Neural Dual Contouring

Abstract

We introduce neural dual contouring (NDC), a new data-driven approach to mesh reconstruction based on dual contouring (DC). Like traditional DC, it produces exactly one vertex per grid cell and one quad for each grid edge intersection, a natural and efficient structure for reproducing sharp features. However, rather than computing vertex locations and edge crossings with hand-crafted functions that depend directly on difficult-to-obtain surface gradients, NDC uses a neural network to predict them. As a result, NDC can be trained to produce meshes from signed or unsigned distance fields, binary voxel grids, or point clouds (with or without normals); and it can produce open surfaces in cases where the input represents a sheet or partial surface. During experiments with five prominent datasets, we find that NDC, when trained on one of the datasets, generalizes well to the others. Furthermore, NDC provides better surface reconstruction accuracy, feature preservation, output complexity, triangle quality, and inference time in comparison to previous learned (e.g., neural marching cubes, convolutional occupancy networks) and traditional (e.g., Poisson) methods. Code and data are available at https://github.com/czq142857/NDC.

Accepted to SIGGRAPH (journal) 2022. Code: https://github.com/czq142857/NDC

References in corpus (7)

• PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
• Open3D: A Modern Library for 3D Data Processing
• Implicit Neural Representations with Periodic Activation Functions
• Point2Mesh: A Self-Prior for Deformable Meshes
• Thingi10K: A Dataset of 10,000 3D-Printing Models
• Neural Marching Cubes
• Learning Deformable Tetrahedral Meshes for 3D Reconstruction

Cited by in corpus (16)

• Robust Zero Level-Set Extraction from Unsigned Distance Fields Based on Double Covering
• Neural-IMLS: Self-supervised Implicit Moving Least-Squares Network for Surface Reconstruction
• Implicit Conversion of Manifold B-Rep Solids by Neural Halfspace Representation
• SpaceMesh: A Continuous Representation for Learning Manifold Surface Meshes
• PPSURF: Combining Patches and Point Convolutions for Detailed Surface Reconstruction
• TetWeave: Isosurface Extraction using On-The-Fly Delaunay Tetrahedral Grids for Gradient-Based Mesh Optimization
• Interaction-Driven Active 3D Reconstruction with Object Interiors
• Neurally Integrated Finite Elements for Differentiable Elasticity on Evolving Domains
• LFS-Aware Surface Reconstruction from Unoriented 3D Point Clouds
• Occupancy-Based Dual Contouring
• PG-NeuS: Robust and Efficient Point Guidance for Multi-View Neural Surface Reconstruction
• Med-PU: Point Cloud Upsampling for High-Fidelity 3D Medical Shape Reconstruction
• Local Surface Parameterizations via Smoothed Geodesic Splines
• Point2Quad: Generating Quad Meshes from Point Clouds via Face Prediction
• Estimation and Visualization of Isosurface Uncertainty from Linear and High-Order Interpolation Methods
• PaMO: Parallel Mesh Optimization for Intersection-Free Low-Poly Modeling on the GPU