Local Semantic-Aware Deep Hashing With Hamming-Isometric Quantization
Local Semantic-Aware Deep Hashing With Hamming-Isometric Quantization
复制标题
DOI：
10.1109/tip.2018.2889269
复制
发表时间：
2019-06
影响因子：
10.6
通讯作者：
Yunbo Wang;Jian Liang;Dong Cao;Zhenan Sun
中科院分区：
文献类型：
--
作者：
Yunbo Wang;Jian Liang;Dong Cao;Zhenan Sun
哈希是一种很有前途的方法，用于紧凑存储和高效检索大数据。与使用手工特征的传统散列方法相比，新兴的深度散列方法采用深度神经网络来学习特征表示和散列函数，这在现实世界的应用中已被证明是更强大和鲁棒的。目前，大多数现有的深度哈希方法都构建成对或三重约束，以获得一对相似数据点之间相似的二进制代码或三重组内相对相似的二进制代码。然而，我们认为，一些关键的本地结构还没有得到充分利用。为此，本文提出了一种新的深度哈希方法--局部语义感知的汉明等距量化深度哈希（LSDH），旨在充分利用哈希函数学习中的局部相似性。具体来说，潜在的语义关系被利用来鲁棒地保持汉明空间中的数据的局部相似性。除了减少二进制量化引入的误差外，设计了一个Hamming等距目标，以最大限度地提高两两类二进制特征与相应二进制码对之间的相似性的一致性，这被证明能够提高二进制码的质量。在三个单标签数据集和一个多标签数据集上的大量实验结果表明，所提出的LSDH比最新的最先进的哈希方法具有更好的性能。
Hashing is a promising approach for compact storage and efficient retrieval of big data. Compared to the conventional hashing methods using handcrafted features, emerging deep hashing approaches employ deep neural networks to learn both feature representations and hash functions, which have been proven to be more powerful and robust in real-world applications. Currently, most of the existing deep hashing methods construct pairwise or triplet-wise constraints to obtain similar binary codes between a pair of similar data points or relatively similar binary codes within a triplet. However, we argue that some critical local structures have not been fully exploited. So, this paper proposes a novel deep hashing method named local semantic-aware deep hashing with Hamming-isometric quantization (LSDH), aiming to make full use of local similarity in hash function learning. Specifically, the potential semantic relation is exploited to robustly preserve local similarity of data in the Hamming space. In addition to reducing the error introduced by binary quantizing, a Hamming-isometric objective is designed to maximize the consistency of similarity between the pairwise binary-like features and corresponding binary codes pair, which is shown to be able to improve the quality of binary codes. Extensive experimental results on several benchmark datasets, including three single-label datasets and one multi-label dataset, demonstrate that the proposed LSDH achieves better performance than the latest state-of-the-art hashing methods.