Linear optimal transport subspaces for point set classification

arXiv:2403.10015v1 Announce Type: cross
Abstract: Learning from point sets is an essential component in many computer vision and machine learning applications. Native, unordered, and permutation invariant set structure space is challenging to model, particularly for point set classification under spatial deformations. Here we propose a framework for classifying point sets experiencing certain types of spatial deformations, with a particular emphasis on datasets featuring affine deformations. Our approach employs the Linear Optimal Transport (LOT) transform to obtain a linear embedding of …

abstract applications arxiv classification computer computer vision cs.cv cs.lg framework linear machine machine learning machine learning applications set space spatial transport type vision