| TR#: | MSC-2019-23 |
| Class: | MSC |
| Title: | Deep Eikonal Solvers |
| Authors: | Moshe Lichtenstein |
| Supervisors: | Ron Kimmel |
| Currently accessibly only within the Technion network | |
| Abstract: | Geodesic distance, a generalization of the Euclidean distance to surfaces with an effective curvature, is a ubiquitous tool used in three-dimensional data analysis. The equation that describes the geodesic distance locally over a surface is a PDE called the Eikonal equation. Typical to the numerical solution of PDEs, estimating the Eikonal solution on a discrete domain requires a trade-off between accuracy and computational complexity. In order to develop a quasilinear-time yet accurate algorithm, a deep learning approach to numerically approximate the solution to the Eikonal equation is introduced.
The proposed method is built on the fast marching scheme which comprises two components: a local numerical solver and an update scheme. We replace the formulaic local numerical solver with a trained neural network to provide highly accurate estimates of local distances for a variety of different geometries and sampling conditions. Our learning approach generalizes not only to flat Euclidean domains but also to curved surfaces enabled by the incorporation of certain invariant features in the neural network architecture. We show a considerable gain in performance, validated by smaller errors and higher orders of accuracy for the numerical solutions of the Eikonal equation computed on different surfaces. The proposed approach leverages the approximation power of neural networks to enhance the performance of numerical algorithms, thereby, connecting the somewhat disparate themes of numerical geometry and learning. |
| Copyright | The above paper is copyright by the Technion, Author(s), or others. Please contact the author(s) for more information |
Remark: Any link to this technical report should be to this page (http://www.cs.technion.ac.il/users/wwwb/cgi-bin/tr-info.cgi/2019/MSC/MSC-2019-23), rather than to the URL of the PDF files directly. The latter URLs may change without notice.
To the list of the MSC technical reports of 2019
To the main CS technical reports page