Advances of Hyperspectral Image Classification Based on Graph Neural Networks

WAN Sheng; YANG Jian; GONG Chen

doi:10.12263/DZXB.20221295

您当前的位置：

首页 >

文章列表页 >

Advances of Hyperspectral Image Classification Based on Graph Neural Networks

SURVEYS AND REVIEWS | 更新时间：2025-12-08

- Advances of Hyperspectral Image Classification Based on Graph Neural Networks
- ACTA ELECTRONICA SINICA Vol. 51, Issue 6, Pages: 1687-1709(2023)
- 作者机构：
  
  南京理工大学计算机科学与工程学院，江苏南京 210014
- 作者简介：
- 基金信息：
  
  National Natural Science Foundation of China(61973162);Natural Science Foundation of Jiangsu Province(BZ2021013);NSF for Distinguished Young Scholar of Jiangsu Province(BK20220080);Fundamental Research Funds for the Central Universities(30920032202;30921013114)
- DOI：10.12263/DZXB.20221295
  CLC： TP751;
- Received：11 November 2022，
  
  Revised：2023-03-09，
  
  Published：25 June 2023
- 稿件说明：
移动端阅览
万升,杨健,宫辰.基于图神经网络的高光谱图像分类研究进展[J].电子学报,2023,51(06):1687-1709.

WAN Sheng,YANG Jian,GONG Chen.Advances of Hyperspectral Image Classification Based on Graph Neural Networks[J].ACTA ELECTRONICA SINICA,2023,51(06):1687-1709.
万升,杨健,宫辰.基于图神经网络的高光谱图像分类研究进展[J].电子学报,2023,51(06):1687-1709. DOI： 10.12263/DZXB.20221295.

WAN Sheng,YANG Jian,GONG Chen.Advances of Hyperspectral Image Classification Based on Graph Neural Networks[J].ACTA ELECTRONICA SINICA,2023,51(06):1687-1709. DOI： 10.12263/DZXB.20221295.

摘要

高光谱成像是遥感领域的一项先进技术，它能够收集和处理来自不同波段的电磁光谱信息，包括可见光、近红外和红外波段.由于高光谱成像技术能够检测到光谱信息的细微变化，因此，其在区分不同类型的地物方面取得了不错的成果.近年来，高光谱图像分类在城市规划和植被监测等应用中引起了许多研究者的关注，其主要目的是将图像中的每个像素分类到一个有意义的类别中.然而，高光谱图像数据量大、特征维数高的问题给像素的精确分类带来了一定挑战.如何有效提取高光谱图像的空谱特征已成为高光谱图像分类中最重要的问题之一.在过去的几年里，深度学习技术依靠强大的特征提取能力，在高光谱图像分类中具有不错的表现.其中，基于图神经网络（Graph Neural Network，GNN）的方法凭借其处理不规则数据的出色能力，为高光谱图像分类提供了新的研究方向.图神经网络是一种能够直接处理图结构数据的深度学习模型.在图神经网络模型中，每个图节点表示一个样本，每条边表示一对样本之间的关系.图卷积操作通过在图节点之间传播信息，来学习图节点的表征，从而模型能够捕捉到图节点之间的复杂关系，并实施诸如节点分类和链接预测的任务.通过将高光谱图像转换为图结构，图神经网络能够在卷积过程中提取图像的空谱特征.其中，每个节点对应于一个像素或区域，像素或区域之间的关系被表示为边.本文从图神经网络的构图方式、图卷积类型、模型架构和优化策略4个角度，梳理了当前基于图神经网络的高光谱图像分类方法的研究内容和进展，并为该领域的技术研究提供了多层次的对比分析.在基于图神经网络的高光谱图像分类方法中，图结构的构建是一项非常重要的任务.目前，主要存在两种基于高光谱数据的建图方法，即基于像素和基于区域的方法.在基于像素的建图方法中，每个像素都被视为一个图节点，这也是构建图结构最直接的方法.其中，每对图节点之间的权重可以由网络预先计算或学习.考虑到基于像素的建图方法复杂度较高，研究者开发了基于区域的建图方法.此类方法将图像的区域作为图节点，目前已被广泛应用于高光谱图像分类方法中.基于区域的建图方法可以有效减小图结构的尺寸，从而实现高效的图卷积运算.此外，图卷积类型的选择也是基于图神经网络的高光谱图像分类中的一个重要问题.当前的图卷积操作主要包含两种类型，即基于谱域的图卷积和基于空间域的图卷积.基于谱域的图卷积从图信号处理的角度引入了滤波器，通常具备较为坚实的数学基础，并且在高光谱图像分类中得到了较为广泛的使用.与基于谱域的方法不同，基于空间域的图卷积利用加权平均函数直接对每个图节点的邻居进行卷积.因此，基于空间域的图卷积通常表现出更强的灵活性和泛化能力.同时，模型架构的选择在基于图神经网络的高光谱图像分类中同样关键.目前主要存在两种类型的模型架构，即单一型和混合型.基于单一型架构的方法仅使用单一类型的深度模型（即图神经网络），早期的基于图神经网络的高光谱图像分类方法大多采用这一架构.然而，图神经网络模型本身存在一些固有的缺陷，这一定程度上限制了单一型架构方法在高光谱图像分类任务中的表现.为了缓解图神经网络模型的固有缺陷，进一步提高算法在高光谱图像分类任务中的性能，研究者开始探索将图神经网络与其他深度模型相融合的混合型模型架构，从而能够利用不同子模型来捕捉高光谱图像中多种类型的空谱特征.模型优化策略的选择对于算法性能也有着显著的影响.目前主要存在两种模型优化策略，即全批量梯度下降和小批量梯度下降.全批量梯度下降法会遍历所有样本计算一次损失函数，然后根据各个参数对应的梯度来更新模型参数，这种优化策略通常用于基于谱域的图神经网络模型.然而，由于每一次的参数更新都会涉及所有样本，因此全批量梯度下降法的复杂度通常较高，尤其是像素或超像素数量较多时.为了提高训练效率，研究者提出使用小批量梯度下降进行模型优化.其中，损失可以根据每个子图计算得到.由于每个子图的规模比原始图结构小得多，因此小批量梯度下降策略通常具有很高的效率.尽管图神经网络模型已经在高光谱图像分类任务中取得了一定成果，但现有方法仍存在一些问题有待改进.例如自适应邻域构建，即邻域的大小和形状可以根据图像的不同区域变化.现有的基于图神经网络的方法通常假设邻域大小和形状是固定不变的，而这一假设对于图像的所有区域而言可能并非最优.此外，另一个问题是可伸缩性和准确性之间的权衡.图神经网络在获得可伸缩性的同时，无可避免地会对原始数据的完整性造成损伤.然而，保持数据的完整性对模型学习是至关重要的.因此，越来越多学者开始研究能够同时保证可扩展性和高分类精度的图神经网络方法.同时，由于高光谱图像可能包含各种类型的噪声，因此噪声的处理亦是一个值得关注的问题.大部分现有的图神经网络方法对噪声不具有鲁棒性，这更加凸显了开发噪声鲁棒图神经网络方法的必要性.总的来讲，由于基于图神经网络的方法能够有效提取高光谱图像中的空谱特征，其在高光谱图像分类领域展现了巨大的潜能.本文从不同角度回顾并总结了当前基于图神经网络的高光谱图像分类方法，包括建图方式、图卷积类型、模型架构和优化策略.同时，还分析了高光谱图像分类所面临的挑战和图神经网络算法的特点，并探讨了未来潜在的研究方向.随着图神经网络方法的不断发展，高光谱图像分类有望在各个领域取得更高的精度和更广泛的应用.

