Efficient end-to-end learning for quantizable representations

Yeonwoo Jeong, Hyun Oh Song

Abstract: Embedding representation learning via neural networks is at the core foundation of modern similarity based search. While much effort has been put in developing algorithms for learning binary hammin...