This repo contains my implementation of many well-known super-resolution deep neural network (DNN) models. Each folder contains a separate model along with instruction on how to train, evaluate, and super-resolve an input image. This page acts as a summary on what models I have implemented and what I plan to implement next.
This is a just-for-fun project.
The following models have been implemented:
- SRCNN - Super-Resolution Convolutional Neural Network: paper.
- ESPCN - Efficient Sub-Pixel Convolutional Neural Network: paper.
- FSRCNN - Accelerating the Super-Resolution Convolutional Neural Network: paper.
- LapSRN - Fast and Accurate Image Super-Resolution with Deep Laplacian Pyramid Networks: paper.