Abstract

Hyperspectral imaging is an advanced imaging technique in the field of remote sensing that collects and processes information from a wide range of electromagnetic spectra

including visible

near-infrared

and infrared wavelengths. Owing to its capacity to detect subtle spectral information

hyperspectral imaging is highly effective at discriminating among different geographic objects. Therefore

hyperspectral image classification

which aims to categorize each image pixel into a certain meaningful class

has recently attracted increasing interest for real-world applications such as urban planning and vegetation monitoring. However

the very large amounts of data and high dimensionality of hyperspectral images make it challenging to classify the image pixels accurately. One of the most important issues in hyperspectral image classification is how to effectively extract the spectral—spatial features of the images. In recent years

deep learning techniques with strong feature extraction abilities have performed well in hyperspectral image classification. Among them

graph neural network (GNN)-based methods have become salient with their excellent ability to handle irregular data

providing a new research direction for hyperspectral image classification.

GNNs are a class of deep learning models that operate on graph-structured data. In GNN models

each node of the graph represents an example

and each edge represents a relationship between a pair of examples. The representations for each node can be learned by propagating information among nodes through graph convolution operations. This enables the model to capture the complex relationships between nodes and perform tasks such as node classification and link prediction. In hyperspectral image classification

GNNs can be used to learn spectral—spatial features by treating the image as a graph

where each pixel or region represents a node and the relationships between neighboring pixels or regions are represented as edges. This paper provides a multi-level comparative analysis of the research progress in GNN-based hyperspectral image classification from the perspectives of graph construction

graph convolution

model architecture

and optimization strategies.

Approaches to graph construction are an important aspect of GNN-based hyperspectral image classification. There are two main approaches to constructing a graph from a hyperspectral image

pixel-based and region-based. In pixel-based methods

each pixel is treated as a node

as this is the most straightforward way to construct a graph. The weight between each pair of nodes can be either precomputed or learned by the networks. Considering the high complexity of pixel-based methods

region-based methods have been developed

in which each region is considered a node and the size of the graph can be greatly reduced

resulting in efficient graph convolution operations. Region-based methods have recently been widely applied to hyperspectral image classification.

