CN113538101A - Method, system and storage medium for customizing home decoration based on artificial intelligence - Google Patents

Method, system and storage medium for customizing home decoration based on artificial intelligence Download PDF

Info

Publication number
CN113538101A
CN113538101A CN202110841711.1A CN202110841711A CN113538101A CN 113538101 A CN113538101 A CN 113538101A CN 202110841711 A CN202110841711 A CN 202110841711A CN 113538101 A CN113538101 A CN 113538101A
Authority
CN
China
Prior art keywords
home decoration
information
target
style
model
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
Application number
CN202110841711.1A
Other languages
Chinese (zh)
Inventor
杨浩松
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CN202110841711.1A priority Critical patent/CN113538101A/en
Publication of CN113538101A publication Critical patent/CN113538101A/en
Pending legal-status Critical Current

Links

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
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G06Q30/0255—Targeted advertisements based on user history
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0277—Online advertisement
    • 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]
    • G06Q30/0631—Item recommendations
    • 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
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/08—Construction

Abstract

The invention discloses a home decoration customizing method, a system and a storage medium based on artificial intelligence, relating to the aspect of home decoration customizing. The home decoration customizing method comprises the following steps: establishing a personal-item preference model based on historical data of home decoration items browsed by a target user to generate target home decoration style information of the user, and generating target house type information according to a target house type graph; establishing a home decoration customization model based on a convolutional neural network, importing the target home decoration style information and the target house type information into the home decoration customization model, and making a home decoration scheme by the home decoration customization model according to the target house type information and the target home decoration style information; displaying the home decoration scheme according to a preset mode; meanwhile, the invention can also realize the switching of different home decoration style effects, and provides the personalized customization of the home decoration scheme for the user.

