TY - JOUR
T1 - AF2GNN
T2 - Graph convolution with adaptive filters and aggregator fusion for hyperspectral image classification
AU - Ding, Yao
AU - Zhang, Zhili
AU - Zhao, Xiaofeng
AU - Hong, Danfeng
AU - Li, Wei
AU - Cai, Wei
AU - Zhan, Ying
N1 - Publisher Copyright:
© 2022 Elsevier Inc.
PY - 2022/7
Y1 - 2022/7
N2 - Hyperspectral image classification (HSIC) is essential in remote sensing image analysis. Applying a graph neural network (GNN) to hyperspectral image (HSI) classification has attracted increasing attention. However, the available GNNs for HSIC only adopt a kind of graph filter and an aggregator, which cannot well deal with the problems of land cover discrimination, noise impaction, and spatial feature learning. To overcome these problems, a graph convolution with adaptive filters and aggregator fusion (AF2GNN) is developed for HSIC. To reduce the number of graph nodes, a superpixel segment algorithm is employed to refine the local spatial features of the HSI. A two-layer 1D CNN is proposed to transform the spectral features of superpixels. In addition, a linear function is designed to combine the different graph filters, with which the graph filter can be adaptively determined by training different weight matrices. Moreover, degree-scalers are defined to combine the multiple filters and present the graph structure. Finally, the AF2GNN is proposed to realize the adaptive filters and aggregator fusion mechanism within a single network. In the proposed network, a softMax function is utilized for graph feature interpretation and pixel-label prediction. Compared with state-of-the-art methods, the proposed method achieves superior experimental results.
AB - Hyperspectral image classification (HSIC) is essential in remote sensing image analysis. Applying a graph neural network (GNN) to hyperspectral image (HSI) classification has attracted increasing attention. However, the available GNNs for HSIC only adopt a kind of graph filter and an aggregator, which cannot well deal with the problems of land cover discrimination, noise impaction, and spatial feature learning. To overcome these problems, a graph convolution with adaptive filters and aggregator fusion (AF2GNN) is developed for HSIC. To reduce the number of graph nodes, a superpixel segment algorithm is employed to refine the local spatial features of the HSI. A two-layer 1D CNN is proposed to transform the spectral features of superpixels. In addition, a linear function is designed to combine the different graph filters, with which the graph filter can be adaptively determined by training different weight matrices. Moreover, degree-scalers are defined to combine the multiple filters and present the graph structure. Finally, the AF2GNN is proposed to realize the adaptive filters and aggregator fusion mechanism within a single network. In the proposed network, a softMax function is utilized for graph feature interpretation and pixel-label prediction. Compared with state-of-the-art methods, the proposed method achieves superior experimental results.
KW - Adaptive filters
KW - Aggregators fusion
KW - Graph neural network
KW - Hyperspectral image classification
UR - http://www.scopus.com/inward/record.url?scp=85129273072&partnerID=8YFLogxK
U2 - 10.1016/j.ins.2022.04.006
DO - 10.1016/j.ins.2022.04.006
M3 - Article
AN - SCOPUS:85129273072
SN - 0020-0255
VL - 602
SP - 201
EP - 219
JO - Information Sciences
JF - Information Sciences
ER -