CN110599308A - Garment recommendation system for identifying and classifying body features based on convolutional neural network - Google Patents
Garment recommendation system for identifying and classifying body features based on convolutional neural network Download PDFInfo
- Publication number
- CN110599308A CN110599308A CN201910880432.9A CN201910880432A CN110599308A CN 110599308 A CN110599308 A CN 110599308A CN 201910880432 A CN201910880432 A CN 201910880432A CN 110599308 A CN110599308 A CN 110599308A
- Authority
- CN
- China
- Prior art keywords
- clothing
- customer
- recommendation
- physical
- information
- 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.)
- Withdrawn
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- 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
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0633—Lists, e.g. purchase orders, compilation or processing
- G06Q30/0635—Processing of requisition or of purchase orders
Abstract
The invention relates to a garment recommendation system for identifying and classifying physical features based on a convolutional neural network, which comprises a customer information automatic acquisition module, a garment style recommendation module and a garment information management module; the automatic customer information acquisition module acquires a photo of a customer in a man-machine interaction mode, extracts the physical and feature characteristics of the customer in the photo by adopting an image recognition technology based on a CNN-SVM multi-classifier algorithm, and takes the physical and feature characteristics as input information of the clothing style recommendation module; the clothing style recommending module adopts expert system technology to recommend personalized clothing styles according to the physical and physical characteristics of customers and generates a clothing recommending list; and the clothing information management module completes order payment management based on the clothing recommendation list generation. The invention can complete the clothing recommendation according to the physical characteristics of the customer.
Description
Technical Field
The invention relates to the technical field of clothing recommendation, in particular to a clothing recommendation system for identifying and classifying physical features based on a convolutional neural network.
Background
The online consumption is more and more pursued by the vast consumer group along with the rapid development of economy and the rise of internet technology, the situation of single offline retail mode in the original clothing field is broken, and online shopping becomes an epoch trend. The most basic consideration when choosing clothes is whether the clothes are matched with the physical characteristics of customers, including selecting clothes with proper style according to the body types of the customers, selecting clothes with proper colors according to the skin colors of the customers and the like, which is not easy for non-clothes matching experts. In order to increase the purchasing rate of customers, strengthen purchasing behaviors of the customers and cultivate loyalty of the customers to online purchased clothes, for the sales situation of the online brand direct stores, a clothing expert recommendation system needs to be researched to replace professionals to help the customers to find clothes meeting the characteristic requirements of the customers in mass network information.
At present, personalized clothing recommendation systems popular in domestic markets, such as amazon, Taobao and the like, are mainly based on recommendation of customer relevance, and the working principle of the personalized clothing recommendation system is to mine similar customers or articles according to personal information of customer interest, browsing records, purchasing records and registration. However, these recommendation systems have problems of data sparsity, cold start and expansibility, when a new customer or article is just added into the system, the corresponding labeling information and customer score record are lacked, the evaluation matrix is sparse, the approximation between customers (articles) cannot be accurately calculated, and the recommendation accuracy is affected. Secondly, part of the recommendation systems acquire the physical feature information of the customers in a man-machine interaction mode, and the acquired physical features have artificial subjectivity and inaccuracy due to the fact that the customers cannot clearly position the external feature types of the customers. Moreover, because the customer inputs the feature information manually, the problems of unclear semantic description, inconvenient operation and the like exist, and the effect of automatically and quickly acquiring the feature data with convenient operation cannot be achieved. Since most garment recommendation systems recommend garments based on customer preferences and the level of knowledge of the customer about the garment expertise varies, the selected garment is not very professional and is not necessarily suitable for the customer from the perspective of the garment expertise, if the garments are recommended only based on customer preferences and scores, it is subjective and does not provide the customer with a suitable garment recommendation.
Disclosure of Invention
The invention aims to provide a garment recommendation system for identifying and classifying physical features based on a convolutional neural network, which can complete garment recommendation according to the physical features of customers.
