A Survey on Deep Hashing Methods

TL;DR

This survey comprehensively reviews deep hashing methods for efficient nearest neighbor search, categorizing algorithms into supervised and unsupervised types, and discusses datasets, evaluation schemes, and future research directions.

Contribution

It provides a detailed taxonomy of deep hashing algorithms, covering supervised, unsupervised, semi-supervised, domain adaptation, and multi-modal approaches, along with evaluation protocols.

Findings

Deep supervised hashing includes pairwise, ranking-based, pointwise, and quantization methods.

Deep unsupervised hashing encompasses similarity reconstruction, pseudo-label, and self-supervised learning methods.

The survey highlights key datasets and evaluation schemes used in deep hashing research.

Abstract

Nearest neighbor search aims to obtain the samples in the database with the smallest distances from them to the queries, which is a basic task in a range of fields, including computer vision and data mining. Hashing is one of the most widely used methods for its computational and storage efficiency. With the development of deep learning, deep hashing methods show more advantages than traditional methods. In this survey, we detailedly investigate current deep hashing algorithms including deep supervised hashing and deep unsupervised hashing. Specifically, we categorize deep supervised hashing methods into pairwise methods, ranking-based methods, pointwise methods as well as quantization according to how measuring the similarities of the learned hash codes. Moreover, deep unsupervised hashing is categorized into similarity reconstruction-based methods, pseudo-label-based methods and prediction-free self-supervised learning-based methods based on their semantic learning manners. We also introduce three related important topics including semi-supervised deep hashing, domain adaption deep hashing and multi-modal deep hashing. Meanwhile, we present some commonly used public datasets and the scheme to measure the performance of deep hashing algorithms. Finally, we discuss some potential research directions in conclusion.