IEEE Trans Neural Netw Learn Syst. 2018 Mar;29(3):608-617. doi: 10.1109/TNNLS.2016.2636870. Epub 2016 Dec 29.
Data-dependent hashing has recently attracted attention due to being able to support efficient retrieval and storage of high-dimensional data, such as documents, images, and videos. In this paper, we propose a novel learning-based hashing method called "supervised discrete hashing with relaxation" (SDHR) based on "supervised discrete hashing" (SDH). SDH uses ordinary least squares regression and traditional zero-one matrix encoding of class label information as the regression target (code words), thus fixing the regression target. In SDHR, the regression target is instead optimized. The optimized regression target matrix satisfies a large margin constraint for correct classification of each example. Compared with SDH, which uses the traditional zero-one matrix, SDHR utilizes the learned regression target matrix and, therefore, more accurately measures the classification error of the regression model and is more flexible. As expected, SDHR generally outperforms SDH. Experimental results on two large-scale image data sets (CIFAR-10 and MNIST) and a large-scale and challenging face data set (FRGC) demonstrate the effectiveness and efficiency of SDHR.
近年来，基于数据的数据哈希因其能够支持高效检索和存储高维数据（如文档、图像和视频）而受到关注。在本文中，我们提出了一种基于“监督离散哈希”（SDH）的新的基于学习的哈希方法，称为“带松弛的监督离散哈希”（SDHR）。SDH 使用普通最小二乘回归和传统的零一矩阵编码类标签信息作为回归目标（码字），从而固定回归目标。在 SDHR 中，回归目标被优化。优化后的回归目标矩阵满足每个示例正确分类的大间隔约束。与使用传统零一矩阵的 SDH 相比，SDHR 利用了学习到的回归目标矩阵，因此更准确地衡量了回归模型的分类错误，并且更灵活。不出所料，SDHR 通常优于 SDH。在两个大规模图像数据集（CIFAR-10 和 MNIST）和一个大规模和具有挑战性的人脸数据集（FRGC）上的实验结果证明了 SDHR 的有效性和效率。
