Graph-Cut-Based Collaborative Node Embeddings for Hyperspectral Images Classification
Graph-Cut-Based Collaborative Node Embeddings for Hyperspectral Images Classification
复制标题
DOI：
10.1109/lgrs.2022.3184817
复制
发表时间：
2022
影响因子：
4.8
通讯作者：
Yuanchao Su;Mengying Jiang;Lianru Gao;Xu Sun;Xueer You;Pengfei Li
中科院分区：
文献类型：
--
作者：
Yuanchao Su;Mengying Jiang;Lianru Gao;Xu Sun;Xueer You;Pengfei Li
在处理高维数据时，节点嵌入（Node embedding， NE）有利于聚合相关性，缓解休斯现象的影响。尽管一些图神经网络可以在实现NE的过程中捕获相关性，但NE的应用仍然面临两个严峻的挑战：大量的模型参数和较差的泛化。在这篇文章中，我们提出了一种新的高光谱图像（HSI）分类方法，称为基于图形切割的协同网元（GCCNE）。具体来说，我们开发了一种基于图切的网元（GCNE）来实现低维特征表示，从而在使用图结构时避免了大量的模型参数。考虑到低维空间中的图割不需要设置锚点以减少计算量，我们采用基于随机子空间（RSs）的集成框架实现GCNE以获得协同特征集，增强了特征表示的泛化性。然后，将协同特征集分别输入到几个基于核的极限学习机（kelm）中，对像素进行分类。RSs的数量与kelm的数量相同。最后，我们获得与每个类相关联的集成结果。使用真实的HSI数据集评估了所提出方法的有效性和竞争力。
Node embedding (NE) is conducive to aggregating correlations and relieving the influence of the Hughes phenomenon when processing high-dimensional data. Although some graph neural networks can capture correlations during achieving NE, the application of NE still faces two rigorous challenges: numerous model parameters and poor generalization. In this letter, we propose a new approach for hyperspectral image (HSI) classification, called the graph-cut-based collaborative NEs (GCCNE). Specifically, we develop a graph-cut-based NE (GCNE) to achieve low-dimensional feature representation, which avoids numerous model parameters when using a graph structure. Considering that the graph cut in a low-dimensional space does not need to set anchors to decrease the calculation amount, we adopt an ensemble framework based on random subspaces (RSs) to implement the GCNE to obtain the collaborative feature sets, enhancing the generalization of feature representation. Afterward, the collaborative feature sets are input in several kernel-based extreme learning machines (KELMs), respectively, classifying pixels. The number of RSs is the same as the number of KELMs. Finally, we acquire an ensemble result associated with each class. The effectiveness and competitiveness of the proposed method are evaluated using real HSI datasets.