Description

Method, system and storage medium for customizing home decoration based on artificial intelligence
Technical Field
The present invention relates to a home decoration customizing method, and more particularly, to a home decoration customizing method, system and storage medium based on artificial intelligence.
Background
With the development of times and social progress, the economy, politics, science and technology, buildings and the like of China are changed to a certain extent, in particular to a customized mode in the modern home decoration design. Under the background of the modern times, people pursue individuation, the requirements on home decoration are continuously improved, more individuation, artistic and functional requirements are met in the relevant customization process, and meanwhile, with the development of the internet technology and the artificial intelligence technology, the impact of the internet on the aspects of convenience, openness, low cost, high efficiency and the like is increasingly prominent in the home decoration industry. Therefore, more and more traditional home decoration industries begin to transform to informatization and intellectualization, and more humanized and intelligent home decoration customization is provided for consumers.
In order to realize the personalized customization of the home decoration scheme, a system is required to be developed for matching, and the system acquires the target home decoration style information and the target house type information of the user; establishing a home decoration customization model based on a convolutional neural network, importing the target home decoration style information into the home decoration customization model, and making a home decoration scheme by the home decoration customization model according to target house type information and target home decoration style information; and displaying the home decoration scheme according to a preset mode. How to establish a home decoration customization model based on a convolutional neural network and how to realize the customization of a home decoration scheme through the home decoration customization model in the system implementation process are all problems which need to be solved urgently.
Disclosure of Invention
In order to solve at least one technical problem, the invention provides a home decoration customization method, a home decoration customization system and a storage medium based on artificial intelligence.
The invention provides a home decoration customizing method based on artificial intelligence in a first aspect, which comprises the following steps:
establishing a personal-item preference model based on historical data of home decoration items browsed by a target user;
extracting personal preferences through the personal-item preference model, and formally describing the personal preferences according to time to generate a personal preference time sequence;
acquiring personal preference information in a preset interval in the personal preference time sequence;
matching the personal preference information with the information of the candidate home decoration items to acquire the interest degree of the user in the candidate home decoration items;
generating target house decoration style information of the user according to the sequencing information of the interest degree, and generating target house type information according to a target house type graph;
establishing a home decoration customization model based on a convolutional neural network, and importing the target home decoration style information and the target house type information into the home decoration customization model;
the home decoration customization model formulates a home decoration scheme according to target house type information and target home decoration style information;
displaying the home decoration scheme according to a preset mode;
the target home decoration style information comprises at least one home decoration style k, different home decoration styles i are subjected to regionalization fusion and are convolved with target house type imaging information, and then the mathematical expression form of the home decoration customized model is as follows:
Figure BDA0003179179190000021
wherein y represents the home decoration scheme information customized by the home decoration customization model, γ represents the model coefficient, i represents the total number of the home decoration styles contained in the target home decoration style information, and j represents the total number of the home decoration styles contained in the target home decoration style informationkRepresenting a home decoration style k feature area mask, gkAnd (e) representing a style code set of the home decoration style k in the target home decoration style information, representing convolution calculation, and representing the image characteristic of the target house type information.
In this scheme, the building of the home decoration customization model based on the convolutional neural network specifically includes:
establishing an initial home decoration customized model based on a convolutional neural network, and initializing model parameters;
acquiring training data, and dividing the initial home decoration customized model into an image conversion module and a home decoration style construction module;
randomly sampling house type imaging information in the training data and setting a corresponding home decoration style label;
presetting weight and gradient information of the image conversion module and the home decoration style construction module, and alternately training the image conversion module and the home decoration style construction module;
for each iteration of t times, training a home decoration style construction module for t-1 iterations, training an image conversion module for 1 iteration, and updating the weight and gradient information;
updating the parameters of an encoder and a decoder of the home decoration customized model according to the updated weight and gradient information;
and (5) after n iterations of the initial home decoration customized model are carried out, training is completed, and a home decoration customized model is generated.
In this scheme, the home decoration customization model formulates a home decoration scheme through target house type information and target home decoration style information, specifically comprising:
acquiring target house type information, splitting the target house type information, and generating imaging information of a wall body, a floor and a ceiling;
extracting the characteristics of the imaging information, and coding the characteristics through a coder in a home decoration custom model to generate image characteristics;
acquiring target home decoration style information of a user, and importing the image characteristics and the target home decoration style information into a home decoration style building module in a home decoration customization model;
performing regional fusion on different home decoration styles according to the target home decoration style information to form a filter bank;
and performing convolution on the filter group corresponding to the target home decoration style and the image characteristics, and decoding by a decoder to obtain a home decoration scheme conforming to the target home decoration style.
In this scheme, still include:
acquiring a home decoration scheme conforming to a target home decoration style, and accessing the home decoration scheme into a furniture database;
collecting interaction records of a user and home decoration items, collecting price information of the home decoration items which the user is interested in and historically purchasing the home decoration items according to the interaction records, and analyzing purchasing power of the user according to the price information;
generating a furniture recommendation label through the purchasing power and the home decoration style, and creating an index task through the furniture recommendation label by the furniture database;
an index module in the furniture database generates a recommended complete furniture scheme according to the index task and attaches a single furniture commodity link and a price in the complete furniture scheme;
meanwhile, the furniture database can adjust the recommended furniture scheme according to the budget of the user.
In this scheme, in the application process of the home decoration customized model, a new home decoration style needs to be added, and then incremental learning of the home decoration customized model on the new home decoration style specifically includes:
performing joint training on the encoder, the decoder and the multiple home decoration style sets by using a preset training strategy;
the parameters of the encoder and the decoder are fixed, and incremental training is carried out on a home decoration style set corresponding to a home decoration style building module of the home decoration customized model according to the new home decoration style;
performing repeated iterative training on a home decoration style building module of the home decoration customized model until convergence;
randomly extracting the new home decoration style after the increment is taken as a target home decoration style, and importing the new home decoration style into the home decoration customized model for verification;
performing imaging processing on the home decoration customization scheme output by the home decoration customization model to generate home decoration customization scheme imaging information;
comparing and judging the home decoration customization scheme imaging information with preset home decoration scheme imaging information to generate a deviation rate;
judging whether the deviation rate is smaller than a preset deviation rate threshold value or not;
if the incremental value is less than the preset value, the home decoration style is successfully constructed, and the home decoration style incremental learning is finished.
The second aspect of the present invention also provides a home decoration customization system based on artificial intelligence, which includes: the device comprises a memory and a processor, wherein the memory comprises a home customization method program based on artificial intelligence, and when the home customization method program based on artificial intelligence is executed by the processor, the method comprises the following steps:
establishing a personal-item preference model based on historical data of home decoration items browsed by a target user;
extracting personal preferences through the personal-item preference model, and formally describing the personal preferences according to time to generate a personal preference time sequence;
acquiring personal preference information in a preset interval in the personal preference time sequence;
matching the personal preference information with the information of the candidate home decoration items to acquire the interest degree of the user in the candidate home decoration items;
generating target house decoration style information of the user according to the sequencing information of the interest degree, and generating target house type information according to a target house type graph;