The technical scheme adopted by the invention for solving the technical problems is as follows: the garment recommendation system based on the convolutional neural network and used for identifying and classifying physical features is provided and comprises a customer information automatic acquisition module, a garment style recommendation module and a garment information management module; the automatic customer information acquisition module acquires a photo of a customer in a man-machine interaction mode, extracts the physical and feature characteristics of the customer in the photo by adopting an image recognition technology based on a CNN-SVM multi-classifier algorithm, and takes the physical and feature characteristics as input information of the clothing style recommendation module; the clothing style recommending module adopts expert system technology to recommend personalized clothing styles according to the physical and physical characteristics of customers and generates a clothing recommending list; and the clothing information management module completes order payment management based on the clothing recommendation list generation.
The automatic customer information acquisition module comprises a training sample library and a customer information library, wherein the training sample library is stored
The method is used for training the sample data sets of the CNN and the SVM, continuously and automatically storing the sample data sets of new customers and updating the sample data sets to improve the correctness of automatically acquiring the physical and morphological characteristics of the customers; the customer information base is used for storing the physical features of the customers extracted from the photos.
The automatic customer information acquisition module combines CNN and SVM based on a CNN-SVM multi-classifier algorithm, and adopts a network structure improved by an AlexNet network; reducing convolution layers in an AlexNet network model of an original caffe framework, reducing network depth, adjusting initial parameter setting of a network, replacing a softmax activating function in the AlexNet network model with an SVM multi-classifier, and further improving system classification accuracy by adopting SVM classification advantages; and the CNN is trained and learned through forward and backward propagation algorithms to extract the original features of the images as the output of a full connection layer, and then the features are adopted to train the SVM multi-classifier to perform a task of identifying and classifying the physical features of the customers in the photos.
The physical characteristics of the customer comprise skin color information, face shape information, body shape information and shoulder shape information.
The clothing style recommending module comprises a rule base, a fact base and an inference machine, wherein the fact base is used for storing the physical and morphological characteristics of the customer transmitted by the customer information automatic acquisition module; the rule base stores clothing matching knowledge based on the production rule; the inference machine simulates the thinking process of clothing experts according to forward reasoning, adopts a blackboard model of a dynamic search mechanism to recommend personalized clothing styles according to the physical and morphological characteristics of customers, and generates a clothing recommendation list.
The blackboard model of the dynamic search mechanism continuously re-divides the priority of each knowledge source according to the number of the result records called by the rule action of the knowledge source in each search process so as to improve the hierarchical structure of the knowledge source with fixed sequence of the traditional blackboard model, ensure that the next knowledge source is searched in a smaller quantity space, and effectively improve the rule matching and search speed of the clothing matching recommendation system.
Advantageous effects
Due to the adoption of the technical scheme, compared with the prior art, the invention has the following advantages and positive effects:
the method adopts the CNN algorithm to obtain the appearance characteristics of the customer through the photo, and the CNN model directly extracts the characteristics of the original image, so that the link of image preprocessing is omitted, loss of texture information in the image preprocessing process is effectively prevented, the image characteristics can be implicitly extracted, the image processing has the advantages of no deformation of translation, rotation and scaling, and then an SVM multi-classifier is used for replacing the last layer of Softmax classification function of the CNN model, so that the respective defects of the CNN model and the CNN classification function are effectively overcome, and the generalization capability of the CNN model and the accuracy of the recognition and classification results are improved.
The invention provides more specialized and intelligent clothing matching recommendation for customers by applying expert system technology, helps customers find clothing matching suitable for their own characteristics, and comprises a knowledge base and an inference machine of the expert system, wherein the knowledge base and the inference machine mainly adopt a production rule form to store customer and clothing information and matching knowledge, and the inference machine adopts a blackboard model with a newly-added dynamic search mechanism.
The clothing expert recommending system based on the CNN provides personalized and specialized clothing recommendation suitable for the skin color, the face shape, the shoulder shape and the body shape of a customer through the appearance features of the customer and the matching opinions of clothing experts. The method has the advantages that the image recognition technology is adopted to objectively obtain the external appearance characteristics of the customer, so that the operations of filling personal information and grading commodities by the customer are omitted, the defect that all characteristics of the body type of the customer are difficult to completely describe by text information is overcome, the obtained external characteristic information of the customer is objective and accurate, and the influence of subjective factors of the customer is reduced.
