Didactic Supervised Super-resolution

This repository contains the pytorch implementation of a sequential super resolution algorithm.The architecture and implementation detail are present in the report in the repository

Getting Started

Copy the entire high res training dataset in png format in a folder called data in the working directory. Copy the low res image to be upscaled in the valid directory in the workspace. create folders result2,result4 and result8 in the workspace.

Prerequisites

python3
pytorch
torchvision 

64 GB RAM or more Atleast 1 GTX 1080ti or better

Training

$python3 tain.py

Testing

$python3 test.py

The above command shall dump 2X,4X and an 8X resolved image in ./result2, ./result4 and ./result8