establishing a home decoration customization model based on a convolutional neural network, and importing the target home decoration style information and the target house type information into the home decoration customization model;
the home decoration customization model formulates a home decoration scheme according to target house type information and target home decoration style information;
displaying the home decoration scheme according to a preset mode;
the target home decoration style information comprises at least one home decoration style k, different home decoration styles i are subjected to regionalization fusion and are convolved with target house type imaging information, and then the mathematical expression form of the home decoration customized model is as follows:
Figure BDA0003179179190000051
wherein y represents the home decoration scheme information customized by the home decoration customization model, γ represents the model coefficient, i represents the total number of the home decoration styles contained in the target home decoration style information, and j represents the total number of the home decoration styles contained in the target home decoration style informationkRepresenting a home decoration style k feature area mask, gkAnd (e) representing a style code set of the home decoration style k in the target home decoration style information, representing convolution calculation, and representing the image characteristic of the target house type information.
In this scheme, the building of the home decoration customization model based on the convolutional neural network specifically includes:
establishing an initial home decoration customized model based on a convolutional neural network, and initializing model parameters;
acquiring training data, and dividing the initial home decoration customized model into an image conversion module and a home decoration style construction module;
randomly sampling house type imaging information in the training data and setting a corresponding home decoration style label;
presetting weight and gradient information of the image conversion module and the home decoration style construction module, and alternately training the image conversion module and the home decoration style construction module;
for each iteration of t times, training a home decoration style construction module for t-1 iterations, training an image conversion module for 1 iteration, and updating the weight and gradient information;
updating the parameters of an encoder and a decoder of the home decoration customized model according to the updated weight and gradient information;
and (5) after n iterations of the initial home decoration customized model are carried out, training is completed, and a home decoration customized model is generated.
In this scheme, the home decoration customization model formulates a home decoration scheme through target house type information and target home decoration style information, specifically comprising:
acquiring target house type information, splitting the target house type information, and generating imaging information of a wall body, a floor and a ceiling;
extracting the characteristics of the imaging information, and coding the characteristics through a coder in a home decoration custom model to generate image characteristics;
acquiring target home decoration style information of a user, and importing the image characteristics and the target home decoration style information into a home decoration style building module in a home decoration customization model;
performing regional fusion on different home decoration styles according to the target home decoration style information to form a filter bank;
and performing convolution on the filter group corresponding to the target home decoration style and the image characteristics, and decoding by a decoder to obtain a home decoration scheme conforming to the target home decoration style.
In this scheme, still include:
acquiring a home decoration scheme conforming to a target home decoration style, and accessing the home decoration scheme into a furniture database;
collecting interaction records of a user and home decoration items, collecting price information of the home decoration items which the user is interested in and historically purchasing the home decoration items according to the interaction records, and analyzing purchasing power of the user according to the price information;
generating a furniture recommendation label through the purchasing power and the home decoration style, and creating an index task through the furniture recommendation label by the furniture database;
an index module in the furniture database generates a recommended complete furniture scheme according to the index task and attaches a single furniture commodity link and a price in the complete furniture scheme;
meanwhile, the furniture database can adjust the recommended furniture scheme according to the budget of the user.
In this scheme, in the application process of the home decoration customized model, a new home decoration style needs to be added, and then incremental learning of the home decoration customized model on the new home decoration style specifically includes:
performing joint training on the encoder, the decoder and the multiple home decoration style sets by using a preset training strategy;
the parameters of the encoder and the decoder are fixed, and incremental training is carried out on a home decoration style set corresponding to a home decoration style building module of the home decoration customized model according to the new home decoration style;
performing repeated iterative training on a home decoration style building module of the home decoration customized model until convergence;
randomly extracting the new home decoration style after the increment is taken as a target home decoration style, and importing the new home decoration style into the home decoration customized model for verification;
performing imaging processing on the home decoration customization scheme output by the home decoration customization model to generate home decoration customization scheme imaging information;
comparing and judging the home decoration customization scheme imaging information with preset home decoration scheme imaging information to generate a deviation rate;
judging whether the deviation rate is smaller than a preset deviation rate threshold value or not;
if the incremental value is less than the preset value, the home decoration style is successfully constructed, and the home decoration style incremental learning is finished.
The third aspect of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium includes a program of an artificial intelligence based home decoration customization method, and when the program of the artificial intelligence based home decoration customization method is executed by a processor, the steps of the artificial intelligence based home decoration customization method are implemented as described in any one of the above items.
The invention discloses a home decoration customizing method, a system and a storage medium based on artificial intelligence, which comprises the following steps: acquiring target home decoration style information and target house type information of a user; establishing a home decoration customization model based on a convolutional neural network, importing the target home decoration style information into the home decoration customization model, and making a home decoration scheme by the home decoration customization model according to target house type information and target home decoration style information; displaying the home decoration scheme according to a preset mode; on one hand, the house imaging information and the house decoration style information expression are separated through the distributed training of the house decoration customized model, meanwhile, the incremental training process of the house decoration customized model has quick training time, the real-time conversion efficiency of a neural network is kept, the learning of various styles in the neural network model can be conveniently carried out, and the interference among different styles is reduced; on the other hand, the invention can also realize the switching of different house ornamentation style effects, and the encoder and the decoder adopt a symmetrical structure in the house ornamentation customized model, thereby greatly reducing the calculated amount.
Drawings
FIG. 1 is a flow chart illustrating an artificial intelligence based home customization method of the present invention;
FIG. 2 is a flow chart illustrating a method of the present invention for building a custom home decoration model based on a convolutional neural network;
FIG. 3 is a flow chart illustrating a method of generating a home appliance solution from a home appliance customization model according to the present invention;
FIG. 4 is a block diagram of an artificial intelligence based home decoration customization system of the present invention.
Detailed Description
In order that the above objects, features and advantages of the present invention can be more clearly understood, a more particular description of the invention will be rendered by reference to the appended drawings. It should be noted that the embodiments and features of the embodiments of the present application may be combined with each other without conflict.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention, however, the present invention may be practiced in other ways than those specifically described herein, and therefore the scope of the present invention is not limited by the specific embodiments disclosed below.
FIG. 1 is a flow chart of a method for customizing home decoration based on artificial intelligence according to the present invention.
As shown in fig. 1, a first aspect of the present invention provides a home decoration customization method based on artificial intelligence, including:
s102, establishing a personal-item preference model based on historical data of home decoration items browsed by a target user;
s104, extracting personal preference through the personal-item preference model, and performing formal description on the personal preference according to time to generate a personal preference time sequence;
s106, acquiring personal preference information in a preset interval in the personal preference time sequence;
s108, matching the personal preference information with the candidate home decoration item information to obtain the interest degree of the user in the candidate home decoration item;
s110, generating target house decoration style information of the user according to the sequencing information of the interest degree, and generating target house type information according to a target house type graph;
s112, establishing a home decoration customization model based on a convolutional neural network, and importing the target home decoration style information and the target house type information into the home decoration customization model;
s114, the home decoration customization model formulates a home decoration scheme through target house type information and target home decoration style information;
s116, displaying the home decoration scheme according to a preset mode;
