1.Download and Compile matconvnet-1.0-beta22
2.Run demosaick.m (if no gpu, adjust the code with gpu=0)
Including three testing datasets: Kodak, Mcmaster and Our WED-CDM dataset.
- MATLAB R2014a
- MatConvNet
- (if GPU version) Cuda-8.0 & cudnn v5.1
@inproceedings{tan2017color,
title={Color image demosaicking via deep residual learning},
author={Tan, Runjie and Zhang, Kai and Zuo, Wangmeng and Zhang, Lei},
booktitle={2017 IEEE International Conference on Multimedia and Expo (ICME)},
pages={793--798},
year={2017},
organization={IEEE}
}
The training dataset and training code will release soon.