Learning Bijective Surface Parameterization for Inferring Signed Distance Functions from Sparse Point Clouds with Grid Deformation
Takeshi Noda, Chao Chen, Junsheng Zhou, Weiqi Zhang, Yu-Shen Liu, Zhizhong Han · 2025
Read Learning Bijective Surface Parameterization for Inferring Signed Distance Functions from Sparse Point Clouds with Grid Deformation by Takeshi Noda, Chao Chen, Junsheng Zhou, Weiqi Zhang, Yu-Shen Liu, Zhizhong Han (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-23670.txt.
Description
Rebuilding a 3D surface from a sparse scatter of points is hard because so much shape detail is missing. Takeshi Noda and co-authors present a learning method that infers a signed distance function, the field that encodes a surface, directly from sparse point clouds. Their key device is a two-way mapping between a flat parameter domain and local surface patches, which stitches the patches into one continuous global shape.
A grid deformation step then refines the fit, nudging grid points to better match the surface. On synthetic and real scanned datasets, the method beat current state-of-the-art approaches. The work matters for scanning and modeling tasks where sensors return only thin point data but a watertight surface is still needed.
Catalog record
| CREATOR | Takeshi Noda, Chao Chen, Junsheng Zhou, Weiqi Zhang, Yu-Shen Liu, Zhizhong Han |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 31 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-23670