Additionally

the type of graph convolution is another important aspect of GNN-based hyperspectral image classification. There are two main types of graph convolution

spectral-based and spatial-based. In spectral-based methods

filters are introduced from the perspective of graph signal processing. Hence

this type of graph convolution typically has a solid mathematical foundation and has been widely adopted in hyperspectral image classification. By contrast

in spatial-based methods

a weighted average function is utilized to perform convolution over the neighbors of each node. Compared with spectral-based models

spatial-based models typically exhibit better flexibility and generalization ability.

Furthermore

selection of model architecture is crucial in GNN-based hyperspectral image classification. There exist two main types of model architectures

monotypic and hybrid. In monotypic architecture

only a single type of deep model (i.e.

graph neural network) is utilized

which was adopted by many early-stage GNN-based methods. However

performance is restricted by the inherent limitations of GNN models. To alleviate these limitations and further improve the performance of hyperspectral image classification tasks

researchers have begun exploring hybrid model architectures that combine GNN with other deep models. These hybrid models allow for the utilization of different sub-models to capture multiple types of spectral—spatial features contained in hyperspectral images.

Finally

the choice of optimization strategy also plays a critical role in the performance of GNN-based hyperspectral image classification. There are two main optimization strategies

full-batch gradient descent and mini-batch gradient descent. The full-batch gradient descent strategy

which is commonly used for spectral-based GNN models

involves computing the loss for all examples and updating model parameters based on the gradients corresponding to each parameter. However

since parameter updating involves all samples

the complexity of full-batch gradient descent can be very high

especially when dealing with a large number of pixels or superpixels. To improve training efficiency

mini-batch gradient descent can be used for model optimization. Here

the original graph structure is divided into subgraphs

and the loss of is computed for each subgraph. Since each subgraph is much smaller than the original graph

this strategy often exhibits greater efficiency.

Despite the promising results achieved by GNN-based hyperspectral image classification

some challenges remain that must be addressed. One of them is adaptive neighborhood construction

where the size and shape of the neighborhood may vary for different regions of the image. Existing GNN-based methods assume a fixed neighborhood size and shape

which may not be optimal for all regions of the image. Another challenge is the tradeoff between scalability and accuracy. GNN models inevitably compromise the integrity of the original data when achieving scalability

although preserving the integrity of the data is crucial for model learning. Therefore

there is a growing need to develop GNN-based methods that can achieve high accuracy while maintaining scalability. Noise processing is also a noteworthy issue

since hyperspectral images may contain various types of noise. Most existing GNN-based methods are not robust to noise

highlighting the need for development of noise-robust methods.

In summary

GNN-based methods have shown great potential for hyperspectral image classification by effectively extracting spectral—spatial features from hyperspectral images. In this paper

we have reviewed and summarized the existing methods from different perspectives

including graph construction

graph convolution

model architectures

and optimization strategies. We have also analyzed the challenges of hyperspectral image classification and the characteristics of GNN algorithms

providing insights into potential future research directions. With the continued development of GNN-based methods

hyperspectral image classification is expected to achieve higher accuracy and broader applicability in various fields.

关键词

Keywords

references

黄鸿 , 王丽华 , 石光耀 . 面向高光谱遥感影像分类的监督多流形鉴别嵌入方法 [J]. 电子学报 , 2020 , 48 ( 6 ): 1099 - 1107 .

HUANG H , WANG L H , SHI G Y . Supervised multi-manifold discriminant embedding method for hyperspectral remote sensing image classification [J]. Acta Electronica Sinica , 2020 , 48 ( 6 ): 1099 - 1107 . (in Chinese)

杜培军 , 夏俊士 , 薛朝辉 , 等 . 高光谱遥感影像分类研究进展 [J]. 遥感学报 , 2016 , 20 ( 2 ): 236 - 256 .

DU P J , XIA J S , XUE Z H , et al . Review of hyperspectral remote sensing image classification [J]. Journal of Remote Sensing , 2016 , 20 ( 2 ): 236 - 256 . (in Chinese)

KRUSE F A . Identification and mapping of minerals in drill core using hyperspectral image analysis of infrared reflectance spectra [J]. International Journal of Remote Sensing , 1996 , 17 ( 9 ): 1623 - 1632 .

LOBO A , GARCIA E , BARROSO G , et al . Machine learning for mineral identification and ore estimation from hyperspectral imagery in tin-tungsten deposits: Simulation under indoor conditions [J]. Remote Sensing , 2021 , 13 ( 16 ): 3258 .

张号逵 , 李映 , 姜晔楠 . 深度学习在高光谱图像分类领域的研究现状与展望 [J]. 自动化学报 , 2018 , 44 ( 6 ): 961 - 977 .

