Hi,I was going through videos that explained how transfer learning works. It briefly described how the architecture of the networks can help us in separating content and style. So, I am currently trying to tackle a problem of extracting texture on an object from a image. For eg: If I have an image of a house which has a sofa in it. Now I want to compare the color/texture of the sofa with a bunch of texture files I have and determine which is the closest to it.
Eg:
This is an example image from which I would want to determine the cushion fabric the matches closest to the one in my repository
Bunch of fabrics that I have. Right now its a single image but I do have them in separate images as well.
Any pointers on how to approach this problem would be great!
Thanks!