Transductive Zero-Shot Hashing via Coarse-to-Fine Similarity Mining

TL;DR

This paper introduces a transductive zero-shot hashing method using coarse-to-fine similarity mining, which effectively transfers knowledge from source to target classes with unlabeled data, improving hashing performance on benchmarks.

Contribution

It proposes a novel joint learning framework with a shared two-stream network and a coarse-to-fine module for transductive zero-shot hashing, addressing bias issues in prior methods.

Findings

Achieves significant improvements over state-of-the-art methods on benchmark datasets.

Effectively transfers similarities from source to target classes using unlabeled data.

Demonstrates robustness and efficiency in zero-shot hashing tasks.

Abstract

Zero-shot Hashing (ZSH) is to learn hashing models for novel/target classes without training data, which is an important and challenging problem. Most existing ZSH approaches exploit transfer learning via an intermediate shared semantic representations between the seen/source classes and novel/target classes. However, due to having disjoint, the hash functions learned from the source dataset are biased when applied directly to the target classes. In this paper, we study the transductive ZSH, i.e., we have unlabeled data for novel classes. We put forward a simple yet efficient joint learning approach via coarse-to-fine similarity mining which transfers knowledges from source data to target data. It mainly consists of two building blocks in the proposed deep architecture: 1) a shared two-streams network, which the first stream operates on the source data and the second stream operates on the unlabeled data, to learn the effective common image representations, and 2) a coarse-to-fine module, which begins with finding the most representative images from target classes and then further detect similarities among these images, to transfer the similarities of the source data to the target data in a greedy fashion. Extensive evaluation results on several benchmark datasets demonstrate that the proposed hashing method achieves significant improvement over the state-of-the-art methods.