Collection of GANs created during development.

The autoencoder and comparison directories contain the results from the project. Other networks in development, failed experiments, etc. should be placed in the development directory.

All files should be Jupyter notebooks.

Models currently are designed for Keras and specifically the tensorflow backend. We ran these notebooks using Google Colab however earlier models were trained using Microsoft Azure.

Important notebooks

• DCGAN_Autoencoder_Sprite_items_weapons.ipynb
• Copy of DCGAN_Autoencoder_Sprite_items_weapons.ipynb
• Comparison DCGAN T.ipynb
• Comparison GAN 2.ipynb
• DCGAN_Sprite_items_weapons.ipynb

This project was developed for CS 534 - Computational Photography fall 2018 course at the University of Wisconsin Madison.

We are,

• Curt Henrichs (Graduate Student)
• Sayem Wani (Undergrad Student)
• Saheen Feroz (Undergrad Student)