it should be noted that, the target home decoration style information includes at least one home decoration style k, different home decoration styles i are regionally fused, and are convolved with the target house type imaging information, so that the mathematical expression form of the home decoration customized model is as follows:
Figure BDA0003179179190000091
wherein y represents the home decoration scheme information customized by the home decoration customization model, γ represents the model coefficient, i represents the total number of the home decoration styles contained in the target home decoration style information, and j represents the total number of the home decoration styles contained in the target home decoration style informationkRepresenting a home decoration style k feature area mask, gkAnd (e) representing a style code set of the home decoration style k in the target home decoration style information, representing convolution calculation, and representing the image characteristic of the target house type information.
It should be noted that browsing the home decoration item history data by the target user includes: the method comprises the steps that webpage information browsed by a user, keyword information searched by the user, home decoration commodity information purchased by the user and the like are extracted through constructing a personal-item preference model and a home decoration customization model, personal preference is extracted, and a home decoration scheme is formulated.
FIG. 2 is a flow chart of a method for building a customized home decoration model based on a convolutional neural network.
According to the embodiment of the invention, the establishment of the home decoration customized model based on the convolutional neural network specifically comprises the following steps:
s202, establishing an initial home decoration customized model based on a convolutional neural network, and initializing model parameters;
s204, acquiring training data, and dividing the initial home decoration customized model into an image conversion module and a home decoration style construction module;
s206, randomly sampling house type imaging information in the training data and setting a corresponding home decoration style label;
s208, presetting the weight and gradient information of the image conversion module and the home decoration style construction module, and alternately training the image conversion module and the home decoration style construction module;
s210, for each iteration of t times, training the home decoration style construction module for t-1 iterations, training the image conversion module for 1 iteration, and updating the weight and gradient information;
s212, updating the parameters of the encoder and the decoder of the home decoration customized model according to the updated weight and gradient information;
and S214, after n iterations of the initial home decoration customized model are carried out, training is completed, and the home decoration customized model is generated.
It should be noted that the home decoration style construction module is composed of a filter set with different home decoration styles, and is used for performing imaging processing on target home decoration style information and extracting feature images in the working process, shielding other areas of the target home decoration style imaging information, reserving relative positions corresponding to the feature images, performing convolution operation on the filters corresponding to the relative positions, collecting feature responses in the filter set, marking non-zero features, and decoding the non-zero features through a decoder to complete home decoration style construction.
It should be noted that in the application process of the home decoration customized model, a new home decoration style needs to be added, and the incremental learning of the home decoration customized model on the new home decoration style specifically includes:
performing joint training on the encoder, the decoder and the multiple home decoration style sets by using a preset training strategy;
the parameters of the encoder and the decoder are fixed, and incremental training is carried out on a home decoration style set corresponding to a home decoration style building module of the home decoration customized model according to the new home decoration style;
performing repeated iterative training on a home decoration style building module of the home decoration customized model until convergence;
randomly extracting the new home decoration style after the increment is taken as a target home decoration style, and importing the new home decoration style into the home decoration customized model for verification;
performing imaging processing on the home decoration customization scheme output by the home decoration customization model to generate home decoration customization scheme imaging information;
comparing and judging the home decoration customization scheme imaging information with preset home decoration scheme imaging information to generate a deviation rate;
judging whether the deviation rate is smaller than a preset deviation rate threshold value or not;
if the incremental value is less than the preset value, the home decoration style is successfully constructed, and the home decoration style incremental learning is finished.
It should be noted that, by means of the distributed training of the home decoration customized model, the house imaging information and the home decoration style information expression are separated, meanwhile, the incremental training of the home decoration customized model not only has fast training time, but also maintains the real-time conversion efficiency of the neural network, so that the learning of multiple styles in the neural network model can be conveniently performed, and the interference between different styles can be reduced.
FIG. 3 is a flow chart illustrating a method for generating a home appliance solution through a home appliance customization model according to the present invention.
According to the embodiment of the invention, the home decoration customization model formulates a home decoration scheme through target house type information and target home decoration style information, and specifically comprises the following steps:
s302, acquiring target house type information, splitting the target house type information, and generating imaging information of a wall body, a floor and a ceiling;
s304, extracting the characteristics of the imaging information, and coding the characteristics through a coder in the home decoration custom model to generate image characteristics;
s306, acquiring target home decoration style information of the user, and importing the image characteristics and the target home decoration style information into a home decoration style building module in a home decoration customization model;
s308, performing regional fusion on different home decoration styles according to the target home decoration style information to form a filter bank;
s310, convolving the filter group corresponding to the target home decoration style with the image characteristics, and decoding by a decoder to obtain a home decoration scheme conforming to the target home decoration style.
It should be noted that, in the personalized customization of home decoration customization, different home decoration styles are usually required to be tested when a user is unsatisfied with a recommended home decoration scheme, an encoder is fixed to obtain intermediate features of a target house, a home decoration style set performs convolution calculation on the intermediate features, a decoder decodes the intermediate features to obtain a home decoration style conversion result, and the encoder and the decoder are arranged in a symmetrical structure, so that the conversion calculation amount is greatly reduced.
It should be noted that, in the home decoration customization system of the present invention, adaptive recommendation and customization are performed by collecting environmental information during the home decoration customization process, specifically:
collecting environment information of a target house, wherein the environment information comprises floor information, light information and orientation information of each room of the house;
performing light shading analysis according to the floor information, the light information and the orientation information of each room, and adding the light shading analysis into the imaging information of the target house through a light simulation module;
judging whether the light brightness is smaller than a preset light brightness threshold value or not;
if the brightness is smaller than the preset value, recommending the user to select the bright color system as the home decoration style main tone, and adding lamp decoration in the home decoration scheme according to the light brightness.
According to the embodiment of the invention, after the home decoration customization model finishes the customization of the home decoration scheme, the furniture scheme is recommended according to the home decoration scheme, and the method specifically comprises the following steps:
generating a home decoration customization scheme according to the home decoration customization model, and accessing the home decoration customization scheme into a furniture database;
collecting interaction records of a user and home decoration items, collecting price information of the home decoration items which the user is interested in and historically purchasing the home decoration items according to the interaction records, and analyzing purchasing power of the user according to the price information;
generating a furniture recommendation label through the purchasing power and the home decoration style, and creating an index task through the furniture recommendation label by the furniture database;
an index module in the furniture database generates a recommended complete furniture scheme according to the index task and attaches a single furniture commodity link and a price in the complete furniture scheme;
meanwhile, the furniture database can adjust the recommended furniture scheme according to the budget of the user.
It should be noted that, when detecting that a target user selects an index task in a furniture database, an index module may be triggered to create at least one index sequence for indexing, the furniture database includes a plurality of database entities, the database entities may implement parameter indexing, the index module may quickly find a matching value of a furniture recommendation tag, and the database verifies index data through a verification module during the indexing task, and the verification data may be continuously collected along with generation of the verification data, and at the same time, old verification data is discarded, thereby increasing efficiency of verifying the index and improving performance of the index. The furniture database generates a recommended complete set of scheme after completing an index task, key words of single furniture commodities in the scheme are obtained, the key words are retrieved, purchasing link information of the single furniture is obtained, price information and preferential activity information of target furniture in the whole network are obtained, an optimal preferential combination is determined through the preferential activity information, final price information of the complete set of furniture scheme is generated according to the price information and the optimal preferential combination of the single furniture, and if the final price exceeds the budget of a user, department furniture in the complete set of furniture scheme is adjusted and replaced, so that the use experience of the user is improved.
