How Well Do WGANs Estimate the Wasserstein Metric?

ML-2021-Project

Project Team

1. Denis Rollov ([email protected])
2. Nikolay Goncharov ([email protected])
3. Svetlana Gabdullina ([email protected])

Utils

All necessary implementations of NNs architectures, losses can be found in the folder WGAN. Jupyter notebooks (*.ipynb) contain all examples and experiments. All code works in Google Colab.

Theoretical part can be found in the project report.

Presentation can be found here.

Link to the video can be found here.