ZHANG H K , LI Y , JIANG Y N . Deep learning for hyperspectral imagery classification: The state of the art and prospects [J]. Acta Automatica Sinica , 2018 , 44 ( 6 ): 961 - 977 . (in Chinese)

SUN W W , DU Q . Hyperspectral band selection: A review [J]. IEEE Geoscience and Remote Sensing Magazine , 2019 , 7 ( 2 ): 118 - 139 .

DENG B , JIA S , SHI D M . Deep metric learning-based feature embedding for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2020 , 58 ( 2 ): 1422 - 1435 .

HARSANYI J C , CHANG C I . Hyperspectral image classification and dimensionality reduction: An orthogonal subspace projection approach [J]. IEEE Transactions on Geoscience and Remote Sensing , 1994 , 32 ( 4 ): 779 - 785 .

MA K Y , CHANG C I . Kernel-based constrained energy minimization for hyperspectral mixed pixel classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 23 .

张绍泉 , 黄志浩 , 邓承志 , 等 . 光谱加权协同稀疏和全变差正则化高光谱图像解混 [J]. 电子学报 , 2020 , 48 ( 12 ): 2453 - 2461 .

ZHANG S Q , HUANG Z H , DENG C Z , et al . Spectral reweighted collaborative sparsity and total variation based hyperspectral unmixing method [J]. Acta Electronica Sinica , 2020 , 48 ( 12 ): 2453 - 2461 . (in Chinese)

HE Z , HU J , WANG Y W . Low-rank tensor learning for classification of hyperspectral image with limited labeled samples [J]. Signal Processing , 2018 , 145 : 12 - 25 .

MA L , CRAWFORD M M , TIAN J W . Local manifold learning-based k-nearest-neighbor for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2010 , 48 ( 11 ): 4099 - 4109 .

GE H M , PAN H Z , WANG L G , et al . Self-training algorithm for hyperspectral imagery classification based on mixed measurement k-nearest neighbor and support vector machine [J]. Journal of Applied Remote Sensing , 2021 , 15 ( 4 ): 042604 .

OKWUASHI O , NDEHEDEHE C E . Deep support vector machine for hyperspectral image classification [J]. Pattern Recognition , 2020 , 103 : 107298 .

MELGANI F , BRUZZONE L . Classification of hyperspectral remote sensing images with support vector machines [J]. IEEE Transactions on Geoscience and Remote Sensing , 2004 , 42 ( 8 ): 1778 - 1790 .

王国胜 , 钟义信 . 支持向量机的若干新进展 [J]. 电子学报 , 2001 , 29 ( 10 ): 1397 - 1400 .

WANG G S , ZHONG Y X . Some new developments on support vector machine [J]. Acta Electronica Sinica , 2001 , 29 ( 10 ): 1397 - 1400 . (in Chinese)

刘建伟 , 孙正康 , 刘泽宇 , 等 . 核典型关联性分析相关特征提取与核逻辑斯蒂回归域自适应学习 [J]. 电子学报 , 2016 , 44 ( 12 ): 2908 - 2915 .

LIU J W , SUN Z K , LIU Z Y , et al . Domain adaptation learning with kernel logistic regression and kernel canonical correlation analysis [J]. Acta Electronica Sinica , 2016 , 44 ( 12 ): 2908 - 2915 . (in Chinese)

LI J , BIOUCAS-DIAS J M , PLAZA A . Semisupervised hyperspectral image segmentation using multinomial logistic regression with active learning [J]. IEEE Transactions on Geoscience and Remote Sensing , 2010 , 48 ( 11 ): 4085 - 4098 .

GHAMISI P , MAGGIORI E , LI S T , et al . New frontiers in spectral-spatial hyperspectral image classification: The latest advances based on mathematical morphology, Markov random fields, segmentation, sparse representation, and deep learning [J]. IEEE Geoscience and Remote Sensing Magazine , 2018 , 6 ( 3 ): 10 - 43 .

周浦城 , 张杰 , 薛模根 , 等 . 基于卷积分析稀疏表示和相位一致性的低照度图像增强 [J]. 电子学报 , 2020 , 48 ( 1 ): 180 - 188 .

ZHOU P C , ZHANG J , XUE M G , et al . Low-light image enhancement based on convolutional analysis sparse representation and phase congruency [J]. Acta Electronica Sinica , 2020 , 48 ( 1 ): 180 - 188 . (in Chinese)

PENG J T , LI L Q , TANG Y Y . Maximum likelihood estimation-based joint sparse representation for the classification of hyperspectral remote sensing images [J]. IEEE Transactions on Neural Networks and Learning Systems , 2019 , 30 ( 6 ): 1790 - 1802 .

WAN S , GONG C , ZHONG P , et al . Hyperspectral image classification with context-aware dynamic graph convolutional network [J]. IEEE Transactions on Geoscience and Remote Sensing , 2021 , 59 ( 1 ): 597 - 612 .

