CN111047407A - Clothing personalized size customization method using variational multidimensional regression - Google Patents
Clothing personalized size customization method using variational multidimensional regression Download PDFInfo
- Publication number
- CN111047407A CN111047407A CN201911282371.2A CN201911282371A CN111047407A CN 111047407 A CN111047407 A CN 111047407A CN 201911282371 A CN201911282371 A CN 201911282371A CN 111047407 A CN111047407 A CN 111047407A
- Authority
- CN
- China
- Prior art keywords
- regression
- variational
- steps
- sample
- theta
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 42
- 239000011159 matrix material Substances 0.000 claims abstract description 27
- 238000013507 mapping Methods 0.000 claims abstract description 12
- 239000000126 substance Substances 0.000 claims description 7
- 230000001419 dependent effect Effects 0.000 claims description 3
- 210000000707 wrist Anatomy 0.000 claims description 3
- 238000004513 sizing Methods 0.000 claims description 2
- 230000002159 abnormal effect Effects 0.000 abstract description 5
- 238000013461 design Methods 0.000 abstract description 4
- 230000001174 ascending effect Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000007493 shaping process Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0621—Item configuration or customization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Finance (AREA)
- Mathematical Physics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Operations Research (AREA)
- Algebra (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method for customizing the personalized size of a garment by using variational multidimensional regression. The method comprises the following steps: constructing a sample set; establishing a multi-dimensional regression prediction model of each size of the clothes; predicting an optimal weight coefficient matrix of the established multidimensional variational regression model by using data sample information of nonlinear weighted mapping and adopting a multidimensional variational regression algorithm; and establishing a prediction model equation, and accurately predicting the corresponding size of the customized garment by using the prediction model equation after obtaining the optimal weight coefficient matrix. The variational multidimensional regression algorithm provided by the invention can effectively solve the overfitting problem and the matrix underrank problem of the regression algorithm sensitive to abnormal values, thereby realizing accurate prediction of the corresponding dimension of the clothing based on a prediction equation and providing personalized accurate custom design for the clothing of a user.
Description
Technical Field
The invention relates to the technical field of variational multi-dimensional regression modeling, in particular to a garment personalized size customization method using variational multi-dimensional regression.
Background
With the improvement of living conditions of people, the demand of high-end consumers on individuation is increasingly obvious, and the demand of the consumers on private customized services is also increasing. At present, some existing garment customization systems usually measure the size of a user by an experienced garment designer on line, and then determine the garment type, so that the off-line cost is high, and the on-line popularization is not facilitated; other garment customization systems adopt a small number of fixed-format templates to match the stature size of a user, so that the garment customization systems cannot better approach the stature size data of the user in practical application and cannot meet the customization requirements of the user.
Disclosure of Invention
The present invention is directed to solving the problems set forth in the background. Therefore, the invention aims to provide a clothing personalized size customization method using variational multidimensional regression, and the variational multidimensional regression algorithm provided by the invention can effectively solve the problems of overfitting and matrix undersequence of the regression algorithm sensitive to abnormal values, thereby realizing accurate prediction of various sizes of clothing based on a prediction equation and providing personalized accurate customization design for the clothing size of a user.
According to the embodiment of the invention, the method for customizing the personalized size of the clothes by using the variational multidimensional regression comprises the following steps:
step 1: carrying out weighted nonlinear mapping based on height, weight and the like on the existing user information by using a function f to obtain a multidimensional data sample of the existing user;
step 2: establishing a multi-dimensional variational regression model;
and step 3: predicting an optimal weight coefficient matrix in the established multidimensional variational regression model for the data sample by adopting a multidimensional variational regression algorithm;
and 4, step 4: and establishing a prediction model equation, and accurately predicting the corresponding size of the customized garment by using the prediction model equation after obtaining the optimal weight coefficient matrix.
