yyyyll0ss/HSI-VecNet

[TGRS 2023] Toward the Vectorization of the Hyperspectral Imagery

HSI-VecNet

This is Pytorch implementation for "Toward the Vectorization of the Hyperspectral Imagery". If you have any questions, please contact yanyl@hnu.edu.cn

Overview

The overall framework for HSI-VecNet.

Running the code

1. Run train.py to train a new model.

• Please put the hyperspectral vectorization datasets into datasets folder.

• The datasets folder is constructed as follows:

  -- indian_pines

  ---- indian_pines.mat

  ---- indian_pines_gt.mat

  ---- indian_pines_coco.json

2. Run inference.py to inference the vectorization results.

3. Run evaluation.py to evaluate the vectorization performance.

• The datasets can be downloaded from: https://pan.baidu.com/s/19CSagulDBTs9nwcHvMahIw Password: ikls

BibTeX

@article{fang2023towards,
  title={Towards the vectorization of hyperspectral imagery},
  author={Fang, Leyuan and Yan, Yinglong and Yue, Jun and Deng, Yue},
  journal={IEEE Transactions on Geoscience and Remote Sensing},
  year={2023},
  publisher={IEEE}
}