PU C Y , HUANG H , LUO L Y . Classfication of hyperspectral image with attention mechanism-based dual-path convolutional network [J]. IEEE Geoscience and Remote Sensing Letters , 2022 , 19 : 1 - 5 .

TU B , ZHOU C L , LIAO X L , et al . Spectral-spatial hyperspectral classification via structural-kernel collaborative representation [J]. IEEE Geoscience and Remote Sensing Letters , 2021 , 18 ( 5 ): 861 - 865 .

HE L , LI J , LIU C Y , et al . Recent advances on spectral-spatial hyperspectral image classification: An overview and new guidelines [J]. IEEE Transactions on Geoscience and Remote Sensing , 2018 , 56 ( 3 ): 1579 - 1597 .

CAO X Y , ZHOU F , XU L , et al . Hyperspectral image classification with Markov random fields and a convolutional neural network [J]. IEEE Transactions on Image Processing , 2018 , 27 ( 5 ): 2354 - 2367 .

王宇 , 陈殿仁 , 沈美丽 , 等 . 基于形态学梯度重构和标记提取的分水岭图像分割 [J]. 中国图象图形学报 , 2008 , 13 ( 11 ): 2176 - 2180 .

WANG Y , CHEN D R , SHEN M L , et al . Watershed segmentation based on morphological gradient reconstruction and marker extraction [J]. Journal of Image and Graphics , 2008 , 13 ( 11 ): 2176 - 2180 . (in Chinese)

GU Y F , LIU T Z , JIA X P , et al . Nonlinear multiple kernel learning with multiple-structure-element extended morphological profiles for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2016 , 54 ( 6 ): 3235 - 3247 .

BENEDIKTSSON J A , PALMASON J A , SVEINSSON J R . Classification of hyperspectral data from urban areas based on extended morphological profiles [J]. IEEE Transactions on Geoscience and Remote Sensing , 2005 , 43 ( 3 ): 480 - 491 .

LI J J , XI B B , LI Y S , et al . Hyperspectral classification based on texture feature enhancement and deep belief networks [J]. Remote Sensing , 2018 , 10 ( 3 ): 396 .

BHATTI U A , YU Z Y , CHANUSSOT J , et al . Local similarity-based spatial-spectral fusion hyperspectral image classification with deep CNN and Gabor filtering [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 15 .

刘小波 , 刘鹏 , 蔡之华 , 等 . 基于深度学习的光学遥感图像目标检测研究进展 [J]. 自动化学报 , 2021 , 47 ( 9 ): 2078 - 2089 .

LIU X B , LIU P , CAI Z H , et al . Research progress of optical remote sensing image object detection based on deep learning [J]. Acta Automatica Sinica , 2021 , 47 ( 9 ): 2078 - 2089 . (in Chinese)

YANG X F , YE Y M , LI X T , et al . Hyperspectral image classification with deep learning models [J]. IEEE Transactions on Geoscience and Remote Sensing , 2018 , 56 ( 9 ): 5408 - 5423 .

FANG L Y , LIU Z L , SONG W W . Deep hashing neural networks for hyperspectral image feature extraction [J]. IEEE Geoscience and Remote Sensing Letters , 2019 , 16 ( 9 ): 1412 - 1416 .

CHEN Y S , LIN Z H , ZHAO X , et al . Deep learning-based classification of hyperspectral data [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2014 , 7 ( 6 ): 2094 - 2107 .

MOU L C , GHAMISI P , ZHU X X . Deep recurrent neural networks for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2017 , 55 ( 7 ): 3639 - 3655 .

魏钰轩 , 陈莹 . 基于自适应层信息熵的卷积神经网络压缩 [J]. 电子学报 , 2022 , 50 ( 10 ): 2398 - 2408 .

WEI Y X , CHEN Y . Convolutional neural network compression based on adaptive layer entropy [J]. Acta Electronica Sinica , 2022 , 50 ( 10 ): 2398 - 2408 . (in Chinese)

YU S Q , JIA S , XU C Y . Convolutional neural networks for hyperspectral image classification [J]. Neurocomputing , 2017 , 219 : 88 - 98 .

JIA P Y , ZHANG M , YU W B , et al . Convolutional neural network based classification for hyperspectral data [C]// 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) . Piscataway : IEEE , 2016 : 5075 - 5078 .

MAKANTASIS K , KARANTZALOS K , DOULAMIS A , et al . Deep supervised learning for hyperspectral data classification through convolutional neural networks [C]// 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) . Piscataway : IEEE , 2015 : 4959 - 4962 .

ZHANG H K , LI Y , ZHANG Y Z , et al . Spectral-spatial classification of hyperspectral imagery using a dual-channel convolutional neural network [J]. Remote Sensing Letters , 2017 , 8 ( 5 ): 438 - 447 .

LEE H , KWON H . Going deeper with contextual CNN for hyperspectral image classification [J]. IEEE Transactions on Image Processing , 2017 , 26 ( 10 ): 4843 - 4855 .

