Evaluation of Hashing Methods Performance on Binary Feature Descriptors

TL;DR

This paper systematically evaluates various data-dependent hashing methods to determine their effectiveness in generating compact binary representations of 512-bit FREAK descriptors, aiming to improve efficiency in binary feature matching.

Contribution

It provides a comprehensive experimental comparison of recent unsupervised, semi-supervised, and supervised hashing methods on large labeled binary datasets.

Findings

Certain hashing methods outperform others in compactness and retrieval accuracy.

Supervised hashing shows significant improvements over unsupervised approaches.

The study identifies the most effective hashing techniques for binary feature descriptors.

Abstract

In this paper we evaluate performance of data-dependent hashing methods on binary data. The goal is to find a hashing method that can effectively produce lower dimensional binary representation of 512-bit FREAK descriptors. A representative sample of recent unsupervised, semi-supervised and supervised hashing methods was experimentally evaluated on large datasets of labelled binary FREAK feature descriptors.