Efficient Nearest Neighbors Search for Large-Scale Landmark Recognition

TL;DR

This paper introduces a new multi-index hashing method called Bag of Indexes (BoI) for efficient approximate nearest neighbor search, significantly reducing query time while maintaining high accuracy in large-scale landmark recognition tasks.

Contribution

The paper presents a novel multi-index hashing approach, BoI, that improves retrieval speed and accuracy for large-scale landmark recognition compared to existing methods.

Findings

Reduces query time significantly

Outperforms state-of-the-art accuracy

Effective on multiple embedding techniques

Abstract

The problem of landmark recognition has achieved excellent results in small-scale datasets. When dealing with large-scale retrieval, issues that were irrelevant with small amount of data, quickly become fundamental for an efficient retrieval phase. In particular, computational time needs to be kept as low as possible, whilst the retrieval accuracy has to be preserved as much as possible. In this paper we propose a novel multi-index hashing method called Bag of Indexes (BoI) for Approximate Nearest Neighbors (ANN) search. It allows to drastically reduce the query time and outperforms the accuracy results compared to the state-of-the-art methods for large-scale landmark recognition. It has been demonstrated that this family of algorithms can be applied on different embedding techniques like VLAD and R-MAC obtaining excellent results in very short times on different public datasets: Holidays+Flickr1M, Oxford105k and Paris106k.