A python implementatation of the Wasserstein distance (also known as earth mover's distance) between two finite mixture models. This function was used to generate the figures from the paper:
WIND: WASSERSTEIN INCEPTION DISTANCE FOR EVALUATING GENERATIVE ADVERSARIAL NETWORK PERFORMANCE
ICASSP 2020
P. Dimitrakopoulos, G. Sfikas, C. Nikou
Here is the list of libraries you need to install to execute the code:
- python = 3.6
- numpy
- scipy