Drawings
FIG. 1 is a schematic diagram of the overall architecture of a clothing expert recommendation system;
FIG. 2 is an illustration of a requirements analysis use case of a garment expert recommendation system;
FIG. 3 is a schematic diagram of a network architecture model;
FIG. 4 is a diagram of a network structure model based on CNN-SVM multi-classifier;
FIG. 5 is a block diagram of a framework of an automatic customer information collection module;
fig. 6 is a schematic diagram of an expert system structure of the clothing expert recommendation system.
Detailed Description
The invention will be further illustrated with reference to the following specific examples. It should be understood that these examples are for illustrative purposes only and are not intended to limit the scope of the present invention. Further, it should be understood that various changes or modifications of the present invention may be made by those skilled in the art after reading the teaching of the present invention, and such equivalents may fall within the scope of the present invention as defined in the appended claims.
The embodiment of the invention relates to a clothes recommendation system for identifying and classifying physical features based on a convolutional neural network, which comprises a customer information automatic acquisition module, a clothes style recommendation module and a clothes information management module; the automatic customer information acquisition module acquires a photo of a customer in a man-machine interaction mode, extracts the physical and feature characteristics of the customer in the photo by adopting an image recognition technology based on a CNN-SVM multi-classifier algorithm, and takes the physical and feature characteristics as input information of the clothing style recommendation module; the clothing style recommending module adopts expert system technology to recommend personalized clothing styles according to the physical and physical characteristics of customers and generates a clothing recommending list; and the clothing information management module completes order payment management based on the clothing recommendation list generation.
As shown in FIG. 1, the clothing expert recommendation system based on CNN body appearance feature recognition and classification mainly comprises three functions
The energy modules are respectively a customer information automatic acquisition module, a clothing style recommendation module and a clothing information management module. Firstly, a customer uploads a photo in an automatic system customer information acquisition module through man-machine interaction, the system automatically extracts the skin color, face shape, body shape and shoulder shape features of the customer in the photo by adopting an image recognition technology based on a CNN-SVM multi-classifier algorithm, and the information is stored in the automatic system customer information acquisition module and is used as input information of a clothing style recommendation module. Secondly, the system adopts an expert system technology to design a clothing style recommending module, which mainly comprises a rule base, a fact base and an inference machine, wherein the customer physical feature information (skin color, face shape, shoulder shape and body shape) acquired by the customer information automatic acquisition module is stored in the fact base, and the rule base stores clothing matching knowledge based on a production rule. The inference machine simulates the thinking process of the clothing expert according to forward reasoning (namely deducing rule conclusion from rule premise), adopts a blackboard model of a dynamic search mechanism to realize the function of personalized clothing style recommendation according to the physical characteristics of the customer, and provides intelligent and personalized clothing recommendation for the customer based on two aspects of different physical characteristics of the customer and the experience of the clothing matching expert. And finally, the customer completes order payment management based on the generated clothing recommendation list, if the customer is interested in the clothing recommendation list displayed by the system, the customer can add a shopping cart and fill order information to complete payment, and a system administrator checks the order and delivers the delivery goods. If the customer is not satisfied with the clothes purchase, the customer can apply for the operations of returning goods and chargeback, canceling orders and the like in the order management module, and a system administrator completes corresponding after-sales service operations.
As shown in fig. 2, the requirement analysis case of the clothing expert recommendation system is that: customer, system management
And experts for matching the costumes with the costumes. If the customer does not register the account, the account registration operation is performed; if the account is registered, the system, the clothing style recommendation, the personal information maintenance, the shopping cart management and the personal order payment and management operation can be directly logged in; the system administrator has the operations of logging in the system, customer information, clothing order information and on-sale clothing information; the clothing matching expert has a login system and clothing matching rule operation.