It should be noted that, according to the budget information and the consumption level of the user, the customized furniture scheme and the recommended furniture scheme are adjusted to generate an optimal scheme, which specifically includes:
acquiring budget information of a user, generating a budget label, and importing the budget label into a home decoration customized model and a furniture database;
the home decoration customization model and the furniture database replace the objects in the home decoration scheme and the furniture scheme in the same style based on the target home decoration style;
meanwhile, a user uploads the existing furniture information to a furniture database, and the furniture database combines a budget label and the existing furniture style to generate a scheme of the rest needed furniture;
if the user is not satisfied with the existing furniture, the furniture database analyzes the target home decoration style through big data to generate an existing furniture transformation scheme, and the existing furniture is recycled through transformation of the existing furniture, so that home decoration budget is saved.
According to the embodiment of the invention, a part of shielding or overlapping area exists due to the view angle in the process of acquiring the target house imaging information, and the texture continuity is damaged in the process of constructing the home decoration style, wherein the parallax consistency in the process of constructing the home decoration style is maintained, and the method specifically comprises the following steps:
giving target house imaging information and imaging information after the home decoration style is built, and acquiring left and right views of the imaging information after the home decoration style is built;
calculating a disparity consistency loss from the left and right views;
according to the symmetry of the left view and the right view, defining an overlapping area of the two views as a symmetric center area;
fusing the information of the two views in the central area to generate central area parallax, central area characteristics and a shielding coefficient of the central area;
removing shielded pixels according to the forward transformation and the parallax of the central area and the shielding coefficient of the central area;
transforming the central region features into original left and right views;
and performing style construction on the overlapping area and an area which is only visible in any one of the two views, and fusing style construction results of the two parts.
FIG. 4 is a block diagram of an artificial intelligence based home decoration customization system of the present invention.
The second aspect of the present invention also provides an artificial intelligence based home decoration customization system 4, which comprises: a memory 41 and a processor 42, wherein the memory includes a program of an artificial intelligence based home decoration customization method, and when the program of the artificial intelligence based home decoration customization method is executed by the processor, the method comprises the following steps:
establishing a personal-item preference model based on historical data of home decoration items browsed by a target user;
extracting personal preferences through the personal-item preference model, and formally describing the personal preferences according to time to generate a personal preference time sequence;
acquiring personal preference information in a preset interval in the personal preference time sequence;
matching the personal preference information with the information of the candidate home decoration items to acquire the interest degree of the user in the candidate home decoration items;
generating target house decoration style information of the user according to the sequencing information of the interest degree, and generating target house type information according to a target house type graph;
establishing a home decoration customization model based on a convolutional neural network, and importing the target home decoration style information and the target house type information into the home decoration customization model;
the home decoration customization model formulates a home decoration scheme according to target house type information and target home decoration style information;
displaying the home decoration scheme according to a preset mode;
the target home decoration style information comprises at least one home decoration style k, different home decoration styles i are subjected to regionalization fusion and are convolved with target house type imaging information, and then the mathematical expression form of the home decoration customized model is as follows:
Figure BDA0003179179190000151
wherein y represents the home decoration scheme information customized by the home decoration customization model, γRepresenting model coefficients, i representing the total number of home styles contained in the target home style information, jkRepresenting a home decoration style k feature area mask, gkAnd (e) representing a style code set of the home decoration style k in the target home decoration style information, representing convolution calculation, and representing the image characteristic of the target house type information.
It should be noted that browsing the home decoration item history data by the target user includes: the method comprises the steps that webpage information browsed by a user, keyword information searched by the user, home decoration commodity information purchased by the user and the like are extracted through constructing a personal-item preference model and a home decoration customization model, personal preference is extracted, and a home decoration scheme is formulated.
According to the embodiment of the invention, the establishment of the home decoration customized model based on the convolutional neural network specifically comprises the following steps:
establishing an initial home decoration customized model based on a convolutional neural network, and initializing model parameters;
acquiring training data, and dividing the initial home decoration customized model into an image conversion module and a home decoration style construction module;
randomly sampling house type imaging information in the training data and setting a corresponding home decoration style label;
presetting weight and gradient information of the image conversion module and the home decoration style construction module, and alternately training the image conversion module and the home decoration style construction module;
for each iteration of t times, training a home decoration style construction module for t-1 iterations, training an image conversion module for 1 iteration, and updating the weight and gradient information;
updating the parameters of an encoder and a decoder of the home decoration customized model according to the updated weight and gradient information;
and (5) after n iterations of the initial home decoration customized model are carried out, training is completed, and a home decoration customized model is generated.
It should be noted that the home decoration style construction module is composed of a filter set with different home decoration styles, and is used for performing imaging processing on target home decoration style information and extracting feature images in the working process, shielding other areas of the target home decoration style imaging information, reserving relative positions corresponding to the feature images, performing convolution operation on the filters corresponding to the relative positions, collecting feature responses in the filter set, marking non-zero features, and decoding the non-zero features through a decoder to complete home decoration style construction.
It should be noted that in the application process of the home decoration customized model, a new home decoration style needs to be added, and the incremental learning of the home decoration customized model on the new home decoration style specifically includes:
performing joint training on the encoder, the decoder and the multiple home decoration style sets by using a preset training strategy;
the parameters of the encoder and the decoder are fixed, and incremental training is carried out on a home decoration style set corresponding to a home decoration style building module of the home decoration customized model according to the new home decoration style;
performing repeated iterative training on a home decoration style building module of the home decoration customized model until convergence;
randomly extracting the new home decoration style after the increment is taken as a target home decoration style, and importing the new home decoration style into the home decoration customized model for verification;
performing imaging processing on the home decoration customization scheme output by the home decoration customization model to generate home decoration customization scheme imaging information;
comparing and judging the home decoration customization scheme imaging information with preset home decoration scheme imaging information to generate a deviation rate;
judging whether the deviation rate is smaller than a preset deviation rate threshold value or not;
if the incremental value is less than the preset value, the home decoration style is successfully constructed, and the home decoration style incremental learning is finished.
It should be noted that, by means of the distributed training of the home decoration customized model, the house imaging information and the home decoration style information expression are separated, meanwhile, the incremental training of the home decoration customized model not only has fast training time, but also maintains the real-time conversion efficiency of the neural network, so that the learning of multiple styles in the neural network model can be conveniently performed, and the interference between different styles can be reduced.
According to the embodiment of the invention, the home decoration customization model formulates a home decoration scheme through target house type information and target home decoration style information, and specifically comprises the following steps:
acquiring target house type information, splitting the target house type information, and generating imaging information of a wall body, a floor and a ceiling;
extracting the characteristics of the imaging information, and coding the characteristics through a coder in a home decoration custom model to generate image characteristics;
acquiring target home decoration style information of a user, and importing the image characteristics and the target home decoration style information into a home decoration style building module in a home decoration customization model;
performing regional fusion on different home decoration styles according to the target home decoration style information to form a filter bank;