Preferably, the step 1 comprises:
establishing an input data set:
X＝{x1,x2,..,xN}
Y＝{y1,y2,..,yN}
wherein each sample is:
xn＝{xin,m＝1,2,...,K}(n＝1,2,...,N)
yn＝{ymn,m＝1,2,...,M}(n＝1,2,...,N)
i represents different input option information, N represents the serial number of the sample, N samples are shared, and xnThe option information for the nth sample is sampleIndependent variable of this, ynThe actual size data of the nth sample is dependent variable of the sample;
the data set is:
D＝{(x1,y1),(x2,y2),...,(xN,yN)}。
preferably, the i ═ 1, 2., K, respectively, represent height, weight, neck circumference, shoulder width, chest circumference, arm length, arm circumference, wrist circumference, waist circumference, and fit preference selected by the user;
m is 1,2,.. times, M, which respectively represents the actual garment length, collar circumference, chest circumference, shoulder width, sleeve circumference, waist circumference, sleeve length, and cuff circumference size of the corresponding sample.
Preferably, the step 2 includes:
establishing a multi-dimensional variational regression model by adopting a sample:
yn＝ΘTf(xn)+e；
wherein, the coefficient matrix Θ ═ θ1,θ2,...,θM) And thetaj＝{θijI 1, 2.. K (j 1, 2.. M) is a weighting parameter for the input, and e is zero-mean gaussian noise.
f is to the input variable xnComprising an encoding of the input information options and a mapping function weighted based on the height weight information.
Preferably, the step 3 comprises:
the weight coefficient matrix of the multi-dimensional variational regression model
Make it
As close to y as possiblenThat is to say make
As close to y as possiblenjSo that v is equal to v for the new input informationiAnd i is 1,2, a.
Wherein s ═ sjJ 1, 2.. said, M }, which is the customized clothing information.
Preferably, e is a zero mean gaussian distribution, and the likelihood function for θ is a gaussian distribution of:
θjthe prior distribution of (a) is:
p(θj|α)＝N(0,α-1I)；
the conjugate prior of the gaussian distribution is the Gamma distribution;
p(α)＝Gamma(a0,b0)；
the joint probability density function is:
p(yn,θj,α)＝p(yn|θj)p(θj|α)p(α)。
preferably, the step 3 specifically further includes a step 3.1:
the equation used to predict the best accurate dimensional information for a custom-made garment is a probability density function q (θ)j,α)＝qθ(θj)qα(α) to approximate the function p (θ) in step 3j,α|yn)；
Solving a probability density function q (theta) from two ends of the above formulajα) to obtain
For the third term at the right end, the approximate probability density q (theta) to be searchedjα) and the posterior probability p (θ)j,α|yn) KL divergence of (d), indicating the degree of similarity of the two:
lnp(yn)≥Eq[lnp(yn,θj,α)]-Eq[q(θj,α)]。
preferably, the step 3 specifically further includes a step 3.2:
when the optimum theta is estimatedjAnd α parameters are such that Eq[lnp(yn,θj,α)]-Eq[q(θj,α)]When the maximum value is obtained, the posterior probability density function p (theta) can be obtainedj,α|yn) The optimal approximation of (c):
the optimal approximation can be derived as:
as can be seen from the above formula, it follows a Gamma distribution, i.e.:
wherein the content of the first and second substances,
a0＝10-8
b0＝10-8
in the same way, the following can be obtained:
this is a gaussian distribution, being:
wherein the content of the first and second substances,
preferably, the step 3 specifically further includes a step 3.3:
to q isα(α) initializing, and updating q by coordinate ascent methodθ(θj) Updated qθ(θj) Then update qα(α) alternately updating the factors until convergence
To obtain thetaj(j ═ 1, 2.. M) to obtain an optimal solution, and a regression coefficient matrix can be obtained
Preferably, the function modeling is implemented as follows:
s1: according to function yn＝ΘTf(xn) + e constructing a regression model;
s2: utilizing functions to sample data information
Performing encoding and non-linear mapping based on height and weight information weighting;
s3: to q isα(α) initializing, using the initialized probability function qα(α) pairing q with the function in step 3.2θ(θj) Performing an update to obtain an updated thetajA (j ═ 1,2,. M) value;
s4: with updated qθ(θj) Using pairs q in step 3.2α(α) updating;
s5: iteration is carried out, and related parameters are alternately updated until convergence;
s6: by
Obtain thetaj(j ═ 1, 2.. M) to obtain an estimated value of the regression coefficient
S7: by using
The prediction model obtains each customized size value of the personalized clothes.