The network structure model of the clothing expert recommendation system shown in fig. 3 mainly adopts a three-layer structure of "browser-Web service area-database system", the browser mainly refers to a man-machine interaction page, the Web server refers to a blackboard model structure, an inference engine with a newly added dynamic search mechanism, the background database refers to a knowledge base, a rule base and a fact base, a customer sends out an HTTP request server through the browser page, the inference engine with the newly added dynamic search mechanism calls the knowledge base by JDBC to perform knowledge check and obtain recommendation results, and then the recommendation results are displayed to the customer by a standard Web page.
The network structure model based on the CNN-SVM multi-classifier shown in FIG. 4 is mainly improved by adopting AlexNet network
The latter network structure: (1) convolution layers are reduced in an AlexNet network model of an original caffe framework, network depth is reduced, and initial parameter setting of the network is adjusted. (2) And a softmax activating function in the AlexNet network model is replaced by an SVM multi-classifier, and the classification accuracy of the system is further improved by adopting the classification advantages of the SVM. The CNN is trained and learned through forward and backward propagation algorithms and used for extracting image original features as output of a full connection layer, then the features are adopted to train the SVM multi-classifier, and once training is completed, a recognition and classification task can be started on test data.
The framework structure of the automatic customer information acquisition module shown in fig. 5 mainly comprises two parts: one part is model training and verification based on a CNN-SVM multi-classifier, and as the number of customers used by the system increases, the uploaded images are updated to a training sample library to train and verify a network model; and the other part adopts a CNN-SVM-based multi-classifier model to automatically identify the uploaded photos of the customer to obtain the appearance features, and automatically judges the skin color, the face shape, the shoulder shape and the body shape of the customer. The module comprises a training sample library and a customer information library, wherein the training sample library stores sample data sets of training CNN and SVM, and meanwhile, as the number of new customers of the system increases, the sample library is continuously and automatically stored in new data sets to update the sample library, so that the correctness of the system for automatically acquiring the physical and morphological characteristics of the customers is improved. The customer information base is used for storing customer information automatically acquired by the system, and the information is used for displaying the customer information and used as input information of a subsequent clothing recommending module.
The expert system structure of the clothing expert recommendation system shown in fig. 6 mainly comprises five parts: the system mainly comprises a customer information automatic acquisition module, a knowledge base, an inference machine, a knowledge acquisition mechanism and an explanation mechanism, wherein the knowledge base mainly comprises a fact base and a rule base, the fact base is the fact information stored in a system database and mainly comprises two parts, namely, customer personal information data which mainly refers to the physical and feature characteristic information acquired by the customer information automatic acquisition module, and on-sale clothing information which mainly refers to clothing color, collar type and style. The second part is that the rule base stores the rule table of the clothing collocation knowledge. The inference machine simulates a clothing expert thinking mode to recommend clothing according to a forward inference mode, namely, based on database facts, a rule result is inferred according to a certain inference strategy or rule knowledge stored in a knowledge base, and the inference machine adopts a blackboard model with a newly added dynamic search mechanism, so that rule matching and search speed is effectively improved. And finally, the system matches the on-sale clothing information which is in accordance with the result in the fact library according to the result obtained by the inference engine, generates a final clothing recommendation list and gives an explanation which is in accordance with the clothing matching of the customer through an explanation mechanism.
The traditional blackboard model searching sequence is fixed according to the knowledge source dividing sequence, so that the result size of each layer of data can influence the efficiency and range of the next layer of data searching. The blackboard model with the new dynamic search mechanism in the embodiment is characterized in that the priority of each knowledge source is continuously divided again according to the number of result records called by the rule action of the knowledge source in each search process to improve the hierarchical structure of the knowledge source in the fixed sequence of the traditional blackboard model, so that the next-level knowledge source is ensured to search in a smaller quantity space, and the rule matching and the search speed of the clothing matching recommendation system are effectively improved.
