RDN

This project aims at providing a fast, modular reference implementation for super-resolution models using pytorch

Introducation

RDN(CVPR2018)

training

preprocess

For training you shall download the DIV2k dataset:- DIV2K
put your train_img,and valid_img to the DIV2K_train_HR and DIV2K_valid_HR.

you shall python main.py process to generate downsample data and then you can train your RDN-Net to use python main.py train .
you can change your para from the config.py All of it realized from pytorch.
Finaly if you want to see the output ,you can download the visdom to see output real time

training_Loss

eval

original

predict

Support

if you have question ,email me [email protected]