EcoSta 2022: Start Registration
View Submission - EcoSta2022
A0347
Title: Most probable paths for anisotropic Brownian motions on manifolds Authors:  Erlend Grong - University of Bergen (Norway) [presenting]
Stefan Sommer - University of Copenhagen (Denmark)
Abstract: The diffusion mean is an intrinsic way of considering the mean of a dataset located on a Riemannian manifold. It also allows considering general types of covariances through a Brownian motion that is not anisotropic. We describe this approach and how computing this means and covariance is related to the most probable paths on manifolds, which are different from geodesics.