and performing convolution on the filter group corresponding to the target home decoration style and the image characteristics, and decoding by a decoder to obtain a home decoration scheme conforming to the target home decoration style.
It should be noted that, in the personalized customization of home decoration customization, different home decoration styles are usually required to be tested when a user is unsatisfied with a recommended home decoration scheme, an encoder is fixed to obtain intermediate features of a target house, a home decoration style set performs convolution calculation on the intermediate features, a decoder decodes the intermediate features to obtain a home decoration style conversion result, and the encoder and the decoder are arranged in a symmetrical structure, so that the conversion calculation amount is greatly reduced.
It should be noted that, in the home decoration customization system of the present invention, adaptive recommendation and customization are performed by collecting environmental information during the home decoration customization process, specifically:
collecting environment information of a target house, wherein the environment information comprises floor information, light information and orientation information of each room of the house;
performing light shading analysis according to the floor information, the light information and the orientation information of each room, and adding the light shading analysis into the imaging information of the target house through a light simulation module;
judging whether the light brightness is smaller than a preset light brightness threshold value or not;
if the brightness is smaller than the preset value, recommending the user to select the bright color system as the home decoration style main tone, and adding lamp decoration in the home decoration scheme according to the light brightness.
According to the embodiment of the invention, after the home decoration customization model finishes the customization of the home decoration scheme, the furniture scheme is recommended according to the home decoration scheme, and the method specifically comprises the following steps:
generating a home decoration customization scheme according to the home decoration customization model, and accessing the home decoration customization scheme into a furniture database;
collecting interaction records of a user and home decoration items, collecting price information of the home decoration items which the user is interested in and historically purchasing the home decoration items according to the interaction records, and analyzing purchasing power of the user according to the price information;
generating a furniture recommendation label through the purchasing power and the home decoration style, and creating an index task through the furniture recommendation label by the furniture database;
an index module in the furniture database generates a recommended complete furniture scheme according to the index task and attaches a single furniture commodity link and a price in the complete furniture scheme;
meanwhile, the furniture database can adjust the recommended furniture scheme according to the budget of the user.
It should be noted that, when detecting that a target user selects an index task in a furniture database, an index module may be triggered to create at least one index sequence for indexing, the furniture database includes a plurality of database entities, the database entities may implement parameter indexing, the index module may quickly find a matching value of a furniture recommendation tag, and the database verifies index data through a verification module during the indexing task, and the verification data may be continuously collected along with generation of the verification data, and at the same time, old verification data is discarded, thereby increasing efficiency of verifying the index and improving performance of the index. The furniture database generates a recommended complete set of scheme after completing an index task, key words of single furniture commodities in the scheme are obtained, the key words are retrieved, purchasing link information of the single furniture is obtained, price information and preferential activity information of target furniture in the whole network are obtained, an optimal preferential combination is determined through the preferential activity information, final price information of the complete set of furniture scheme is generated according to the price information and the optimal preferential combination of the single furniture, and if the final price exceeds the budget of a user, department furniture in the complete set of furniture scheme is adjusted and replaced, so that the use experience of the user is improved.
It should be noted that, according to the budget information and the consumption level of the user, the customized furniture scheme and the recommended furniture scheme are adjusted to generate an optimal scheme, which specifically includes:
acquiring budget information of a user, generating a budget label, and importing the budget label into a home decoration customized model and a furniture database;
the home decoration customization model and the furniture database replace the objects in the home decoration scheme and the furniture scheme in the same style based on the target home decoration style;
meanwhile, a user uploads the existing furniture information to a furniture database, and the furniture database combines a budget label and the existing furniture style to generate a scheme of the rest needed furniture;
if the user is not satisfied with the existing furniture, the furniture database analyzes the target home decoration style through big data to generate an existing furniture transformation scheme, and the existing furniture is recycled through transformation of the existing furniture, so that home decoration budget is saved.
According to the embodiment of the invention, a part of shielding or overlapping area exists due to the view angle in the process of acquiring the target house imaging information, and the texture continuity is damaged in the process of constructing the home decoration style, wherein the parallax consistency in the process of constructing the home decoration style is maintained, and the method specifically comprises the following steps:
giving target house imaging information and imaging information after the home decoration style is built, and acquiring left and right views of the imaging information after the home decoration style is built;
calculating a disparity consistency loss from the left and right views;
according to the symmetry of the left view and the right view, defining an overlapping area of the two views as a symmetric center area;
fusing the information of the two views in the central area to generate central area parallax, central area characteristics and a shielding coefficient of the central area;
removing shielded pixels according to the forward transformation and the parallax of the central area and the shielding coefficient of the central area;
transforming the central region features into original left and right views;
and performing style construction on the overlapping area and an area which is only visible in any one of the two views, and fusing style construction results of the two parts.
The third aspect of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium includes a program of an artificial intelligence based home decoration customization method, and when the program of the artificial intelligence based home decoration customization method is executed by a processor, the steps of the artificial intelligence based home decoration customization method are implemented as described in any one of the above items.
The invention discloses a home decoration customizing method, a system and a storage medium based on artificial intelligence, which comprises the following steps: acquiring target home decoration style information and target house type information of a user; establishing a home decoration customization model based on a convolutional neural network, importing the target home decoration style information into the home decoration customization model, and making a home decoration scheme by the home decoration customization model according to target house type information and target home decoration style information; displaying the home decoration scheme according to a preset mode; on one hand, the house imaging information and the house decoration style information expression are separated through the distributed training of the house decoration customized model, meanwhile, the incremental training process of the house decoration customized model has quick training time, the real-time conversion efficiency of a neural network is kept, the learning of various styles in the neural network model can be conveniently carried out, and the interference among different styles is reduced; on the other hand, the invention can also realize the switching of different house ornamentation style effects, and the encoder and the decoder adopt a symmetrical structure in the house ornamentation customized model, thereby greatly reducing the calculated amount.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described device embodiments are merely illustrative, for example, the division of the unit is only a logical functional division, and there may be other division ways in actual implementation, such as: multiple units or components may be combined, or may be integrated into another system, or some features may be omitted, or not implemented. In addition, the coupling, direct coupling or communication connection between the components shown or discussed may be through some interfaces, and the indirect coupling or communication connection between the devices or units may be electrical, mechanical or other forms.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units; can be located in one place or distributed on a plurality of network units; some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, all the functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately regarded as one unit, or two or more units may be integrated into one unit; the integrated unit can be realized in a form of hardware, or in a form of hardware plus a software functional unit.
Those of ordinary skill in the art will understand that: all or part of the steps for realizing the method embodiments can be completed by hardware related to program instructions, the program can be stored in a computer readable storage medium, and the program executes the steps comprising the method embodiments when executed; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
Alternatively, the integrated unit of the present invention may be stored in a computer-readable storage medium if it is implemented in the form of a software functional module and sold or used as a separate product. Based on such understanding, the technical solutions of the embodiments of the present invention may be essentially implemented or a part contributing to the prior art may be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, a ROM, a RAM, a magnetic or optical disk, or various other media that can store program code.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.

