Near-Isometric Binary Hashing for Large-scale Datasets

TL;DR

This paper introduces Near-Isometric Binary Hashing (NIBH), a scalable data-dependent hashing method that minimizes worst-case distortion to improve large-scale dataset indexing and retrieval performance.

Contribution

The paper proposes a novel hashing scheme based on worst-case distortion minimization and develops an efficient algorithm that outperforms existing methods on large datasets.

Findings

NIBH achieves superior distance and ranking preservation.

NIBH outperforms ten state-of-the-art hashing schemes.

The algorithm scales well to large datasets.

Abstract

We develop a scalable algorithm to learn binary hash codes for indexing large-scale datasets. Near-isometric binary hashing (NIBH) is a data-dependent hashing scheme that quantizes the output of a learned low-dimensional embedding to obtain a binary hash code. In contrast to conventional hashing schemes, which typically rely on an $\ell_2$-norm (i.e., average distortion) minimization, NIBH is based on a $\ell_{\infty}$-norm (i.e., worst-case distortion) minimization that provides several benefits, including superior distance, ranking, and near-neighbor preservation performance. We develop a practical and efficient algorithm for NIBH based on column generation that scales well to large datasets. A range of experimental evaluations demonstrate the superiority of NIBH over ten state-of-the-art binary hashing schemes.