LEARNING PHASE FIELD MEAN CURVATURE FLOWS WITH NEURAL NETWORKS

We introduce in this paper new, efficient numerical methods based on neural networks for the approximation of the mean curvature flow of either oriented or non-orientable surfaces. To learn the correct interface evolution law, our neural networks are trained on phase field representations of exact evolving interfaces. The structure of the networks draws inspiration from splitting schemes used for the discretization of the Allen-Cahn equation. But when the latter approximates the mean curvature motion of oriented interfaces only, the approach we propose extends very naturally to the non-orientable case. In addition, although trained on smooth flows only, our networks can handle singularities as well. Furthermore, they can be coupled easily with additional constraints which allows us to show various applications illustrating the flexibility and efficiency of our approach: mean curvature flows with volume constraint, multiphase mean curvature flows, numerical approximation of Steiner trees, numerical approximation of minimal surfaces