Utilizing machine learning, computers can now be taught to generate real images. This project uses a general adversarial network (GAN) to generate images. A GAN is composed of two neural networks: one plays the painter and one the judge. The painter uses responses from the judge to improve its painting, eventually generating life-like images. However, getting a GAN to work as intended in this regard is notoriously difficult; one goal of the project is to improve the training process.