HKNAS: Classification of Hyperspectral Imagery Based on Hyper Kernel Neural Architecture Search (TNNLS 2023)

Di Wang, Bo Du, Liangpei Zhang and Dacheng Tao

Pytorch implementation of our paper for Neural Architecture Search based hyperspectral image classification.

Fig.1 - Search Space.

Fig.2 - Search Algorithm.

Usage

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

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

3. For 3-D HK-CLS and 3-D HK-SEG, setting the 3-D convolution form in main.py
4. Search, Training, Validation, Testing and Predicion (Taka an example of WHU-Hi-HongHu dataset):

• 1-D HK-CLS

cd 1DHKCLS
CUDA_VISIBLE_DEVICES=0 python main.py --flag 'honghu' --exp_num 10 --block_num 3 --layer_num 1

• 3-D HK-CLS

cd 3DHKCLS
CUDA_VISIBLE_DEVICES=0 python main.py --flag 'honghu' --exp_num 10 --block_num 3 --layer_num 3

• 3-D HK-SEG

cd 3DHKSEG
CUDA_VISIBLE_DEVICES=0 python main.py --flag 'honghu' --exp_num 10 --block_num 3 --layer_num 1

Citation

@ARTICLE{hknas,
  author={Wang, Di and Du, Bo and Zhang, Liangpei and Tao, Dacheng},
  journal={IEEE Transactions on Neural Networks and Learning Systems}, 
  title={HKNAS: Classification of Hyperspectral Imagery Based on Hyper Kernel Neural Architecture Search}, 
  year={2023},
  volume={},
  number={},
  pages={1-15},
  doi={10.1109/TNNLS.2023.3270369}}

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] Graph Convolution based Hyperspectral Image Classification
    Spectral-Spatial Global Graph Reasoning for Hyperspectral Image Classification, IEEE TNNLS, 2023 | Paper | Github
    Di Wang^∗, Bo Du, and Liangpei Zhang

[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