GAN Experiments

Missing Data [2D space]

Given a GAN can we fill in corrupted or missing data in images of certain types - for e.g. first restrict to numbers(MNIST), letters(EMNIST) and then onto objects(COCO)

Run Environment

Build System:
macOS
Python 3 Kernel in Jupyter Notebooks

Dependencies

Tensorflow
Keras
Tqdm

Datasets

MNIST
EMNIST
COCO