MCNet

This is an implementation of Mixed 2D/3D Convolutional Network for Hyperspectral Image Super-Resolution.

Dataset

Three public datasets, i.e., CAVE, Harvard, Foster, are employed to verify the effectiveness of the proposed MCNet. Since there are too few images in these datasets for deep learning algorithm, we augment the training data. With respect to the specific details, please see the implementation details section.

Moreover, we also provide the code about data pre-processing in folder "data pre-processing". The folder contains three parts, including training set augment, test set pre-processing, and band mean for all training set.

Requirement

python 2.7, Pytorch 0.3.1, cuda 9.0

Training

The ADAM optimizer with beta_1 = 0.9, beta _2 = 0.999 is employed to train our network. The learning rate is initialized as 10^-4 for all layers, which decreases by a half at every 35 epochs.

You can train or test directly from the command line as such:

# python train.py --cuda --datasetName CAVE --upscale_factor 4

# python test.py --cuda --model_name checkpoint/model_4_epoch_XXX.pth

Result

To qualitatively measure the proposed MCNet, three evaluation methods are employed to verify the effectiveness of the algorithm, including peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and spectral angle mapping (SAM).

Scale	CAVE	Harvard	Foster
x2	45.102 / 0.9738 / 2.241	46.263 / 0.9827 / 1.883	58.878 / 0.9988 / 4.061
x3	41.031 / 0.9526 / 2.809	42.681 / 0.9627 / 2.214	55.017 / 0.9970 / 5.126
x4	39.026 / 0.9319 / 3.292	40.081 / 0.9367 / 2.410	52.225 / 0.9941 / 5.685

Citation

Please consider cite MCNet for hyperspectral image super-resolution if you find it helpful.

@article{Li2020mixed,

title={Mixed  {2D/3D} Convolutional Network  for Hyperspectral Image Super-Resolution},
author={Q. Li and Q. Wang and X. Li},
journal={Remote Sensing},
volume={12},
number={10},
pages={1660},
year={2020}}

If you has any questions, please send e-mail to liqmges@gmail.com.

qianngli / MCNet