The beneficial effects of the invention are as follows:
the method encodes the existing multi-sample multi-dimensional information and performs nonlinear mapping based on the relations of height, weight and the like; establishing a multi-dimensional regression prediction model of each size of the clothes; converting the problem of solving the weighting coefficient in the model into an optimization problem of approximating the posterior probability by a probability density function; and performing model training by using the mapped data sample information through variational processing, alternately and iteratively updating the probability density function through a coordinate ascending method, and obtaining an optimized weight matrix during convergence.
The variational multidimensional regression algorithm provided by the invention can effectively solve the overfitting problem and the matrix underrank problem of the regression algorithm sensitive to abnormal values, thereby realizing accurate prediction of various dimensions of clothes based on a prediction equation and providing personalized accurate custom design for the clothes dimensions of users.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
fig. 1 is a flow chart of a method for customizing the personalized size of a garment by using variational multidimensional regression according to the invention.
Detailed Description
The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic views illustrating only the basic structure of the present invention in a schematic manner, and thus show only the constitution related to the present invention.
Referring to fig. 1, a method for personalized sizing of a garment using variational multidimensional regression, the method comprising the steps of:
step 1: carrying out weighted nonlinear mapping based on height, weight and the like on the existing user information by using a function f to obtain a multidimensional data sample of the existing user;
step 2: establishing a multi-dimensional variational regression model;
and step 3: predicting an optimal weight coefficient matrix in the established multidimensional variational regression model for the data sample by adopting a multidimensional variational regression algorithm;
and 4, step 4: and establishing a prediction model equation, and accurately predicting the corresponding size of the customized garment by using the prediction model equation after obtaining the optimal weight coefficient matrix.
In one embodiment, there may be four steps:
the method comprises the following steps: constructing multiple sample sets xnAnd ynAnd in xnInternally inputting information selected by the user at ynActual letter of internal input sampleInformation;
step two: establishing a multi-dimensional variational regression model;
step three: predicting an optimal weight coefficient matrix in the established multidimensional variational regression model for the data sample by adopting a multidimensional variational regression algorithm;
step four: and establishing a prediction model equation, and accurately predicting the corresponding size of the customized garment by using the prediction model equation after obtaining the optimal weight coefficient matrix.
The step 1 comprises the following steps:
establishing aggregate input data:
X＝{x1,x2,..,xN}
Y＝{y1,y2,..,yN}
wherein each sample is:
xn＝{xin,m＝1,2,...,K}(n＝1,2,...,N)
yn＝{ymn,m＝1,2,...,M}(n＝1,2,...,N)
i represents different input option information, N represents the serial number of the sample, N samples are shared, and xnThe option information for the nth sample is the argument of the sample, ynThe actual size data of the nth sample is dependent variable of the sample;
the data set is:
D＝{(x1,y1),(x2,y2),...,(xN,yN)}。
the value of K is an arbitrary constant, that is, the value of K includes other user-defined information in addition to the above-mentioned information.
Height (cm), weight (kg), neck circumference (thin, common, thick), shoulder width (narrow, common, wide), chest circumference (big, normal, obese), arm length (short, normal, long), arm circumference (thin, common, thick), wrist circumference (thin, common, thick), waist circumference (normal, slightly fat, large), version preference (shaping, loose, standard) and the like. Besides height and weight, other inputs are one of the information selected by the user with the related options.
M is 1,2, the length of the collar, the circumference of the chest, the width of the shoulder, the circumference of the sleeves, the circumference of the waist and the length of the sleeves, the circumference of the cuffs and the like of the corresponding samples, and the value of M is an arbitrary constant, that is, the size information of the samples is included in addition to the size information given above.
