Skip to content

dman14/Super-Resolution-using-Hierarchical-VAEs

Repository files navigation

Super-Resolution-using-Hierarchical-VAEs

02456 Deep learning, DTU Compute, Autumn 2020 Final Project

Information

The Poster for the presentation of the project can be found here

The Report of the project can be seen here

Description

The purpose of the project was to achieve and develop a super-resolution model based on a Hierarchical VAE, here the Ladder VAE implementation was used to build upon it.

The Project has three parts, those showing the progress up to the hierarchical VAE, those parts being a SR-CNN implementation, a SR-VAE model, and finally the SR-LVAE model.

Usage

To run any of the three models, the Operational IPytthon Notebooks can be run, each for their individual model.

For the Training/Testing data, the datasets can be downloaded from here. For Training the General100 dataset was used, while for Testing the Set14 was used.

Also, all the packages from requirements.txt are needed in order to run the notebooks.

Results

About

Deep Learning course Project

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published