JSPEN

Deep Spatial-Spectral Joint-Sparse Prior Encoding Network for Hyperspectral Target Detection

The python code implementation of the paper "Deep Spatial-Spectral Joint-Sparse Prior Encoding Network for Hyperspectral Target Detection"

Requirements

Ubuntu 18.04 cuda 11.0
Python 3.7 Pytorch 1.11

Usage

Brief description

dataset floder stores training and testing dataset. model floder stores the pretrained model for Texas Coast dataset.

training

run the train.py include train_background and train_target

testing

run the test.py to generate the loss map

Citation

@ARTICLE{10549817,
  author={Dong, Wenqian and Wu, Xiaoyang and Qu, Jiahui and Gamba, Paolo and Xiao, Song and Vizziello, Anna and Li, Yunsong},
  journal={IEEE Transactions on Cybernetics}, 
  title={Deep Spatial—Spectral Joint-Sparse Prior Encoding Network for Hyperspectral Target Detection}, 
  year={2024},
  volume={54},
  number={12},
  pages={7780-7792},
  keywords={Object detection;Hyperspectral imaging;Adaptation models;Optimization;Training;Linear programming;Feature extraction;Hyperspectral target detection;interpretability;joint-sparse;model-encoding network},
  doi={10.1109/TCYB.2024.3403729}}