The step 2 comprises the following steps:
establishing a multi-dimensional variational regression model by adopting a sample:
yn＝ΘTf(xn)+e；
wherein, the coefficient matrix Θ ═ θ1,θ2,...,θM) And thetaj＝{θijI 1, 2.. K (j 1, 2.. M) is a weighting parameter for the input, and e is zero-mean gaussian noise.
f is to the input variable xnComprising an encoding of the input information options and a mapping function weighted based on the height weight information.
The step 3 comprises the following steps:
the regression coefficient is usually obtained by a least square method, but the least square method causes an overfitting phenomenon because data information contains an abnormal value, and in addition, when the matrix is obtained in a pseudo-inverse manner, the matrix may not be of full rank, so that the prediction error of the constructed regression model is large, and the algorithm has defects.
In order to accurately, conveniently and individually customize the clothes, a multi-dimensional variational regression algorithm is provided to predict the optimal weight coefficient matrix
Make it
As close to y as possiblenThat is to say make
As close to y as possiblenjSo that v is equal to v for the new input informationiAnd i is 1,2, a.
Wherein s ═ sjJ 1, 2.. said, M }, which is the customized clothing information;
e is a zero mean gaussian distribution, and the likelihood function for θ is gaussian:
θjthe prior distribution of (a) is:
p(θj|α)＝N(0,α-1I)；
the conjugate prior of the gaussian distribution is the Gamma distribution;
p(α)＝Gamma(a0,b0)；
the joint probability density function is:
p(yn,θj,α)＝p(yn|θj)p(θj|α)p(α)；
said step 3 comprises a step 3.1: the equation used to predict the best accurate dimensional information for a custom-made garment is a probability density function q (θ)j,α)＝qθ(θj)qα(α) to approximate the function p (θ) in step 3j,α|yn)；
Solving a probability density function q (theta) from two ends of the above formulajα) to obtain
For the third term on the right hand side, the approximate probability density q (theta) to be foundjα) and the posterior probability p (θ)j,α|yn) KL divergence of (d), indicating the degree of similarity of the two:
lnp(yn)≥Eq[lnp(yn,θj,α)]-Eq[q(θj,α)]；
step 3 further comprises step 3.2:
when the optimum theta is estimatedjAnd α parameters are such that Eq[lnp(yn,θj,α)]-Eq[q(θj,α)]When the maximum value is obtained, the posterior probability density function p (theta) can be obtainedj,α|yn) The optimal approximation of (c):
the optimal approximation can be derived as:
as can be seen from the above formula, it follows a Gamma distribution, i.e.:
wherein the content of the first and second substances,
a0＝10-8
b0＝10-8
in the same way, the following can be obtained:
this is a gaussian distribution, being:
wherein the content of the first and second substances,
step 3 of the method further comprises step 3.3:
to q isα(α) initializing, and updating q by coordinate ascent methodθ(θj) Updated qθ(θj) Then update qα(α) alternately updating the factors until convergence
To obtain thetaj(j ═ 1, 2.. M) to obtain an optimal solution, and a regression coefficient matrix can be obtained
The step 4 further comprises: using a matrix of regression coefficients
And establishing a prediction model equation so as to accurately determine the corresponding dimension of the customized garment.
The function modeling is realized by the following specific method:
s1: according to function yn＝ΘTf(xn) + e constructing a regression prediction model;
s2: utilizing functions to sample data information
Performing encoding and non-linear mapping based on height and weight information weighting;
s3: to q isα(α) initializing, using the initialized probability function qα(α) pairing q with the function in step 3.2θ(θj) Performing an update to obtain an updated thetajA (j ═ 1,2,. M) value;
s4: with updated qθ(θj) Using pairs q in step 3.2α(α) updating;
s5: iteration is carried out, and related parameters are alternately updated until convergence;
s6: by
Obtain thetaj(j ═ 1, 2.. M) to obtain an estimated value of the regression coefficient
S7: by using
The prediction model obtains each customized size value of the personalized clothes.
