Neural Dual Contouring

TL;DR

Neural Dual Contouring (NDC) is a neural network-based mesh reconstruction method that improves accuracy and feature preservation over traditional and previous learned approaches, capable of handling various input formats and generalizing across datasets.

Contribution

NDC introduces a neural network to predict vertex positions and edge crossings in dual contouring, enabling flexible, accurate, and efficient mesh reconstruction from diverse data types.

Findings

NDC outperforms previous methods in accuracy and feature preservation.

NDC generalizes well across different datasets.

NDC offers faster inference times.

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.