Build modern image retrieval systems using deep learning. Learn to use Siamese networks and triplet loss for embedding visual representations.