FSRCNN

Unofficial implementation of FSRCNN using PyTorch.

Steps

1. Run data_aug.m to augment training set.
2. Run generate_train.m and generate_test.m to generate h5 files for data loader. (Note that you may need to change the path in the matlab code for different image source)
3. Run main.py.

Requirement

1. python >= 3.4
2. pytorch >= 0.4
3. torchvision
4. tensorflow

Reference

Chao Dong, Chen Change Loy, Xiaoou Tang. Accelerating the Super-Resolution Convolutional Neural Network, in Proceedings of European Conference on Computer Vision (ECCV), 2016