Multiscale Dynamic Graph Convolutional Network for Hyperspectral Image Classification
Xidian University · Nanjing University of Information Science and Technology · +3 more institutions
Abstract
Convolutional neural network (CNN) has demonstrated impressive ability to represent hyperspectral images and to achieve promising results in hyperspectral image classification. However, traditional CNN models can only operate convolution on regular square image regions with fixed size and weights, and thus, they cannot universally adapt to the distinct local regions with various object distributions and geometric appearances. Therefore, their classification performances are still to be improved, especially in class boundaries. To alleviate this shortcoming, we consider employing the recently proposed graph convolutional network (GCN) for hyperspectral image classification, as it can conduct the convolution on…
Citation impact
- FWCI
- 35.88
- Percentile
- 100%
- References
- 91
Authors
6- SWSheng WanCorresponding
Xidian University, Nanjing University of Information Science and Technology, Nanjing University of Science and Technology
- CGChen Gong
Xidian University, Nanjing University of Information Science and Technology, Nanjing University of Science and Technology
- PZPing Zhong
National University of Defense Technology
- BDBo Du
Wuhan University
- LZLefei Zhang
Wuhan University
Topics & keywords
- Hyperspectral imaging
- Pattern recognition (psychology)
- Computer science
- Discriminative model
- Convolutional neural network
- Graph
- Artificial intelligence
- Convolution (computer science)
- Reduced inequalities
Funding
- CAChina Association for Science and TechnologyAward: 2018QNRC001
- NNNational Natural Science Foundation of ChinaAwards: 61671456, 61602246, 61971428, U1713208, 61973162
- NSNatural Science Foundation of Jiangsu ProvinceAward: BK20171430
- FRFundamental Research Funds for the Central UniversitiesAward: 30918011319