Sliced Wasserstein Distance for Learning Gaussian Mixture Models | IEEE Conference Publication | IEEE Xplore