This project is loosely based on the paper "Learning Fine-grained Image Similarity with Deep Ranking"(https://static.googleusercontent.com/media/research.google.com/en//pubs/archive/42945.pdf). A simpler version is used for generating the triplets by employing a different network architecture. There are three important stages in the implementation of the Deep ranking model :

a. Generation of triplets
b. Feeding the triplets into the network
c. Ranking layer