This repository is an official PyTorch implementation of the paper "Dual Convolutional Neural Networks for Low-Level Vision" .
The code is built on EDSR (PyTorch) and tested on Ubuntu 16.04 environment (Python3.6, PyTorch_0.4.1, CUDA8.0, cuDNN5.1) with NVIDIA 1080Ti GPUs.
- ubuntu16.04
- Python 3.6(Recommend to use Anaconda)
- PyTorch0.4.1
- numpy
- skimage
- imageio
- matplotlib
- tqdm
- cv2
This repository includes Dual CNN in Super-Resolution task and Denoise task.
- If you are interested in Super-Resolution. please read the readme.md in
./Super-Resolution/README.md - If you are interested in Denoise. please read the readme.md in
./Denoise/README.md
