Abstract:Hashing can facilitate efficient retrieval and storage for large-scale images due to the binary representation. In the real applications, the trade-off between retrieval accuracy and speed is essential for designing a hashing framework, which is reflected by variable hash code lengths. In light of this, the existing hashing methods need to train different models for different lengths of hash codes, leading to considerable training time cost and hashing flexibility reduction. Given that a sample can be represented by various hash codes with different lengths, there are some helpful relationships that can boost the performance of hashing methods. However, the existing hashing methods do not fully utilize these relationships. To address the aforementioned issues, we propose a new model, known as supervised discrete multiple-length hashing (SDMLH), to simultaneously learn hash codes with multiple lengths. In this proposed SDMLH method, three types of information are respectively derived, from the hash codes with different lengths. The original features of the samples, and the label, are applied for hash learning. Unlike the existing hashing methods, SDMLH can fully employ the assistance among hash codes with different lengths and learn them in one step. Furthermore, given a hash length meeting the demand of users, we propose a hash fusion strategy to obtain the hash code with this desirable length by fusing the multiple-length hash codes. This obtained hash code outperforms the one learned directly. In addition, SDMLH can generate the hash code of any length that is shorter than the sum length of given multiple hash codes with the fusion strategy. To the best of our knowledge, SDMLH is one of the first attempts for learning multiple-length hash codes simultaneously. We conduct extensive experiments based on three benchmark datasets, demonstrating the superiority of this proposed method.

Deep Multiple Length Hashing Via Multi-task Learning

Multi-task Learning for Deep Semantic Hashing

Supervised Discrete Multiple-Length Hashing for Image Retrieval

Deep Collaborative Multi-View Hashing for Large-Scale Image Search

Collaborative Learning for Extremely Low Bit Asymmetric Hashing

Long-Tail Hashing

Scalable Multimedia Retrieval By Deep Learning Hashing With Relative Similarity Learning

Deep Multi-Region Hashing.

Efficient Training of Very Deep Neural Networks for Supervised Hashing

Hidden multi-distance loss-based full-convolution hashing

Learning discrete class-specific prototypes for deep semantic hashing

Large-Scale Multi-Task Image Labeling with Adaptive Relevance Discovery and Feature Hashing

Deep Hashing: A Joint Approach for Image Signature Learning.

Supervised Short-Length Hashing.

Deep Multi-Label Hashing For Large-Scale Visual Search Based On Semantic Graph

Discriminative deep metric learning for asymmetric discrete hashing

Multi-Label Deep Sparse Hashing

DHWP: Learning High-Quality Short Hash Codes Via Weight Pruning

Binary Multidimensional Scaling for Hashing

Enhanced Discrete Multi-Modal Hashing: More Constraints Yet Less Time to Learn

Deep Multimodal Hashing with Orthogonal Regularization