FAQ.md

Image similarity

Frequently asked questions

• General
  • Which problems can be solved using image similarity, and which ones cannot?
  • When should I use image similarity and not another method?

Which problems can be solved using image similarity?

Image similarity is often used to build Image Retrieval systems where, given a query image, the goal is to find all similar images in a reference set. These systems can be used e.g. on a shopping website to suggest users comparable products, or to cluster large corporate image datasets into groups with similar content and appearance.

When should I use image similarity and not another method?

Image similarity can be used if the object-of-interest is relatively large in the image, e.g. more than 20% image width/height. If the object is smaller, then object detection methods should be used in a pre-processing step to locate and crop the image to the area of interest.