WAN S , GONG C , ZHONG P , et al . Multiscale dynamic graph convolutional network for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2020 , 58 ( 5 ): 3162 - 3177 .

WU Z H , PAN S R , CHEN F W , et al . A comprehensive survey on graph neural networks [J]. IEEE Transactions on Neural Networks and Learning Systems , 2021 , 32 ( 1 ): 4 - 24 .

刘启超 , 肖亮 , 刘芳 , 等 . SSCDenseNet: 一种空-谱卷积稠密网络的高光谱图像分类算法 [J]. 电子学报 , 2020 , 48 ( 4 ): 751 - 762 .

LIU Q C , XIAO L , LIU F , et al . SSCDenseNet: A spectral-spatial convolutional dense network for hyperspectral image classification [J]. Acta Electronica Sinica , 2020 , 48 ( 4 ): 751 - 762 . (in Chinese)

KIPF T N , WELLING M . Semi-supervised classification with graph convolutional networks [C]// Proceedings of the International Conference on Learning Representations . Toulon : ICLR , 2017 : 1 - 14 .

QIN A Y , SHANG Z W , TIAN J Y , et al . Spectral-spatial graph convolutional networks for semisupervised hyperspectral image classification [J]. IEEE Geoscience and Remote Sensing Letters , 2019 , 16 ( 2 ): 241 - 245 .

MOU L C , LU X Q , LI X L , et al . Nonlocal graph convolutional networks for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2020 , 58 ( 12 ): 8246 - 8257 .

ZHANG S , TONG H H , XU J J , et al . Graph convolutional networks: A comprehensive review [J]. Computational Social Networks , 2019 , 6 ( 1 ): 1 - 23 .

WAN S , PAN S R , ZHONG P , et al . Dual interactive graph convolutional networks for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 14 .

WAN S , PAN S R , ZHONG S W , et al . Multi-level graph learning network for hyperspectral image classification [J]. Pattern Recognition , 2022 , 129 : 108705 .

HONG D F , GAO L R , YAO J , et al . Graph convolutional networks for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2021 , 59 ( 7 ): 5966 - 5978 .

SPERDUTI A , STARITA A . Supervised neural networks for the classification of structures [J]. IEEE Transactions on Neural Networks , 1997 , 8 ( 3 ): 714 - 735 .

GORI M , MONFARDINI G , SCARSELLI F . A new model for learning in graph domains [C]// Proceedings of 2005 IEEE International Joint Conference on Neural Networks . Piscataway : IEEE , 2005 : 729 - 734 .

SCARSELLI F , GORI M , TSOI A C , et al . The graph neural network model [J]. IEEE Transactions on Neural Networks , 2009 , 20 ( 1 ): 61 - 80 .

BRUNA J , ZAREMBA W , SZLAM A , et al . Spectral networks and locally connected networks on graphs [C]// Proceedings of the 2nd International Conference on Learning Representations . Banff : ICLR , 2014 : 1 - 14 .

DEFFERRARD M , BRESSON X , VANDERGHEYNST P . Convolutional neural networks on graphs with fast localized spectral filtering [C]// Proceedings of the 30th International Conference on Neural Information Processing Systems . New York : ACM , 2016 : 3844 - 3852 .

LI R Y , WANG S , ZHU F Y , et al . Adaptive graph convolutional neural networks [J]. Proceedings of the AAAI Conference on Artificial Intelligence , 2018 , 32 ( 1 ): 3546 - 3553 .

ZHUANG C Y , MA Q . Dual graph convolutional networks for graph-based semi-supervised classification [C]// Proceedings of the 2018 World Wide Web Conference . New York : ACM , 2018 : 499 - 508 .

XU B B , SHEN H W , CAO Q , et al . Graph wavelet neural network [C]// Proceedings of the International Conference on Learning Representations . Vancouver : ICLR , 2019 : 1 - 13 .

MICHELI A . Neural network for graphs: A contextual constructive approach [J]. IEEE Transactions on Neural Networks , 2009 , 20 ( 3 ): 498 - 511 .

BACCIU D , ERRICA F , MICHELI A . Contextual graph Markov model: A deep and generative approach to graph processing [C]// Proceedings of the 35th International Conference on Machine Learning . Stockholm : IMLS , 2018 : 294 - 303 .

ATWOOD J , TOWSLEY D . Diffusion-convolutional neural networks [C]// Proceedings of the 30th International Conference on Neural Information Processing Systems . New York : Curran Associates Inc. , 2016 : 2001 - 2009 .

TRAN D V , NAVARIN N , SPERDUTI A . On filter size in graph convolutional networks [C]// 2018 IEEE Symposium Series on Computational Intelligence (SSCI) . Piscataway : IEEE , 2019 : 1534 - 1541 .

ZHU X J . Semi-Supervised Learning Literature Survey [R]. Madison : University of Wisconsin-Madison , 2005 .

