Simultaneous feature learning and hash coding with deep neural networks

TLDR

Similarity‑preserving hashing is widely used for large‑scale image retrieval, but conventional methods rely on hand‑crafted features followed by separate coding steps that can produce sub‑optimal binary codes. The authors propose a supervised deep neural network architecture that directly maps images to binary hash codes. The architecture comprises a convolutional feature extractor, a divide‑and‑encode module that assigns each branch to a hash bit, and a triplet ranking loss that enforces relative similarity. Experiments on benchmark datasets demonstrate that the simultaneous feature learning and hash coding approach significantly outperforms existing supervised and unsupervised hashing methods.

Abstract

Similarity-preserving hashing is a widely-used method for nearest neighbour search in large-scale image retrieval tasks. For most existing hashing methods, an image is first encoded as a vector of hand-engineering visual features, followed by another separate projection or quantization step that generates binary codes. However, such visual feature vectors may not be optimally compatible with the coding process, thus producing sub-optimal hashing codes. In this paper, we propose a deep architecture for supervised hashing, in which images are mapped into binary codes via carefully designed deep neural networks. The pipeline of the proposed deep architecture consists of three building blocks: 1) a sub-network with a stack of convolution layers to produce the effective intermediate image features; 2) a divide-and-encode module to divide the intermediate image features into multiple branches, each encoded into one hash bit; and 3) a triplet ranking loss designed to characterize that one image is more similar to the second image than to the third one. Extensive evaluations on several benchmark image datasets show that the proposed simultaneous feature learning and hash coding pipeline brings substantial improvements over other state-of-the-art supervised or unsupervised hashing methods.