/StatueStyleGAN

This repository contains the tensorflow implementation of A Style-Based Generator Architecture for Generative Adversarial Networks applied on Statues.

Primary LanguageJupyter Notebook

StatueStyleGAN

This repository contains the tensorflow implementation of "A Style-Based Generator Architecture for Generative Adversarial Networks" paper by Karras et al. (https://arxiv.org/abs/1812.04948).

The dataset used for this work was manually gathered from pinterest (http://bit.ly/2GGNMIZ) and image preprocessing techniques such as cropping and resizing was peformed on these images afterwards.

The training and testing code can be found in the jupyter notebook.

Pretrained Model

The Pretrained model for this work can be found here: http://bit.ly/2yGCnEF

Excerpts of Training data

These are some of the images that was gotten from pinterest and used for training. Training data

Results

These are the artifical statues that was generated by the generator model. Generated Statues