# Neural Dual Contouring Source: [Neural Dual Contouring]() Collection index: [llms.txt]() Data updated: 2025-10-26T00:15:46.115Z Authors: Zhiqin Chen, Andrea Tagliasacchi, Thomas Funkhouser, Hao Zhang Published: 2022-02-04T07:03:39.000Z ## 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. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Mesh Reconstruction on Thingi10K (test)](): Mesh Reconstruction; Thingi10K (test); Chamfer Distance (CD): 3.52. - [Point Cloud Reconstruction on Thingi32, DeepFashion3D, and Objaverse (test)](): Point Cloud Reconstruction; Thingi32, DeepFashion3D, and Objaverse (test); F1: 87.4. - [Mesh Reconstruction on ABC (test)](): Mesh Reconstruction; ABC (test); Chamfer Distance: 3.39. - [Surface Reconstruction on ShapeNet (CD, NC, IoU, LSD-A)](): Surface Reconstruction; ShapeNet; CD: 2.9. - [Mesh Reconstruction on ABC analytical (test)](): Mesh Reconstruction; ABC analytical (test); CD: 3.6. - [Mesh Reconstruction on Thingi10K analytical (test)](): Mesh Reconstruction; Thingi10K analytical (test); CD: 4. ## Original sources and implementation [Original paper]() [PDF]()