Multi-hop double-branch network for small sample hyperspectral image classification

MuRong Shuang, Yulong Wang, Yanjun ZhangJournal of Spatial Science 2024 · DOI

Abstract

Neural networks have been extensively utilized in the classification of hyperspectral images. The study introduces a novel dual-branch structured network named MGACN for hyperspectral image classification under the condition of small sample sizes. This model a multi-hop graph network (MHGCN) and a graph attention convolutional network (GACNN). Graph Convolutional Network (GCN) integrates multi-hop mechanism to obtain and aggregate information between long-distance non-neighbor nodes. The Graph Attention Network (GAT) combines 2D-CNN by incorporating a noise reduction mechanism, as well as position and channel attention mechanisms. The classification accuracy reaches 98.43%, 99.02%, and 98.55% on the Indian Pines, Salinas, and Pavia University datasets, which is significantly better than that of the comparison model.
BibTeX
@article{shuang2024multi,
  author = {MuRong Shuang and Yulong Wang and Yanjun Zhang},
  title = {{Multi-hop double-branch network for small sample hyperspectral image classification}},
  journal = {Journal of Spatial Science},
  year = {2024},
  doi = {10.1080/14498596.2024.2350585},
}

Loading related content...

Figure