In this project, I implemented a Generative Adversarial Network using PyTorch based on the repo: https://github.com/diegoalejogm/gans/blob/master/1.%20Vanilla%20GAN%20PyTorch.ipynb, to generate fake images of parked vehicles. The discriminator network was trained on the dataset within "Parking_Lot_Vehicle_Images_UFPR04.zip" which contains 46,125 images of parked vehicles taken from security cameras in a parking lot. The zip folder has to be extracted into the root folder.
"Project_4_Git.ipynb" contains all information regarding the code implementation in the form of markdowns and comments. An overview of the implementaion as well as analysis of results can be found in the document "Project 4 Report".