HAMILTON W L , YING R , LESKOVEC J . Inductive representation learning on large graphs [C]// Proceedings of the 31st International Conference on Neural Information Processing Systems . New York : Curran Associates Inc. , 2017 : 1025 - 1035 .

VASWANI A , SHAZEER N , PARMAR N , et al . Attention is all you need [C]// Proceedings of the 31st International Conference on Neural Information Processing Systems . New York : Curran Associates Inc. , 2017 : 6000 - 6010 .

VELIČKOVIĆ P , CUCURULL G , CASANOVA A , et al . Graph attention networks [C]// Proceedings of the International Conference on Learning Representations . Vancouver : ICLR , 2018 : 1 - 12 .

MONTI F , BOSCAINI D , MASCI J , et al . Geometric deep learning on graphs and manifolds using mixture model CNNs [C]// 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . Piscataway : IEEE , 2017 : 5425 - 5434 .

GILMER J , SCHOENHOLZ S S , RILEY P F , et al . Neural message passing for Quantum chemistry [C]// Proceedings of the 34th International Conference on Machine Learning . Sydney : JMLR.org , 2017 : 1263 - 1272 .

WANG X L , GIRSHICK R , GUPTA A , et al . Non-local neural networks [C]// 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition . Piscataway : IEEE , 2018 : 7794 - 7803 .

LIU Q C , XIAO L , YANG J X , et al . CNN-enhanced graph convolutional network with pixel- and superpixel-level feature fusion for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2021 , 59 ( 10 ): 8657 - 8671 .

DING Y , ZHAO X F , ZHANG Z L , et al . Graph sample and aggregate-attention network for hyperspectral image classification [J]. IEEE Geoscience and Remote Sensing Letters , 2022 , 19 : 1 - 5 .

BAI J , DING B X , XIAO Z , et al . Hyperspectral image classification based on deep attention graph convolutional network [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 16 .

DING Y , GUO Y Y , CHONG Y W , et al . Global consistent graph convolutional network for hyperspectral image classification [J]. IEEE Transactions on Instrumentation and Measurement , 2021 , 70 : 1 - 16 .

MA Z T , JIANG Z G , ZHANG H P . Hyperspectral image classification using feature fusion hypergraph convolution neural network [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 14 .

LIU S Y , CAO Y , WANG Y B , et al . DFL-LC: Deep feature learning with label consistencies for hyperspectral image classification [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2021 , 14 : 3669 - 3681 .

ZUO X B , YU X C , LIU B , et al . FSL-EGNN: Edge-labeling graph neural network for hyperspectral image few-shot classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 18 .

DING Y , CHONG Y W , PAN S M , et al . Spatial-spectral unified adaptive probability graph convolutional networks for hyperspectral image classification [J]. IEEE Transactions on Neural Networks and Learning Systems , 2023 , 34 ( 7 ): 3650 - 3664 .

GUO F M , LI Z W , XIN Z Q , et al . Dual graph U-nets for hyperspectral image classification [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2021 , 14 : 8160 - 8170 .

XI B B , LI J J , LI Y S , et al . Semisupervised cross-scale graph prototypical network for hyperspectral image classification [J]. IEEE Transactions on Neural Networks and Learning Systems , 2022 , DOI: 10.1109/TNNLS.2022.3158280 http://dx.doi.org/10.1109/TNNLS.2022.3158280 .

LIU Q W , DONG Y N , ZHANG Y X , et al . A fast dynamic graph convolutional network and CNN parallel network for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 15 .

LI W , WANG J J , GAO Y H , et al . Graph-feature-enhanced selective assignment network for hyperspectral and multispectral data classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 14 .

HE X , CHEN Y S , GHAMISI P . Dual graph convolutional network for hyperspectral image classification with limited training samples [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 18 .

ACHANTA R , SHAJI A , SMITH K , et al . SLIC superpixels compared to state-of-the-art superpixel methods [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence , 2012 , 34 ( 11 ): 2274 - 2282 .

DING Y , ZHAO X F , ZHANG Z L , et al . Multiscale graph sample and aggregate network with context-aware learning for hyperspectral image classification [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2021 , 14 : 4561 - 4572 .

DING Y , ZHAO X F , ZHANG Z L , et al . Semi-supervised locality preserving dense graph neural network with ARMA filters and context-aware learning for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 12 .

LIU Q C , XIAO L , YANG J X , et al . Multilevel superpixel structured graph U-nets for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 15 .

JIA S , JIANG S G , ZHANG S Y , et al . Graph-in-graph convolutional network for hyperspectral image classification [J]. IEEE Transactions on Neural Networks and Learning Systems , 2022 , DOI: 10.1109/TNNLS.2022.3182715 http://dx.doi.org/10.1109/TNNLS.2022.3182715 .

CHEN J , JIAO L C , LIU X , et al . Automatic graph learning convolutional networks for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 16 .