The method encodes the existing multi-sample multi-dimensional information and performs nonlinear mapping based on the relations of height, weight and the like; establishing a multi-dimensional regression prediction model of each size of the clothes; converting the problem of solving the weighting coefficient in the model into an optimization problem of approximating the posterior probability by a probability density function; and performing model training by using the mapped data sample information through variational processing, alternately and iteratively updating the probability density function through a coordinate ascending method, and obtaining an optimized weight vector during convergence.
The variational multidimensional regression algorithm provided by the invention can effectively solve the overfitting problem and the matrix underrank problem of the regression algorithm sensitive to abnormal values, thereby realizing accurate prediction of various dimensions of clothes based on a prediction equation and providing personalized accurate custom design for the clothes dimensions of users.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art should be considered to be within the technical scope of the present invention, and the technical solutions and the inventive concepts thereof according to the present invention should be equivalent or changed within the scope of the present invention.
Claims (10)
1. A method for customizing the personalized size of clothes by using variational multidimensional regression is characterized in that: the method comprises the following steps:
step 1: carrying out weighted nonlinear mapping based on height, weight and the like on the existing user information by using a function f to obtain a multidimensional data sample of the existing user;
step 2: establishing a multi-dimensional variational regression model;
and step 3: predicting an optimal weight coefficient matrix in the established multidimensional variational regression model for the data sample by adopting a multidimensional variational regression algorithm;
and 4, step 4: and establishing a prediction model equation, and accurately predicting the corresponding size of the customized garment by using the prediction model equation after obtaining the optimal weight coefficient matrix.
2. The method of claim 1, wherein the method comprises the steps of: the step 1 comprises the following steps:
establishing an input data set:
X＝{x1,x2,..,xN}
Y＝{y1,y2,..,yN}
wherein each sample is:
xn＝{xin,i＝1,2,...,K}(n＝1,2,...,N)
yn＝{ymn,m＝1,2,...,M}(n＝1,2,...,N)
i represents different input option information, N represents the serial number of the sample, N samples are shared, and xnThe input option information for the nth sample is the independent variable of the sample, ynThe actual size data of the nth sample is dependent variable of the sample;
the data set is:
D＝{(x1,y1),(x2,y2),...,(xN,yN)}。
3. the method of claim 2, wherein the method comprises the steps of: the i-1, 2., K respectively represents height, weight, neck circumference, shoulder width, chest circumference, arm length, arm circumference, wrist circumference, waist circumference and version preference selected by the user;
m is 1, 2.. times, M, which respectively represents the actual garment length, collar circumference, chest circumference, shoulder width, sleeve circumference, waist circumference, sleeve length, and cuff circumference size of the corresponding user sample.
4. The method of claim 1, wherein the method comprises the steps of: the step 2 comprises the following steps:
establishing a multi-dimensional variational regression model by adopting a sample:
yn＝ΘTf(xn)+e；
wherein, the coefficient matrix Θ ═ θ1,θ2,...,θM) And thetaj＝{θijI 1, 2.. K } (j 1, 2.. M) is a weighting parameter for the input, e is zero-mean gaussian noise;
f is to the input variable xnComprising an encoding of the input information options and a mapping function weighted based on the height weight information.
5. The method of claim 1, wherein the method comprises the steps of: the step 3 comprises the following steps:
the optimal weight coefficient matrix is predicted by adopting a multi-dimensional variational regression algorithm
Make it
As close to y as possiblenThat is to say make
As close to y as possiblenjSo that v is equal to v for the new input informationi1, 2.., K }, using an output prediction model equation:
the dimensions of the custom-made garment are accurately predicted. Wherein s ═ sjJ 1, 2.. said, M }, which is the customized garment size information.
