Project uses Python, Jupyter Notebook, and Tensorflow to generate images of common clothing items (shoes, sweaters, etc.). Dataset taken from

keras.datasets.fashion_mnist

This project uses a Variational Autoencoder trained on the dataset of clothing to generate random and entirely new images of clothing.

The program then uses a denoising Autoencoder, trained to reconstruct images from altered "noisy" training images, to increase the image definition of the generated images.

This was my first project with autoencoders, and I hope in the future to be able to generate more complex and detailed images (like the ones seen in OpenAI's DALLE 2)