collect the first clothing recommendation set matching each global parameter in the global parameter set; Step 3: Mapping the local parameter set to the self-contour framework of each individual member, and obtaining a second clothing recommendation set for each individual member; Step 4: Perform clothing integration processing based on the first clothing recommendation set and all second clothing recommendation sets to obtain clothing recommendation results, wherein the clothing recommendation results include the overall dress of the team and the individual dress of each member.

2. The integrated clothing style recommendation method based on team needs according to claim 1 is characterized in that: The team requirements are parsed based on the team theme to obtain a global parameter set and a local parameter set around the team theme, including: Based on the team theme, a demand analysis model is obtained by matching from a theme-model comparison table; Analyze the team requirements according to the requirements analysis model to obtain a plurality of first initial parameters and a plurality of second initial parameters; Based on the required role of each first initial parameter in the parsing process, an initial radius is set to the corresponding first initial parameter; Setting a first weight based on the first initial parameters corresponding to all initial radii; At the same time, an equal weight is assigned to each second initial parameter as a second weight; Counting all first initial parameters set with first weights as a global parameter set; All second initial parameters set with second weights are counted as a local parameter set.

3. The integrated clothing style recommendation method based on team needs according to claim 2 is characterized in that: The first weight is set based on the first initial parameters corresponding to all the initial radii, including: Determine the circumference and the circumference area of ​​the circle corresponding to the first initial parameter based on the set initial radius; Performing a first normal distribution process on all the circumferences of the circles to obtain a first radius, and at the same time, performing a second normal distribution process on all the circumference areas to obtain a second radius; Selecting a maximum radius and a minimum radius from all initial radii, and setting a radius threshold based on a difference between the maximum radius and the minimum radius and in combination with the first radius and the second radius; ； in, Indicates the corresponding radius threshold; Respectively represent the average values ​​based on all first radii and all second radii; Respectively represent the maximum radius and the minimum radius; Respectively represent the number of all first radii and the number of all second radii; represents the radius fine-tuning function; Performing intersection processing on the first initial quantity of the long side radius outside the first normal distribution and the first initial quantity of the long side radius outside the second normal distribution to obtain an intersection initial parameter; If the number of the intersection initial parameters is 0, setting a first importance value for each first initial parameter according to a difference between an initial radius of each first initial parameter and a radius threshold; ； in, represents the first important value of the i1th first initial parameter; represents the initial radius of the i1th first initial parameter; If the number of the intersection initial parameters is not 0, then according to the number of long side radii corresponding to each first initial parameter and in combination with the difference between the initial radius of each first initial parameter and the radius threshold, a second important value is set for each first initial parameter; ； in, represents the second most important value of the i1th first initial parameter; Indicates that the initial radius corresponding to the i1th first initial parameter belongs to the number of long side radii; Indicates the number of parameters whose initial radius belongs to the long side radius among all the first initial parameters; ln represents the sign of the logarithmic function; Based on the importance values ​​of all first initial parameters, first weights are set to corresponding first initial parameters.

4. The integrated clothing style recommendation method based on team needs according to claim 1 is characterized in that: Collecting a first clothing recommendation set matching each global parameter in the global parameter set, including: According to the parameter attribute of the global parameter, a first clothing recommendation set matching the parameter attribute is matched from an attribute-clothing database.

5. The integrated clothing style recommendation method based on team needs according to claim 1 is characterized in that: Mapping the local parameter set to the self-profile framework of each individual member includes: Determine the action position of each local parameter in the local parameter set, compare the action position with the self-contour framework, and combine the initial setting performance result of each local parameter to obtain the initial local sequence corresponding to the individual member; Obtain the outstanding advantages and convertible advantages of each individual member's own profile structure, and adjust the initial local sequence in combination with the outstanding characteristics of the team theme; The second clothing recommendation set corresponding to the individual member is obtained by adjusting the local sequence.

6. The integrated clothing style recommendation method based on team needs according to claim 5 is characterized in that: The initial local sequence is adjusted, including: Acquire the transferable and outstanding advantages of each individual member based on their own profile framework; Determine a first imbalance coefficient corresponding to the convertible advantage of the individual member, and at the same time, determine a second imbalance coefficient corresponding to the outstanding advantage of the individual member; ； in, represents the first imbalance coefficient corresponding to the j1th convertible advantage of the individual member; represents the second imbalance coefficient corresponding to the j2th prominent advantage of the individual member; It indicates the number of times the historical costumes corresponding to the j1-th convertible advantage of a single member are recommended and adjusted successfully; represents the total number of historical occurrences consistent with the j1th convertible advantage of the corresponding individual member; represents the number of failed historical recommended adjustments consistent with the j1th convertible advantage of the corresponding individual member; It represents the advantage prominence coefficient corresponding to the j1th convertible advantage of the individual member; It represents the advantage prominence coefficient corresponding to the j2th prominence of the individual member; Indicates the total number of convertible advantages involved in the corresponding individual member; It represents the total number of convertible advantages involved by all individual members; The variance of the advantage salience coefficient representing the convertible advantages involved by all individual members; Extracting the adaptation factor based on each action position from the first imbalance factors and the second imbalance factors of all individual members, and calculating the overall adaptation value of the corresponding action position; If the overall adaptation value is greater than the preset adaptation value of the corresponding action position, the corresponding sequence is adjusted according to the first imbalance factor and the second imbalance factor involved in the corresponding action position and in combination with the prominent characteristics of the team theme to achieve partial clothing adjustment; Otherwise, the sequence of the corresponding action positions is not adjusted; According to the final result of each action position, an adjusted local sequence is obtained.

7. The integrated clothing style recommendation method based on team needs according to claim 5 is characterized in that: According to the adjustment of the local sequence, the second clothing recommendation set corresponding to the individual member is obtained, including: comparing the adjusted local sequence with a personalized clothing database; According to the comparison result, a second clothing recommendation set corresponding to the individual member is obtained.

8. The integrated clothing style recommendation method based on team needs according to claim 1 is characterized in that: Performing clothing integration processing according to the first clothing recommendation set and all second clothing recommendation sets to obtain clothing recommendation results, including: Inputting the first clothing recommendation set and each second clothing recommendation set into a team clothing analysis table in sequence, wherein the team clothing analysis table includes overall clothing elements and unique clothing elements of different individual members; Count the coordination between each column's unique clothing element and each overall clothing element; If all coordination is greater than the corresponding preset, the corresponding unique clothing elements will be retained; Otherwise, the elements whose coordination is less than the preset among all the unique clothing elements of each individual member are locked and regarded as the first elements, and the elements to be adjusted and the personnel to be adjusted are screened by counting the first number of the first elements involved in each individual member and counting the second number of the first elements involved in each column of unique clothing elements; Weakening the local parameter set of the personnel to be adjusted according to the global parameter set and in combination with the division of labor roles of the personnel to be adjusted; performing information standardization processing on the element information of the unique clothing element in the column corresponding to the element to be adjusted according to the clothing set with the maximum recommendation coefficient in the first clothing recommendation set; According to the weakening processing results and the information standardization processing results, the clothing recommendation results are obtained.

Images