%0 Journal Article
%T 基于质心的自适应字典学习的多视图低秩稀疏子空间聚类算法
Multi-View Low Rank Sparse Subspace Clustering Algorithm Based on Centroid Adaptive Dictionary Learning
%A 李祥
%A 由从哲
%J Computer Science and Application
%P 1297-1307
%@ 2161-881X
%D 2023
%I Hans Publishing
%R 10.12677/CSA.2023.136128
%X 随着大数据时代的到来，图像处理逐渐向高维方向发展。而高维数据通常被认为位于多个低维数据的并当中。通过将高维子空间划分为几个低维子空间，可以更好地了解高维子空间的底层结构。现有的聚类方法大多通过在每个视图上构造一个关联矩阵来解决多视图子空间聚类问题。本文提出了一种基于自适应字典学习的多视图低秩稀疏子空间聚类方法。即在多视图低秩稀疏表示模型中，引入了一种基于正交约束的自适应字典学习策略。该方法从原始数据中自适应学习字典，使模型对噪声具有鲁棒性。同时，通过优化方法得到投影矩阵和低秩稀疏特征。在本文提出的算法上对三类标准数据集进行测试，结果表明，本文所提出的算法聚类效果优于其它同类型的算法。
With the advent of the era of big data, image processing is gradually developing towards high-dimensional direction. High-dimensional data is generally considered to be in the union of multiple low-dimensional data. By dividing the high-dimensional subspace into several low-dimensional subspaces, we can better understand the underlying structure of the high-dimensional subspace. Most of the existing clustering methods solve the problem of multi-view subspace clustering by constructing an association matrix on each view. In this paper, an adaptive dictionary learning based multi-view low rank sparse subspace clustering method is proposed. In the multi-view low rank sparse representation model, an adaptive dictionary learning strategy based on orthogonal constraints is introduced. The method learns the dictionary adaptively from the original data and makes the model robust to noise. At the same time, the projection matrix and the low rank sparse feature are obtained by the optimization method. The results show that the proposed algorithm performs better than other algorithms of the same type in clustering three kinds of standard data sets.
%K 多视图，低秩表示，稀疏子空间聚类，字典学习
Multiple Views
%K Low Rank Representation
%K Sparse Subspace Clustering
%K Dictionary Learning
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=67890