lucas2606-rs / enl-fcn

ENL-FCN for HSI classification

Efficient Deep Learning of Non-local Features for Hyperspectral Image Classification [arXiv]

Introduction

This work focuses on integrating long-range contextual information for HSI classification. Concretely, the efficient non-local module is embedded in the FCN as a learning unit to capture the long-range contextual information with computation efficiency.

Requirements

• Ununtu 18.0
• python 3.7
• Pytorch 1.4

Training and Testing

put HSI dataset in Datasets folder
run the Demo_IP.py for Indian Pines dataset training and testing

Citation

@article{shen2020,
  title={Efficient Deep Learning of Non-local Features for Hyperspectral Image Classification},
  author={Shen, Yu and Zhu, Sijie and Chen, Chen and Xiao, Liang and Du, Qian and Chen, Jianyu and Pan, Delu},
  journal={IEEE Transactions on Geoscience and Remote Sensing},
  year={2020}
}

Implementation

This work is implemented based on CCNet.