Real world data often lie on low-dimensional Riemannian manifolds embedded in high-dimensional spaces. This motivates learning degenerate normalizing flows that map between the ambient space and a low-dimensional latent space. However, if the manifold has a non-trivial topology, it can never be correctly learned using a single flow. Instead multiple flows must be `glued together'. In this paper, we first propose the general training scheme for learning such a collection of flows, and secondly we develop the first numerical algorithms for computing geodesics on such manifolds. Empirically, we demonstrate that this leads to highly significant improvements in topology estimation.
View on arXiv@article{yu2025_2505.24665,
title={ Learning geometry and topology via multi-chart flows },
author={ Hanlin Yu and Søren Hauberg and Marcelo Hartmann and Arto Klami and Georgios Arvanitidis },
journal={arXiv preprint arXiv:2505.24665},
year={ 2025 }
}