Mesh Draping: Parametrization?Free Neural Mesh Transfer

Despite recent advances in geometric modelling, 3D mesh modelling still involves a considerable amount of manual labour by experts. In this paper, we introduce Mesh Draping: neural method for transferring existing structure from one shape to another. The drapes the source over target geometry and at same time seeks preserve carefully designed characteristics mesh. At its core, our deforms using progressive positional encoding (PE). We show that leveraging gradually increasing frequencies guide optimization, are able achieve stable high-quality transfer. Our approach is simple requires little user guidance, compared contemporary surface mapping techniques which rely on parametrization or careful tuning. Most importantly, Draping parameterization-free method, thus applicable variety representations, including point clouds, polygon soups non-manifold meshes. demonstrate transferred meshing remains faithful design characteristics, fits well.