/SSFIN

Primary LanguagePythonMIT LicenseMIT

Multi-task Interaction learning for Spatiospectral Image Super-Resolution

usage: main.py [-h] --upscale_factor UPSCALE_FACTOR [--batchSize BATCHSIZE]
               [--testBatchSize TESTBATCHSIZE] [--nEpochs NEPOCHS] [--lr LR]
               [--threads THREADS] [--seed SEED]

PyTorch Super Res Example

optional arguments:
  -h, --help            show this help message and exit
  --upscale_factor      super resolution upscale factor
  --batchSize           training batch size
  --testBatchSize       testing batch size
  --nEpochs             number of epochs to train for
  --lr                  Learning Rate. Default=0.0001
  --threads             number of threads for data loader to use Default=4
  --seed                random seed to use. Default=123

This example trains a spatiospectral super-resolution network on the CAVE dataset, using the first 22 HSIs for training, the remaining 10 HSIs for testing. A trained model can be downloaded at https://drive.google.com/file/d/1x60giSaTPZWZoG25sRhEbhFxVN8C_3u_/view?usp=sharing

Example Usage:

Train

python main.py

Test

python main.py --mode 0 --nEpochs 100

The code is for the work:

@article{ma2022multi,
  title={Multi-task Interaction learning for Spatiospectral Image Super-Resolution},
  author={Qing Ma, Junjun Jiang, Xianming Liu, and Jiayi Ma},
  journal={IEEE Transactions on Image Processing},
  volume={},
  number={},
  pages={},
  year={2022},
}