6. The method of claim 5, wherein the method comprises the steps of: the algorithm of the step 3 comprises the following steps:
θjthe prior distribution of (a) is:
p(θj|α)＝N(0,α-1I)；
the conjugate prior of the gaussian distribution is the Gamma distribution;
p(α)＝Gamma(a0,b0)；
the joint probability density function is:
p(yn,θj,α)＝p(yn|θj)p(θj|α)p(α)。
7. the method of claim 5, wherein the method comprises the steps of: the step 3 specifically comprises a step 3.1:
using a probability density function q (theta)j,α)＝qθ(θj)qα(α) to approximate the function p (theta)j,α|yn)；
Solving a probability density function q (theta) from two ends of the above formulajα) to obtain
For the third term at the right end, the approximate probability density q (theta) to be searchedjα) and the posterior probability p (θ)j,α|yn) KL divergence of (d), indicating the degree of similarity of the two:
ln p(yn)≥Eq[lnp(yn,θj,α)]-Eq[q(θj,α)]。
8. the method of claim 5, wherein the method comprises the steps of: the step 3 specifically further comprises a step 3.2:
when the optimum theta is estimatedjAnd α parameters are such that Eq[lnp(yn,θj,α)]-Eq[q(θj,α)]When the maximum value is obtained, the posterior probability density function p (theta) can be obtainedj,α|yn) Is the most important ofExcellent approximation:
the optimal approximation can be derived as:
as can be seen from the above formula, it follows a Gamma distribution, i.e.:
wherein the content of the first and second substances,
a0＝10-8
b0＝10-8
in the same way, the following can be obtained:
this is a gaussian distribution, being:
wherein the content of the first and second substances,
wherein the content of the first and second substances,
9. the method of claim 5, wherein the method comprises the steps of: the step 3 specifically further comprises a step 3.3:
to q isα(α) initializing, and updating q by coordinate ascent methodθ(θj) Updated qθ(θj) Then update qα(α) alternately updating the factors until convergence
To obtain thetaj(j ═ 1, 2.. M) to obtain an optimal solution, and a regression coefficient matrix can be obtained
10. The method of personalized sizing of garments using variational multidimensional regression according to any of claims 1 to 9, characterized in that: the function modeling is realized by the following specific method:
s1: according to function yn＝ΘTf(xn) + e constructing a regression model;
s2: utilizing functions to sample data information
(N ═ 1, 2.. times, N) encoding and weighted based on height and weight information nonlinear mapping;
s3: to q isα(α) initializing, using the initialized probability function qα(α) pairing q with the function in step 3.2θ(θj) Performing an update to obtain an updated thetajA (j ═ 1,2,. M) value;
s4: with updated qθ(θj) Using pairs q in step 3.2α(α) updating;
s5: iteration is carried out, and related parameters are alternately updated until convergence;
s6: by
Obtain thetaj(j ═ 1, 2.. M) to obtain an estimated value of the regression coefficient
S7: by using
The prediction model obtains each customized size value of the personalized clothes.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911282371.2A CN111047407A (en) | 2019-12-13 | 2019-12-13 | Clothing personalized size customization method using variational multidimensional regression |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911282371.2A CN111047407A (en) | 2019-12-13 | 2019-12-13 | Clothing personalized size customization method using variational multidimensional regression |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111047407A true CN111047407A (en) | 2020-04-21 |
Family
ID=70236160
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911282371.2A Pending CN111047407A (en) | 2019-12-13 | 2019-12-13 | Clothing personalized size customization method using variational multidimensional regression |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111047407A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113706688A (en) * | 2021-07-22 | 2021-11-26 | 定智衣(上海)服装科技有限公司 | Dynamic human body size characteristic modeling method |
| CN114331216A (en) * | 2022-01-21 | 2022-04-12 | 杭州贝嘟科技有限公司 | Garment size evaluation method, electronic device, and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101013508A (en) * | 2007-02-12 | 2007-08-08 | 西安工程大学 | Method for constructing divisional composite three-dimensional parameterized digital mannequin |
| CN107154071A (en) * | 2016-03-02 | 2017-09-12 | 南京航空航天大学 | The method that Case-based Reasoning generates individual face body Model according to anthropological measuring size data |
