Search papers, labs, and topics across Lattice.
This paper introduces Global-Neighborhood Alignment Hashing (GNAH), an innovative unsupervised cross-modal hashing method that effectively learns binary codes from a limited number of unlabeled image-text pairs. By leveraging a Prototype-Anchored Global Alignment module and a Contrastive Stochastic Neighborhood Alignment module, GNAH captures and preserves the semantic structure of vision-language models while mitigating overfitting. Experimental results show that GNAH significantly outperforms existing unsupervised methods in data-constrained scenarios, highlighting its potential for practical applications in cross-modal retrieval.
GNAH achieves superior unsupervised cross-modal retrieval performance with only a fraction of the data, transforming how we approach data efficiency in this domain.
Compared to supervised cross-modal hashing (CMH), unsupervised CMH reduces the reliance on manual labeling by learning binary codes from unlabeled image-text pairs. However, existing unsupervised CMH methods often rely on large-scale image-text pairs, which are costly to collect. To address this limitation, we propose Global-Neighborhood Alignment Hashing (GNAH), a novel approach that preserves the semantic structure of vision-language foundation models within a compact binary Hamming space using only a limited number of image-text pairs. Specifically, GNAH captures global structural information from the continuous latent space and transfers it into the binary Hamming space through a Prototype-Anchored Global Alignment module. In addition, GNAH extends conventional pairwise contrastive learning by modeling stochastic neighborhood relationships via a Contrastive Stochastic Neighborhood Alignment module, thereby alleviating overfitting to sparse pairwise correlations. Extensive experiments demonstrate that GNAH consistently outperforms existing unsupervised cross-modal retrieval methods under data-constrained settings, offering a practical solution for real-world CMH applications.