Claims (9)

1. A home decoration customizing method based on artificial intelligence is characterized by comprising the following steps:
establishing a personal-item preference model based on historical data of home decoration items browsed by a target user;
extracting personal preferences through the personal-item preference model, and formally describing the personal preferences according to time to generate a personal preference time sequence;
acquiring personal preference information in a preset interval in the personal preference time sequence;
matching the personal preference information with the information of the candidate home decoration items to acquire the interest degree of the user in the candidate home decoration items;
generating target house decoration style information of the user according to the sequencing information of the interest degree, and generating target house type information according to a target house type graph;
establishing a home decoration customization model based on a convolutional neural network, and importing the target home decoration style information and the target house type information into the home decoration customization model;
the home decoration customization model formulates a home decoration scheme according to target house type information and target home decoration style information;
displaying the home decoration scheme according to a preset mode;
the target home decoration style information comprises at least one home decoration style k, different home decoration styles i are subjected to regionalization fusion and are convolved with target house type imaging information, and then the mathematical expression form of the home decoration customized model is as follows:
Figure FDA0003179179180000011
wherein y represents the home decoration scheme information customized by the home decoration customization model, γ represents the model coefficient, i represents the total number of the home decoration styles contained in the target home decoration style information, and j represents the total number of the home decoration styles contained in the target home decoration style informationkRepresenting a home decoration style k feature area mask, gkAnd (e) representing a style code set of the home decoration style k in the target home decoration style information, representing convolution calculation, and representing the image characteristic of the target house type information.
2. The artificial intelligence based home decoration customization method according to claim 1, wherein the building of the home decoration customization model based on the convolutional neural network is specifically as follows:
establishing an initial home decoration customized model based on a convolutional neural network, and initializing model parameters;
acquiring training data, and dividing the initial home decoration customized model into an image conversion module and a home decoration style construction module;
randomly sampling house type imaging information in the training data and setting a corresponding home decoration style label;
presetting weight and gradient information of the image conversion module and the home decoration style construction module, and alternately training the image conversion module and the home decoration style construction module;
for each iteration of t times, training a home decoration style construction module for t-1 iterations, training an image conversion module for 1 iteration, and updating the weight and gradient information;
updating the parameters of an encoder and a decoder of the home decoration customized model according to the updated weight and gradient information;
and (5) after n iterations of the initial home decoration customized model are carried out, training is completed, and a home decoration customized model is generated.
3. The artificial intelligence based home decoration customization method according to claim 1, wherein the home decoration customization model formulates a home decoration scheme through target house layout information and target home decoration style information, and specifically comprises:
acquiring target house type information, splitting the target house type information, and generating imaging information of a wall body, a floor and a ceiling;
extracting the characteristics of the imaging information, and coding the characteristics through a coder in a home decoration custom model to generate image characteristics;
acquiring target home decoration style information of a user, and importing the image characteristics and the target home decoration style information into a home decoration style building module in a home decoration customization model;
performing regional fusion on different home decoration styles according to the target home decoration style information to form a filter bank;
and performing convolution on the filter group corresponding to the target home decoration style and the image characteristics, and decoding by a decoder to obtain a home decoration scheme conforming to the target home decoration style.
4. The method for customizing home decoration based on artificial intelligence, according to claim 1, further comprising:
acquiring a home decoration scheme conforming to a target home decoration style, and accessing the home decoration scheme into a furniture database;
collecting interaction records of a user and home decoration items, collecting price information of the home decoration items which the user is interested in and historically purchasing the home decoration items according to the interaction records, and analyzing purchasing power of the user according to the price information;
generating a furniture recommendation label through the purchasing power and the home decoration style, and creating an index task through the furniture recommendation label by the furniture database;
an index module in the furniture database generates a recommended complete furniture scheme according to the index task and attaches a single furniture commodity link and a price in the complete furniture scheme;
meanwhile, the furniture database can adjust the recommended furniture scheme according to the budget of the user.
5. The method for customizing home decoration based on artificial intelligence of claim 1, wherein in the application process of the home decoration customization model, a new home decoration style needs to be added, and then the incremental learning of the home decoration customization model on the new home decoration style specifically comprises:
performing joint training on the encoder, the decoder and the multiple home decoration style sets by using a preset training strategy;
the parameters of the encoder and the decoder are fixed, and incremental training is carried out on a home decoration style set corresponding to a home decoration style building module of the home decoration customized model according to the new home decoration style;
performing repeated iterative training on a home decoration style building module of the home decoration customized model until convergence;
randomly extracting the new home decoration style after the increment is taken as a target home decoration style, and importing the new home decoration style into the home decoration customized model for verification;
performing imaging processing on the home decoration customization scheme output by the home decoration customization model to generate home decoration customization scheme imaging information;
comparing and judging the home decoration customization scheme imaging information with preset home decoration scheme imaging information to generate a deviation rate;
judging whether the deviation rate is smaller than a preset deviation rate threshold value or not;
if the incremental value is less than the preset value, the home decoration style is successfully constructed, and the home decoration style incremental learning is finished.
6. An artificial intelligence based home decoration customization system, comprising: the device comprises a memory and a processor, wherein the memory comprises a home customization method program based on artificial intelligence, and when the home customization method program based on artificial intelligence is executed by the processor, the method comprises the following steps:
establishing a personal-item preference model based on historical data of home decoration items browsed by a target user;
extracting personal preferences through the personal-item preference model, and formally describing the personal preferences according to time to generate a personal preference time sequence;
acquiring personal preference information in a preset interval in the personal preference time sequence;
matching the personal preference information with the information of the candidate home decoration items to acquire the interest degree of the user in the candidate home decoration items;
generating target house decoration style information of the user according to the sequencing information of the interest degree, and generating target house type information according to a target house type graph;
establishing a home decoration customization model based on a convolutional neural network, and importing the target home decoration style information and the target house type information into the home decoration customization model;
the home decoration customization model formulates a home decoration scheme according to target house type information and target home decoration style information;
displaying the home decoration scheme according to a preset mode;
the target home decoration style information comprises at least one home decoration style k, different home decoration styles i are subjected to regionalization fusion and are convolved with target house type imaging information, and then the mathematical expression form of the home decoration customized model is as follows:
Figure FDA0003179179180000041
wherein y represents the home decoration scheme information customized by the home decoration customization model, γ represents the model coefficient, i represents the total number of the home decoration styles contained in the target home decoration style information, and j represents the total number of the home decoration styles contained in the target home decoration style informationkRepresenting a home decoration style k feature area mask, gkAnd (e) representing a style code set of the home decoration style k in the target home decoration style information, representing convolution calculation, and representing the image characteristic of the target house type information.
7. The artificial intelligence based home decoration customization system according to claim 6, wherein the home decoration customization model is established based on a convolutional neural network, and specifically comprises:
establishing an initial home decoration customized model based on a convolutional neural network, and initializing model parameters;
acquiring training data, and dividing the initial home decoration customized model into an image conversion module and a home decoration style construction module;
randomly sampling house type imaging information in the training data and setting a corresponding home decoration style label;
presetting weight and gradient information of the image conversion module and the home decoration style construction module, and alternately training the image conversion module and the home decoration style construction module;
for each iteration of t times, training a home decoration style construction module for t-1 iterations, training an image conversion module for 1 iteration, and updating the weight and gradient information;
updating the parameters of an encoder and a decoder of the home decoration customized model according to the updated weight and gradient information;
and (5) after n iterations of the initial home decoration customized model are carried out, training is completed, and a home decoration customized model is generated.
8. The system of claim 6, wherein the home decoration customization model formulates a home decoration scheme according to the target house type information and the target home decoration style information, and specifically comprises:
acquiring target house type information, splitting the target house type information, and generating imaging information of a wall body, a floor and a ceiling;
extracting the characteristics of the imaging information, and coding the characteristics through a coder in a home decoration custom model to generate image characteristics;
acquiring target home decoration style information of a user, and importing the image characteristics and the target home decoration style information into a home decoration style building module in a home decoration customization model;
performing regional fusion on different home decoration styles according to the target home decoration style information to form a filter bank;
and performing convolution on the filter group corresponding to the target home decoration style and the image characteristics, and decoding by a decoder to obtain a home decoration scheme conforming to the target home decoration style.
9. A computer-readable storage medium characterized by: the computer readable storage medium comprises a program of an artificial intelligence based home decoration customizing method, which when executed by a processor, implements the steps of an artificial intelligence based home decoration customizing method according to any one of claims 1 to 5.
CN202110841711.1A 2021-07-26 2021-07-26 Method, system and storage medium for customizing home decoration based on artificial intelligence Pending CN113538101A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110841711.1A CN113538101A (en) 2021-07-26 2021-07-26 Method, system and storage medium for customizing home decoration based on artificial intelligence

