Fast Entropic Regularized Optimal Transport Using Semidiscrete Cost Approximation
Evgeny Tenetov,Gershon Wolansky,Ron Kimmel
DOI: https://doi.org/10.1137/17m1162925
IF: 2.968
2018-01-01
SIAM Journal on Scientific Computing
Abstract:The optimal transportation theory was successfully applied to different tasks on geometric domains as images and triangle meshes. In these applications the transport problem is defined on a Riemannian manifold with geodesic distance $d(x,y)$. Usually, the cost function used is the geodesic distance $d$ or the squared geodesic distance $d^2$. These choices result in the 1-Wasserstein distance, also known as the earth mover's distance (EMD), or the 2-Wasserstein distance. The entropy regularized optimal transport problem can be solved using the Bregman projection algorithm. This algorithm can be implemented using only matrix multiplications of matrix $\exp(-C/varepsilon)$ (pointwise exponent) and pointwise vector multiplications, where $C$ is a cost matrix, and $\varepsilon$ is the regularization parameter. In this paper, we obtain a low-rank decomposition of this matrix and exploit it to accelerate the Bregman projection algorithm. Our low-rank decomposition is based on the semidiscrete approximation of the cost function, which is valid for a large family of cost functions, including $d^p (x, y)$, where $p \geq 1$. Our method requires the calculation of only a small portion of the distances.
mathematics, applied