Sliced Wasserstein Distance for Learning Gaussian Mixture Models
Clone or download
Latest commit 24a1a48 Jun 7, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data SWGMM figure Jun 7, 2018
README.md Updated Readme Jun 7, 2018
SWM_GMM_10_G.mp4 Added Files Jun 7, 2018
SWM_GMM_RMSProp_Demo_GitHub.ipynb Added Files Jun 7, 2018
utils.py Added Files Jun 7, 2018

README.md

swgmm

This repository contains a demo implementation of the method described in:

"Sliced Wasserstein Distance for Learning Gaussian Mixture Models", CVPR'18

which defines the sliced-Wasserstein means problem, and describes a novel technique for fitting Gaussian Mixture Models to data. In short, the method minimizes the sliced-Wasserstein distance between the data distribution and a GMM with respect to the GMM parameters.