HyperBCS: Biscale Convolutional Self-Attention Network for Hyperspectral Coastal Wetlands Classification
Official implementation of HyperBCS by Junshen Luo, Zhi He, Haomei Lin, Heqian Wu.
git clone https://github.com/JeasunLok/HyperBCS.git && cd HyperBCS
conda create -n HyperBCS python=3.7
conda activate HyperBCS
pip install -r requirements.txt
pip install torch==1.13.1+cu117 -f https://download.pytorch.org/whl/torch_stable.html
Download our datasets then place them in data
folder
Baiduyun: https://pan.baidu.com/s/1hyye2fVxoUaOJ6YR_RUSJg (access code: js66)
Google Drive: https://drive.google.com/drive/folders/1jjg6Jlyb92pVrUzbdr5fHSMzYQnr2U47
Dataset MongCai
python main_argparse.py -mt HyperBCS -hbcsm 3D -e 100 -lr 5e-3 -bs 32 -d MongCai
Dataset CamPha
python main_argparse.py -mt HyperBCS -hbcsm 3D -e 100 -lr 5e-3 -bs 32 -d CamPha
python main_argparse.py -h
Please kindly cite the papers if this code is useful and helpful for your research.
J. Luo, Z. He, H. Lin and H. Wu, "Biscale Convolutional Self-Attention Network for Hyperspectral Coastal Wetlands Classification," in IEEE Geoscience and Remote Sensing Letters, vol. 21, pp. 1-5, 2024, Art no. 6002705, doi: 10.1109/LGRS.2024.3351551.
@article{luo2024biscale,
title={Biscale Convolutional Self-Attention Network for Hyperspectral Coastal Wetlands Classification},
author={Luo, Junshen and He, Zhi and Lin, Haomei and Wu, Heqian},
journal={IEEE Geoscience and Remote Sensing Letters},
year={2024},
publisher={IEEE}
}
Junshen Luo: [email protected]
Junshen Luo is with School of Geography and Planning, Sun Yat-sen University, Guangzhou 510275, China