Spectral-Spatial Global Graph Reasoning for Hyperspectral Image Classification (TNNLS 2023)

Di Wang, Bo Du, and Liangpei Zhang

Pytorch implementation of our paper for graph convolution based hyperspectral image classification.

Figure - The proposed SSGRN.

Usage

1. Install Pytorch 1.9 with Python 3.8.
2. Clone this repo.

git clone https://github.com/DotWang/SSGRN.git

3. Prepare a suitable GCC version, then install the SSN

cd utils/src
python setup.py install

4. Training, validation, testing and prediction with trainval.py :

For example, when implementing SSGRN on Salinas Valley dataset

 CUDA_VISIBLE_DEVICES=0 python -u trainval.py \
    --dataset 'salina' --network 'ssgrn' \
    --norm 'norm' \
    --input_mode 'whole' \
    --input_size 128 128 --overlap_size 54 \
    --experiment-num 10 --lr 1e-3 \
    --epochs 1000 --batch-size 1 \
    --val-batch-size 1 \
    --se_groups 256 --sa_groups 256

Then the evaluated accuracies, the trained models and the classification maps are separately saved.

When training on the Houston dataset, using the mode of part and setting the input_size

    --input_mode 'part' \
    --input_size 349 635 --overlap_size 0 \

Paper and Citation

If this repo is useful for your research, please cite our paper.

@ARTICLE{ssgrn,
  author={Wang, Di and Du, Bo and Zhang, Liangpei},
  journal={IEEE Transactions on Neural Networks and Learning Systems}, 
  title={Spectral-Spatial Global Graph Reasoning for Hyperspectral Image Classification}, 
  year={2024},
  volume={35},
  number={9},
  pages={12924-12937},
  keywords={Feature extraction;Cognition;Convolutional neural networks;Learning systems;Hyperspectral imaging;Training;Image classification;Adaptively;global perception;graph convolution;hyperspectral image classification (HSIC);spectral-spatial},
  doi={10.1109/TNNLS.2023.3265560}}

Thanks

GCN-Pytorch   MDGCN   SSN   SSN-Pytorch

Relevant Projects

[1] Pixel and Patch-level Hyperspectral Image Classification
    Adaptive Spectral–Spatial Multiscale Contextual Feature Extraction for Hyperspectral Image Classification, IEEE TGRS, 2020 | Paper | Github
    Di Wang^∗, Bo Du, Liangpei Zhang and Yonghao Xu

[2] Image-level/Patch-free Hyperspectral Image Classification
    Fully Contextual Network for Hyperspectral Scene Parsing, IEEE TGRS, 2021 | Paper | Github
    Di Wang^∗, Bo Du, and Liangpei Zhang

[3] Neural Architecture Search for Hyperspectral Image Classification
    HKNAS: Classification of Hyperspectral Imagery Based on Hyper Kernel Neural Architecture Search, IEEE TNNLS, 2023 | Paper | Github
    Di Wang^∗, Bo Du, Liangpei Zhang, and Dacheng Tao

[4] ImageNet Pretraining and Transformer based Hyperspectral Image Classification
    DCN-T: Dual Context Network with Transformer for Hyperspectral Image Classification, IEEE TIP, 2023 | Paper | Github
    Di Wang^∗, Jing Zhang, Bo Du, Liangpei Zhang, and Dacheng Tao