System and method for social style mapping

Abstract

A system is disclosed that obtains an image of an apparel item and text describing the apparel item. The system also determines descriptive metadata of the apparel item by extracting information from the text. Further, the system further determines categorization metadata of the apparel item by analyzing the image. Additionally, the system stores the image, the descriptive metadata, and the categorization metadata in a virtual closet. Moreover, the system provides a recommendation of an outfit from a stylist to a user based on the virtual closet.

What we claim is: 
     
         1 . A method of operating a computer based assistive styling system to provide recommendations to users regarding clothing choices, wherein the method comprises:
 receiving a plurality of images from a plurality of users, the received images being of an article of clothing;   processing, using a processor, the received images to generate metadata associated with the received images, the metadata including one of a plurality of categories, where the categories are defined using machine learning;   receiving a stylist recommendation request from a user, the request including an image of an article of clothing;   comparing, using the processor, the metadata associated with the image included in the stylist recommendation request, including the category of such metadata, with metadata from other images which were previously received by the computer based assistive styling system, to find at least one match result for the image included in the stylist recommendation request;   using the at least one match result of the comparison of metadata to create a styling recommendation for the user who sent the stylist recommendation request; and   providing the styling recommendation to the user who sent the stylist recommendation request.   
     
     
         2 . The method of  claim 1 , wherein the metadata is generated by the processor and the metadata is generated using natural language processing carried out on text of descriptive information which the computer based assistive styling system receives along with the images. 
     
     
         3 . The method of  claim 2 , wherein the metadata generated using natural language processing is used by the comparison of metadata to find the at least one match result. 
     
     
         4 . The method of  claim 1 , wherein computer intelligence is used during the generation of the metadata to categorize the images based on color. 
     
     
         5 . The method of  claim 4 , wherein the use of computer intelligence in the generation of the metadata to categorize the images based on color includes determining a color model by:
 i. defining a set of color categories,   ii. providing a model color sample for one or more categories of the set of color categories, and   iii. generating a color map corresponding to one or more of the color categories using the respective color sample,   wherein the color maps define a combination of color ranges corresponding to the color categories, and one or more of the color maps is indexed by a corresponding luminance value of the respective color sample.   
     
     
         6 . The method of  claim 5 , wherein one or more color map forms a map of Cr versus Cb, where Cr is red minus luma, and Cb is blue minus luma. 
     
     
         7 . The method of  claim 6 , wherein the color of a particular item in an image is categorized by determining the color, which is derived by Cr and Cb, and a luminance value Y, of a pixel or a co-located group of several pixels in the image. 
     
     
         8 . The method of  claim 5 , wherein the color model is trained by providing reference images having known colors, Cr, Cb and luminance values Y. 
     
     
         9 . The method of  claim 5  wherein a confidence score is assigned to a selected color category based on a nearness of colors to edges of color areas in the color map. 
     
     
         10 . The method of  claim 1 , wherein the processor, based on matching module instructions, uses rules-based decisions in performing the comparison of metadata, and also uses at least one of (i) clothing items which the user already has in a wardrobe of the user, (ii) clothing items in retail catalogs, and (iii) current or seasonal fashion trends. 
     
     
         11 . The method of  claim 1 , wherein the generation of the metadata to categorize the images is performed by defining an item model by:
 i. defining a set of item categories using shape definition, machine learning or a combination of both,   ii. providing a model sample for each item category in the set of item categories, and   iii. generating a category map corresponding to each of the item categories using the respective model sample.   
     
     
         12 . The method of  claim 1 , wherein a virtual stylist provides the styling recommendation to the user who sent the styling recommendation request. 
     
     
         13 . The method of  claim 12 , wherein the virtual stylist uses information about the user, including any of body size, dimensions, preferred brands, preferred price range, wardrobe items and saved outfits, to provide the styling recommendation to the user who sent the styling recommendation request. 
     
     
         14 . The method of  claim 12 , wherein recommendation options of the virtual stylist are limited in that if a pairing of two items of clothing is to be recommended, the pairing comprises (i) an item of clothing from a wardrobe of the user and (ii) an item of clothing from a retailer's catalog. 
     
     
         15 . The method of  claim 1 , wherein the user who sent the stylist recommendation request shares the styling recommendation with other users using a social messaging scheme. 
     
     
         16 . The method of  claim 15 , wherein messaging between the user who sent the stylist recommendation request and the other users is enabled, where the user who sent the stylist recommendation request sends a message to the other users with a wish list of recommendations. 
     
     
         17 . The method of  claim 12 , wherein the virtual stylist uploads a purchase history of a particular user, into a virtual wardrobe storage in the system, for use in making the styling recommendation. 
     
     
         18 . The method of  claim 1 , wherein the image of the article of clothing sent by the user as part of the styling recommendation request is an image of an article of clothing that the user already owns, and the styling recommendation request is a request for other articles of clothing that would complement the article of clothing that the user already owns. 
     
     
         19 . The method of  claim 10 , wherein the stylist recommendation request further comprises location and event information and wherein the processor, based on matching module instructions, further uses the location and event information to create the styling recommendation for the user. 
     
     
         20 . A computer based assistive styling system comprising:
 a processor; and   a computer-readable data storage device storing program instructions that, when executed by the processor, cause the system to:   receive a plurality of images from a plurality of users, the received images being of an article of clothing;   process, the received images to generate metadata associated with the received images, the metadata including one of a plurality of categories, where the categories are defined using machine learning;   receive a stylist recommendation request from a user, the request including an image of an article of clothing;   compare the metadata associated with the image included in the stylist recommendation request, including the category of such metadata, with metadata from other images which were previously received by the computer based assistive styling system, to find at least one match result for the image included in the stylist recommendation request;   use the at least one match result of the comparison of metadata to create a styling recommendation for the user who sent the stylist recommendation request; and   provide the styling recommendation to the user who sent the stylist recommendation request.