Hybrid Spatial-spectral Neural Network for Hyperspectral Image Denosing
python test_icvl_gaussian.py --arch=proposed_base_icvl --device='cuda' --index=1 --ckpt=model_zoo/proposed_dw/icvl_gaussian_base.ckpt --save_dir=./results/proposed_base
python test_icvl_complex.py --arch=proposed_base_icvl --device='cuda' --index=0 --ckpt=model_zoo/proposed_dw/icvl_complex_base.ckpt --save_dir=./results/proposed_base
python test_realistic.py --arch=proposed_base_real --device='cuda' --index=0 --ckpt=model_zoo/proposed_dw/realistic_base.ckpt --save_dir=./results/proposed_baseAll trained model parameter files can be found on Google Drive.
| ICVL iid Gaussian noise | ICVL Complex Noise | Realistic Dataset |
|---|---|---|
| Baidu Drive from SST | BaiDu Drive from SST | Google Drive |
@InProceedings{liang2024hybrid,
author = {Liang, Hao and Ke Chengjie and Li, Kun},
title = {Hybrid Spatial-spectral Neural Network for Hyperspectral Image Denoising},
booktitle = {Proceedings of the European conference on computer vision (ECCV) workshops},
year = {2024},
organization={Springer}
}