In this project, we'll define and train a DCGAN on a dataset of faces. Our goal is to get a generator network to generate new images of faces that look as realistic as possible! The project will be broken down into a series of tasks from loading in data to defining and training adversarial networks. At the end of the notebook, we'll be able to visualize the results of our trained Generator to see how it performs; our generated samples should look like fairly realistic faces with small amounts of noise.
We'll be using the CelebFaces Attributes Dataset (CelebA) to train our adversarial networks. This dataset is more complex than the number datasets (like MNIST or SVHN) so we should define deeper networks and train them for a longer time to get good results. It is suggested to utilize a GPU for training.
Since the project's main focus is on building the GANs, pre-processed data is available for us. Each of the CelebA images has been cropped to remove parts of the image that don't include a face, then resized down to 64x64x3 NumPy images.
if you are working locally, you can download this data by clicking here
https://github.com/udacity/deep-learning
https://www.udacity.com/course/deep-learning-nanodegree--nd101