SHAHRAKI F F , PRASAD S . Graph convolutional neural networks for hyperspectral data classification [C]// 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP) . Piscataway : IEEE , 2019 : 968 - 972 .

YANG P , TONG L , QIAN B , et al . Hyperspectral image classification with spectral and spatial graph using inductive representation learning network [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2020 , 14 : 791 - 800 .

SELLAMI A , TABBONE S . Deep neural networks-based relevant latent representation learning for hyperspectral image classification [J]. Pattern Recognition , 2022 , 121 : 108224 .

WANG H Y , CHENG Y H , PHILIP CHEN C L , et al . Semisupervised classification of hyperspectral image based on graph convolutional broad network [J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing , 2021 , 14 : 2995 - 3005 .

XU K J , ZHAO Y , ZHANG L M , et al . Spectral-spatial residual graph attention network for hyperspectral image classification [J]. IEEE Geoscience and Remote Sensing Letters , 2022 , 19 : 1 - 5 .

PU S , SONG Y , CHEN Y , et al . Hyperspectral image classification with localized spectral filtering-based graph attention network [J]. ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences , 2022 , 3 : 155 - 161 .

CAI W W , WEI Z G . Remote sensing image classification based on a cross-attention mechanism and graph convolution [J]. IEEE Geoscience and Remote Sensing Letters , 2022 , 19 : 1 - 5 .

SHA A S , WANG B , WU X F , et al . Semisupervised classification for hyperspectral images using graph attention networks [J]. IEEE Geoscience and Remote Sensing Letters , 2021 , 18 ( 1 ): 157 - 161 .

DONG Y N , LIU Q W , DU B , et al . Weighted feature fusion of convolutional neural network and graph attention network for hyperspectral image classification [J]. IEEE Transactions on Image Processing , 2022 , 31 : 1559 - 1572 .

YANG B , CAO F L , YE H L . A novel method for hyperspectral image classification: Deep network with adaptive graph structure integration [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 12 .

QU J H , XU Y S , DONG W Q , et al . Dual-branch difference amplification graph convolutional network for hyperspectral image change detection [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 12 .

HU H J , YAO M L , HE F , et al . Graph neural network via edge convolution for hyperspectral image classification [J]. IEEE Geoscience and Remote Sensing Letters , 2022 , 19 : 1 - 5 .

ZHAO X F , NIU J H , LIU C T , et al . Hyperspectral image classification based on graph transformer network and graph attention mechanism [J]. IEEE Geoscience and Remote Sensing Letters , 2022 , 19 : 1 - 5 .

CHEN R , LI G H , DAI C L . DRGCN: Dual residual graph convolutional network for hyperspectral image classification [J]. IEEE Geoscience and Remote Sensing Letters , 2022 , 19 : 1 - 5 .

ZHANG H Y , ZOU J Q , ZHANG L P . EMS-GCN: An end-to-end mixhop superpixel-based graph convolutional network for hyperspectral image classification [J]. IEEE Transactions on Geoscience and Remote Sensing , 2022 , 60 : 1 - 16 .

WANG Y , SUN Y B , LIU Z W , et al . Dynamic graph CNN for learning on point clouds [J]. ACM Transactions on Graphics , 38 ( 5 ): 146 .

JIA Y Q , SHELHAMER E , DONAHUE J , et al . Caffe: Convolutional architecture for fast feature embedding [C]// Proceedings of the 22nd ACM international conference on Multimedia . New York : ACM , 2014 : 675 - 678 .

ABADI M , BARHAM P , CHEN J M , et al . TensorFlow: A system for large-scale machine learning [C]// Proceedings of the 12th USENIX conference on Operating Systems Design and Implementation . Berkeley : USENIX Association , 2016 : 265 - 283 .

PASZKE A , GROSS S , MASSA F , et al . PyTorch: An imperative style, high-performance deep learning library [C]// Proceedings of the 33rd International Conference on Neural Information Processing Systems . New York : Curran Associates Inc. , 2019 : 8026 - 8037 .

XI B B , LI J J , LI Y S , et al . Semi-supervised graph prototypical networks for hyperspectral image classification [C]// 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS . Piscataway : IEEE , 2021 : 2851 - 2854 .

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

RWK-GNN: Fraud Detection for Imbalanced Graphs with Feature Enhancement and Subkernel Decomposition

Variable Horizon Multi-Directional Scanning Method for Time Series Anomaly Detection

Differentially Private with Sparse and Smooth Self-Distillation

Related Author

YANG Jian

YU Hao-miao

LIU Wei

MENG Liu-chang

CHEN Kai-rui

SONG You

HUANG Yu-zhe

GUAN Yong-yuan

Related Institution

College of Computer Science， Nankai University

School of Software, Beihang University

School of Computer Science and Engineering, School of Cyber Science and Engineering, Nanjing University of Science and Technology

School of Information, Renmin University of China

College of Information Engineering, Capital Normal University

⁰