The clothing information management module also comprises an on-sale clothing information base and a binding order database, the on-sale clothing information (namely clothing pictures, numbers, styles, shoulder shapes, stocks and the like) of stores is stored in the on-sale clothing information base, the clothing order database stores the ordering information record of customers, and when the clothing is put on the shelf or put off the shelf, a system administrator needs to update the system database. When a customer places an order for payment, cancels an order, applies for goods return and refund and the like, a system operator needs to maintain and manage in an order database, and the high-efficiency after-sale service of the whole garment recommendation system is ensured. The customer can complete order payment management based on the generated clothing recommendation list, if the customer is interested in the clothing recommendation list displayed by the system, the customer can add a shopping cart and fill order information to complete payment, and a system administrator checks orders and delivers delivery goods. If the customer is not satisfied with the clothes purchase, the customer can apply for the operations of returning goods and chargeback, canceling orders and the like in the order management module, and a system administrator completes corresponding after-sales service operations.
The invention can provide personalized and specialized clothes recommendation suitable for the skin color, face shape, shoulder shape and body shape of the customer by combining the appearance characteristics of the customer and the matching opinions of the clothing experts. The method has the advantages that the image recognition technology is adopted to objectively obtain the external appearance characteristics of the customer, so that the operations of filling personal information and grading commodities by the customer are omitted, the defect that all characteristics of the body type of the customer are difficult to completely describe by text information is overcome, the obtained external characteristic information of the customer is objective and accurate, and the influence of subjective factors of the customer is reduced.
Claims (6)
1. A garment recommendation system for identifying and classifying physical features based on a convolutional neural network is characterized by comprising a customer information automatic acquisition module, a garment style recommendation module and a garment information management module; the automatic customer information acquisition module acquires a photo of a customer in a man-machine interaction mode, extracts the physical and feature characteristics of the customer in the photo by adopting an image recognition technology based on a CNN-SVM multi-classifier algorithm, and takes the physical and feature characteristics as input information of the clothing style recommendation module; the clothing style recommending module adopts expert system technology to recommend personalized clothing styles according to the physical and physical characteristics of customers and generates a clothing recommending list; and the clothing information management module completes order payment management based on the clothing recommendation list generation.
2. The convolutional neural network-based clothing recommendation system for body appearance feature recognition and classification as claimed in claim 1, wherein the automatic customer information acquisition module comprises a training sample library and a customer information library, the training sample library stores sample data sets for training CNN and SVM, and the sample data sets of new customers are continuously and automatically stored and updated to improve the correctness of automatically acquiring the body appearance features of the customers; the customer information base is used for storing the physical features of the customers extracted from the photos.
3. The convolutional neural network-based clothing recommendation system for body feature recognition and classification as claimed in claim 2, wherein the automatic customer information acquisition module combines CNN and SVM based on CNN-SVM multi-classifier algorithm, and adopts a network structure improved by AlexNet network; reducing convolution layers in an AlexNet network model of an original caffe framework, reducing network depth, adjusting initial parameter setting of a network, replacing a softmax activating function in the AlexNet network model with an SVM multi-classifier, and further improving system classification accuracy by adopting SVM classification advantages; and the CNN is trained and learned through forward and backward propagation algorithms to extract the original features of the images as the output of a full connection layer, and then the features are adopted to train the SVM multi-classifier to perform a task of identifying and classifying the physical features of the customers in the photos.
4. The convolutional neural network-based body appearance feature recognition and classification garment recommendation system of claim 1, wherein the body appearance features of the customer comprise skin color information, face shape information, body shape information and shoulder shape information.
5. The convolutional neural network-based clothes recommendation system for body appearance feature recognition and classification as claimed in claim 1, wherein the clothes style recommendation module comprises a rule base, a fact base and an inference engine, the fact base is used for storing the body appearance features of the customers transmitted by the customer information automatic acquisition module; the rule base stores clothing matching knowledge based on the production rule; the inference machine simulates the thinking process of clothing experts according to forward reasoning, adopts a blackboard model of a dynamic search mechanism to recommend personalized clothing styles according to the physical and morphological characteristics of customers, and generates a clothing recommendation list.