| CN109214605A (en) * | 2018-11-12 | 2019-01-15 | 国网山东省电力公司电力科学研究院 | Power-system short-term Load Probability prediction technique, apparatus and system |
-
2019
- 2019-12-13 CN CN201911282371.2A patent/CN111047407A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101013508A (en) * | 2007-02-12 | 2007-08-08 | 西安工程大学 | Method for constructing divisional composite three-dimensional parameterized digital mannequin |
| CN107154071A (en) * | 2016-03-02 | 2017-09-12 | 南京航空航天大学 | The method that Case-based Reasoning generates individual face body Model according to anthropological measuring size data |
| CN109214605A (en) * | 2018-11-12 | 2019-01-15 | 国网山东省电力公司电力科学研究院 | Power-system short-term Load Probability prediction technique, apparatus and system |
Non-Patent Citations (1)
| Title |
|---|
| 李鹏飞等: "基于机器视觉的服装尺寸在线测量系统" * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113706688A (en) * | 2021-07-22 | 2021-11-26 | 定智衣(上海)服装科技有限公司 | Dynamic human body size characteristic modeling method |
| CN114331216A (en) * | 2022-01-21 | 2022-04-12 | 杭州贝嘟科技有限公司 | Garment size evaluation method, electronic device, and storage medium |
| CN114331216B (en) * | 2022-01-21 | 2022-06-17 | 杭州贝嘟科技有限公司 | Garment size evaluation method, electronic device, and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Jansakul et al. | Score tests for zero-inflated Poisson models | |
| CN108090093B (en) | Method and device for generating recommendation result | |
| Weyer et al. | Asymptotic properties of SPS confidence regions | |
| CN111047407A (en) | Clothing personalized size customization method using variational multidimensional regression | |
| CN107066628A (en) | Wear the clothes recommendation method and device | |
| CN109800821A (en) | Method, image processing method, device, equipment and the medium of training neural network | |
| CN110084221A (en) | A kind of serializing face critical point detection method of the tape relay supervision based on deep learning | |
| Bloem-Reddy et al. | Random-walk models of network formation and sequential Monte Carlo methods for graphs | |
| CN110222925A (en) | Performance quantization wire examination method, device and computer readable storage medium | |
| CN112084878A (en) | Method for judging attitude standard degree of operator | |
| CN115048938A (en) | Statement emotion analysis method and device based on semantic and syntax dual channels | |
| CN108804565A (en) | A kind of knowledge based indicates the Harmonious Matrix decomposition method of study | |
| CN109214333A (en) | Convolutional neural networks structure, face character recognition methods, device and terminal device | |
| CN111882005A (en) | Data type determination method and device, terminal equipment and storage medium | |
| Wang et al. | QSAR models on aminopyrazole-substituted resorcylate compounds as Hsp90 inhibitors | |
| CN115424660A (en) | Method and device for predicting multi-source information relation by using prediction model | |
| CN111160443A (en) | Activity and user identification method based on deep multitask learning | |
| CN111047406A (en) | Telecommunication package recommendation method, device, storage medium and equipment | |
| Kumar et al. | A min-max approach to fuzzy clustering, estimation, and identification | |
| CN109243562A (en) | A kind of image makings method for improving based on Elman artificial neural network and genetic algorithms | |
| CN109753685A (en) | A kind of logic error rate calculation method | |
| CN111814804A (en) | Human body three-dimensional size information prediction method and device based on GA-BP-MC neural network | |
| CN115906863B (en) | Emotion analysis method, device, equipment and storage medium based on contrast learning | |
| CN114169255B (en) | Image generation system and method | |
| Hossain et al. | Estimation techniques for regression model with zero-inflated poisson data |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20220630 Address after: Room 706, No. 1, Lane 168, Taihong Road, Minhang District, Shanghai 201100 Applicant after: Shanghai bailuo Information Technology Co.,Ltd. Address before: 210042 3rd floor, 699-1 Xuanwu Avenue, Xuanwu District, Nanjing City, Jiangsu Province Applicant before: Nanjing zhonglue Information Technology Co.,Ltd. |
|
| WD01 | Invention patent application deemed withdrawn after publication | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200421 |