Multimodal diff-hash

TL;DR

This paper introduces a simple multimodal similarity-preserving hashing algorithm that effectively maps diverse data types into Hamming space, outperforming existing methods in preserving intra- and inter-modal similarities.

Contribution

The paper proposes a novel, straightforward algorithm for multimodal hashing that improves upon state-of-the-art performance in similarity preservation.

Findings

Significantly outperforms previous methods

Effectively preserves intra- and inter-modal similarities

Efficiently maps multimodal data into Hamming space

Abstract

Many applications require comparing multimodal data with different structure and dimensionality that cannot be compared directly. Recently, there has been increasing interest in methods for learning and efficiently representing such multimodal similarity. In this paper, we present a simple algorithm for multimodal similarity-preserving hashing, trying to map multimodal data into the Hamming space while preserving the intra- and inter-modal similarities. We show that our method significantly outperforms the state-of-the-art method in the field.