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110841711.1A CN113538101A (en) 2021-07-26 2021-07-26 Method, system and storage medium for customizing home decoration based on artificial intelligence

Publications (1)

Publication Number Publication Date
CN113538101A true CN113538101A (en) 2021-10-22

Family

ID=78120745

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110841711.1A Pending CN113538101A (en) 2021-07-26 2021-07-26 Method, system and storage medium for customizing home decoration based on artificial intelligence

Country Status (1)

Country Link
CN (1) CN113538101A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114442885A (en) * 2021-12-29 2022-05-06 广东三维家信息科技有限公司 Home decoration design method, device, equipment and storage medium
CN115063124A (en) * 2022-08-18 2022-09-16 江苏艾佳家居用品有限公司 Digital intelligent home decoration design method and system paying attention to long-term living demand
CN115795627A (en) * 2022-12-28 2023-03-14 广州极点三维信息科技有限公司 Furniture feature construction method, system, device and medium
CN116644503A (en) * 2023-07-21 2023-08-25 深圳小米房产网络科技有限公司 House decoration design method and system based on artificial intelligence

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108959669A (en) * 2017-05-19 2018-12-07 深圳市掌网科技股份有限公司 The Home Fashion & Design Shanghai method and system of intelligence
CN110084293A (en) * 2019-04-18 2019-08-02 贝壳技术有限公司 A kind of determination method and apparatus in complete bright pattern house
US20190251446A1 (en) * 2018-02-15 2019-08-15 Adobe Inc. Generating visually-aware item recommendations using a personalized preference ranking network
CN111553012A (en) * 2020-04-28 2020-08-18 广东博智林机器人有限公司 Home decoration design method and device, electronic equipment and storage medium
US20210019453A1 (en) * 2019-07-15 2021-01-21 Ke.Com (Beijing) Technology Co., Ltd. Artificial intelligence systems and methods for interior design
CN112651838A (en) * 2021-01-12 2021-04-13 刘晓辉 Financial transaction recommendation method and system based on artificial intelligence

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108959669A (en) * 2017-05-19 2018-12-07 深圳市掌网科技股份有限公司 The Home Fashion & Design Shanghai method and system of intelligence
US20190251446A1 (en) * 2018-02-15 2019-08-15 Adobe Inc. Generating visually-aware item recommendations using a personalized preference ranking network
CN110084293A (en) * 2019-04-18 2019-08-02 贝壳技术有限公司 A kind of determination method and apparatus in complete bright pattern house
US20210019453A1 (en) * 2019-07-15 2021-01-21 Ke.Com (Beijing) Technology Co., Ltd. Artificial intelligence systems and methods for interior design
CN111553012A (en) * 2020-04-28 2020-08-18 广东博智林机器人有限公司 Home decoration design method and device, electronic equipment and storage medium
CN112651838A (en) * 2021-01-12 2021-04-13 刘晓辉 Financial transaction recommendation method and system based on artificial intelligence

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114442885A (en) * 2021-12-29 2022-05-06 广东三维家信息科技有限公司 Home decoration design method, device, equipment and storage medium
CN115063124A (en) * 2022-08-18 2022-09-16 江苏艾佳家居用品有限公司 Digital intelligent home decoration design method and system paying attention to long-term living demand
CN115795627A (en) * 2022-12-28 2023-03-14 广州极点三维信息科技有限公司 Furniture feature construction method, system, device and medium
CN115795627B (en) * 2022-12-28 2023-09-26 广州极点三维信息科技有限公司 Furniture feature construction method, system, device and medium
CN116644503A (en) * 2023-07-21 2023-08-25 深圳小米房产网络科技有限公司 House decoration design method and system based on artificial intelligence
CN116644503B (en) * 2023-07-21 2023-12-08 深圳小米房产网络科技有限公司 House decoration design method and system based on artificial intelligence

Similar Documents

Publication Publication Date Title
CN113538101A (en) Method, system and storage medium for customizing home decoration based on artificial intelligence
CN109657156A (en) A kind of personalized recommendation method generating confrontation network based on circulation
Bahrehmand et al. Optimizing layout using spatial quality metrics and user preferences
CN106327240A (en) Recommendation method and recommendation system based on GRU neural network
CN111723292B (en) Recommendation method, system, electronic equipment and storage medium based on graph neural network
JP2004529406A (en) Method and apparatus for dynamic real-time market segmentation
CN107391582B (en) The information recommendation method of user preference similarity is calculated based on context ontology tree
CN106168980A (en) Multimedia resource recommends sort method and device
Ding et al. Product color emotional design adaptive to product shape feature variation
CN110263245B (en) Method and device for pushing object to user based on reinforcement learning model
CN108153792A (en) A kind of data processing method and relevant apparatus
Yang et al. Combining users’ cognition noise with interactive genetic algorithms and trapezoidal fuzzy numbers for product color design
CN113918832A (en) Graph convolution collaborative filtering recommendation system based on social relationship
Ahamed et al. A recommender system based on deep neural network and matrix factorization for collaborative filtering
CN111949886A (en) Sample data generation method and related device for information recommendation
CN114861050A (en) Feature fusion recommendation method and system based on neural network
Zakeri Ranking based on optimal points multi-criteria decision-making method
CN108073578B (en) Method and system for object recommendation
CN113610610B (en) Session recommendation method and system based on graph neural network and comment similarity
CN117216281A (en) Knowledge graph-based user interest diffusion recommendation method and system
CN116862454B (en) Indoor building design management method and system
CN113051468A (en) Movie recommendation method and system based on knowledge graph and reinforcement learning
CN111242654B (en) Method and system for generating advertisement picture
CN115809339A (en) Cross-domain recommendation method, system, device and storage medium
Yannou et al. Evolutionary and interactive sketching tool for innovative car shape design

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