gabriel-ozeas/mnist-wgan

A Keras implementation of the Wasserstein GAN using the MNIST dataset

---

Getting Started

Clone the repository, navigate to the local directory and run:

python main.py

Sample images created by the generator at epochs 0 and 40.

Prerequisites

• Keras >= 2.0.8
• TensorFlow >= 1.3.0
• Numpy >= 1.13.3
• Matplotlib >= 2.0.2
• Seaborn >= 0.7.1

Built With

• TensorFlow - A deep learning library
• Keras - A high-level deep learning package built upon TensorFlow

Authors

• David Reiman - davidreiman

References

• Wasserstein GAN - Original paper by Arjovsky et al.
• Wasserstein implementation - Original implementation by Arjovsky et al.
• Wasserstein GAN in Keras - Keras implementation of Wasserstein GAN