6. The clothing recommendation system based on the convolutional neural network for body feature recognition and classification as claimed in claim 5, wherein the blackboard model of the dynamic search mechanism continuously re-divides the priorities of the knowledge sources according to the number of the result records called by the rule actions of the knowledge sources in each search process so as to improve the hierarchical structure of the knowledge sources in the fixed sequence of the traditional blackboard model, ensure that the next knowledge source searches in a smaller quantity space, and effectively improve the rule matching and the search speed of the clothing collocation recommendation system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910880432.9A CN110599308A (en) | 2019-09-18 | 2019-09-18 | Garment recommendation system for identifying and classifying body features based on convolutional neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910880432.9A CN110599308A (en) | 2019-09-18 | 2019-09-18 | Garment recommendation system for identifying and classifying body features based on convolutional neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110599308A true CN110599308A (en) | 2019-12-20 |
Family
ID=68860513
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910880432.9A Withdrawn CN110599308A (en) | 2019-09-18 | 2019-09-18 | Garment recommendation system for identifying and classifying body features based on convolutional neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110599308A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112232888A (en) * | 2020-11-06 | 2021-01-15 | 深圳市护家科技有限公司 | Intelligent analysis system and method for consumer behaviors |
| CN112651927A (en) * | 2020-12-03 | 2021-04-13 | 北京信息科技大学 | Raman spectrum intelligent identification method based on convolutional neural network and support vector machine |
| US11893847B1 (en) | 2022-09-23 | 2024-02-06 | Amazon Technologies, Inc. | Delivering items to evaluation rooms while maintaining customer privacy |
-
2019
- 2019-09-18 CN CN201910880432.9A patent/CN110599308A/en not_active Withdrawn
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112232888A (en) * | 2020-11-06 | 2021-01-15 | 深圳市护家科技有限公司 | Intelligent analysis system and method for consumer behaviors |
| CN112651927A (en) * | 2020-12-03 | 2021-04-13 | 北京信息科技大学 | Raman spectrum intelligent identification method based on convolutional neural network and support vector machine |
| US11893847B1 (en) | 2022-09-23 | 2024-02-06 | Amazon Technologies, Inc. | Delivering items to evaluation rooms while maintaining customer privacy |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10360623B2 (en) | Visually generated consumer product presentation | |
| CN109189904A (en) | Individuation search method and system | |
| CN108428166A (en) | The clothes commending system of figure and features feature recognition classification based on convolutional neural networks | |
| CN110110181A (en) | A kind of garment coordination recommended method based on user styles and scene preference | |
| KR20210098884A (en) | A method of providing a fashion item recommendation service using a body shape and purchase history | |
| CN110599308A (en) | Garment recommendation system for identifying and classifying body features based on convolutional neural network | |
| US20210350391A1 (en) | Methods and systems for providing a personalized user interface | |
| CN108829847A (en) | Commodity search method and system based on multi-modal shopping preferences | |
| Gharaei et al. | Content-based clothing recommender system using deep neural network | |
| Papachristou et al. | Machine learning for clothing manufacture as a mean to respond quicker and better to the demands of clothing brands: a Greek case study | |
| WO2020090054A1 (en) | Information processing system, information processing device, server device, program and method | |
| Zhu et al. | New product design with popular fashion style discovery using machine learning | |
| Turkut et al. | An online recommendation system using deep learning for textile products | |
| KR20220039697A (en) | Method, apparatus and computer program for style recommendation | |
| KR20220019737A (en) | Method, apparatus and computer program for fashion item recommendation | |
| Choi et al. | Developing an AI-based automated fashion design system: reflecting the work process of fashion designers | |
| KR20210131198A (en) | Method, apparatus and computer program for advertising recommended product | |
| KHALID et al. | Design and implementation of clothing fashion style recommendation system using deep learning. | |
| CN109146606A (en) | A kind of brand recommended method, electronic equipment, storage medium and system | |
| KR102378072B1 (en) | Method, apparatus and computer program for style recommendation | |
| Shamoi et al. | Apparel online shop reflecting customer perception | |
| KR20210063665A (en) | Recommendation item based on user event information and apparatus performing the same | |
| Wu et al. | An application of generative AI for knitted textile design in fashion | |
| JP2020071859A (en) | Information processing system, information processing apparatus, server device, program, or method | |
| US11170430B1 (en) | System, method, apparatus, and computer program product for persona based gift searches for all occasions |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20191220 |
|
| WW01 | Invention patent application withdrawn after publication |