Xi'an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi'an, China.
IEEE Trans Image Process. 2018;27(1):106-120. doi: 10.1109/TIP.2017.2755766.
Hashing has been an important and effective technology in image retrieval due to its computational efficiency and fast search speed. The traditional hashing methods usually learn hash functions to obtain binary codes by exploiting hand-crafted features, which cannot optimally represent the information of the sample. Recently, deep learning methods can achieve better performance, since deep learning architectures can learn more effective image representation features. However, these methods only use semantic features to generate hash codes by shallow projection but ignore texture details. In this paper, we proposed a novel hashing method, namely hierarchical recurrent neural hashing (HRNH), to exploit hierarchical recurrent neural network to generate effective hash codes. There are three contributions of this paper. First, a deep hashing method is proposed to extensively exploit both spatial details and semantic information, in which, we leverage hierarchical convolutional features to construct image pyramid representation. Second, our proposed deep network can exploit directly convolutional feature maps as input to preserve the spatial structure of convolutional feature maps. Finally, we propose a new loss function that considers the quantization error of binarizing the continuous embeddings into the discrete binary codes, and simultaneously maintains the semantic similarity and balanceable property of hash codes. Experimental results on four widely used data sets demonstrate that the proposed HRNH can achieve superior performance over other state-of-the-art hashing methods.Hashing has been an important and effective technology in image retrieval due to its computational efficiency and fast search speed. The traditional hashing methods usually learn hash functions to obtain binary codes by exploiting hand-crafted features, which cannot optimally represent the information of the sample. Recently, deep learning methods can achieve better performance, since deep learning architectures can learn more effective image representation features. However, these methods only use semantic features to generate hash codes by shallow projection but ignore texture details. In this paper, we proposed a novel hashing method, namely hierarchical recurrent neural hashing (HRNH), to exploit hierarchical recurrent neural network to generate effective hash codes. There are three contributions of this paper. First, a deep hashing method is proposed to extensively exploit both spatial details and semantic information, in which, we leverage hierarchical convolutional features to construct image pyramid representation. Second, our proposed deep network can exploit directly convolutional feature maps as input to preserve the spatial structure of convolutional feature maps. Finally, we propose a new loss function that considers the quantization error of binarizing the continuous embeddings into the discrete binary codes, and simultaneously maintains the semantic similarity and balanceable property of hash codes. Experimental results on four widely used data sets demonstrate that the proposed HRNH can achieve superior performance over other state-of-the-art hashing methods.
哈希技术由于其计算效率和快速搜索速度，在图像检索中是一种重要且有效的技术。传统的哈希方法通常通过利用手工制作的特征来学习哈希函数以获得二进制代码，但这些方法不能最优地表示样本的信息。最近，深度学习方法可以实现更好的性能，因为深度学习架构可以学习更有效的图像表示特征。然而，这些方法仅使用语义特征通过浅层投影生成哈希码，而忽略了纹理细节。在本文中，我们提出了一种新的哈希方法，即层次递归神经网络哈希（HRNH），以利用层次递归神经网络生成有效的哈希码。本文有三个贡献。首先，提出了一种深度哈希方法，广泛利用空间细节和语义信息，其中我们利用层次卷积特征构建图像金字塔表示。其次，我们提出的深度网络可以直接利用卷积特征图作为输入，以保留卷积特征图的空间结构。最后，我们提出了一种新的损失函数，该函数考虑了将连续嵌入量化为离散二进制代码的量化误差，同时保持哈希码的语义相似性和平衡属性。在四个广泛使用的数据集上的实验结果表明，所提出的 HRNH 可以优于其他最新的哈希方法。
