Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2511.14283 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909910806036480 |
|---|---|
| author | Xi, Zi-Chen Huang, Jiahui Chen, Hao-Xiang Williams, Francis Xu, Qun-Ce Mu, Tai-Jiang Hu, Shi-Min |
| author_facet | Xi, Zi-Chen Huang, Jiahui Chen, Hao-Xiang Williams, Francis Xu, Qun-Ce Mu, Tai-Jiang Hu, Shi-Min |
| contents | We proposed a generalized method, NeuralSSD, for reconstructing a 3D implicit surface from the widely-available point cloud data. NeuralSSD is a solver-based on the neural Galerkin method, aimed at reconstructing higher-quality and accurate surfaces from input point clouds. Implicit method is preferred due to its ability to accurately represent shapes and its robustness in handling topological changes. However, existing parameterizations of implicit fields lack explicit mechanisms to ensure a tight fit between the surface and input data. To address this, we propose a novel energy equation that balances the reliability of point cloud information. Additionally, we introduce a new convolutional network that learns three-dimensional information to achieve superior optimization results. This approach ensures that the reconstructed surface closely adheres to the raw input points and infers valuable inductive biases from point clouds, resulting in a highly accurate and stable surface reconstruction. NeuralSSD is evaluated on a variety of challenging datasets, including the ShapeNet and Matterport datasets, and achieves state-of-the-art results in terms of both surface reconstruction accuracy and generalizability. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_14283 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | NeuralSSD: A Neural Solver for Signed Distance Surface Reconstruction Xi, Zi-Chen Huang, Jiahui Chen, Hao-Xiang Williams, Francis Xu, Qun-Ce Mu, Tai-Jiang Hu, Shi-Min Computer Vision and Pattern Recognition Graphics Machine Learning I.5; I.3 We proposed a generalized method, NeuralSSD, for reconstructing a 3D implicit surface from the widely-available point cloud data. NeuralSSD is a solver-based on the neural Galerkin method, aimed at reconstructing higher-quality and accurate surfaces from input point clouds. Implicit method is preferred due to its ability to accurately represent shapes and its robustness in handling topological changes. However, existing parameterizations of implicit fields lack explicit mechanisms to ensure a tight fit between the surface and input data. To address this, we propose a novel energy equation that balances the reliability of point cloud information. Additionally, we introduce a new convolutional network that learns three-dimensional information to achieve superior optimization results. This approach ensures that the reconstructed surface closely adheres to the raw input points and infers valuable inductive biases from point clouds, resulting in a highly accurate and stable surface reconstruction. NeuralSSD is evaluated on a variety of challenging datasets, including the ShapeNet and Matterport datasets, and achieves state-of-the-art results in terms of both surface reconstruction accuracy and generalizability. |
| title | NeuralSSD: A Neural Solver for Signed Distance Surface Reconstruction |
| topic | Computer Vision and Pattern Recognition Graphics Machine Learning I.5; I.3 |
| url | https://arxiv.org/abs/2511.14283 |