NeuroNURBS: Learning Efficient Surface Representations for 3D Solids
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916485434179584 |
|---|---|
| author | Fan, Jiajie Gholami, Babak Bäck, Thomas Wang, Hao |
| author_facet | Fan, Jiajie Gholami, Babak Bäck, Thomas Wang, Hao |
| contents | Boundary Representation (B-Rep) is the de facto representation of 3D solids in Computer-Aided Design (CAD). B-Rep solids are defined with a set of NURBS (Non-Uniform Rational B-Splines) surfaces forming a closed volume. To represent a surface, current works often employ the UV-grid approximation, i.e., sample points uniformly on the surface. However, the UV-grid method is not efficient in surface representation and sometimes lacks precision and regularity. In this work, we propose NeuroNURBS, a representation learning method to directly encode the parameters of NURBS surfaces. Our evaluation in solid generation and segmentation tasks indicates that the NeuroNURBS performs comparably and, in some cases, superior to UV-grids, but with a significantly improved efficiency: for training the surface autoencoder, GPU consumption is reduced by 86.7%; memory requirement drops by 79.9% for storing 3D solids. Moreover, adapting BrepGen for solid generation with our NeuroNURBS improves the FID from 30.04 to 27.24, and resolves the undulating issue in generated surfaces. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2411_10848 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | NeuroNURBS: Learning Efficient Surface Representations for 3D Solids Fan, Jiajie Gholami, Babak Bäck, Thomas Wang, Hao Computer Vision and Pattern Recognition Computational Engineering, Finance, and Science Boundary Representation (B-Rep) is the de facto representation of 3D solids in Computer-Aided Design (CAD). B-Rep solids are defined with a set of NURBS (Non-Uniform Rational B-Splines) surfaces forming a closed volume. To represent a surface, current works often employ the UV-grid approximation, i.e., sample points uniformly on the surface. However, the UV-grid method is not efficient in surface representation and sometimes lacks precision and regularity. In this work, we propose NeuroNURBS, a representation learning method to directly encode the parameters of NURBS surfaces. Our evaluation in solid generation and segmentation tasks indicates that the NeuroNURBS performs comparably and, in some cases, superior to UV-grids, but with a significantly improved efficiency: for training the surface autoencoder, GPU consumption is reduced by 86.7%; memory requirement drops by 79.9% for storing 3D solids. Moreover, adapting BrepGen for solid generation with our NeuroNURBS improves the FID from 30.04 to 27.24, and resolves the undulating issue in generated surfaces. |
| title | NeuroNURBS: Learning Efficient Surface Representations for 3D Solids |
| topic | Computer Vision and Pattern Recognition Computational Engineering, Finance, and Science |
| url | https://arxiv.org/abs/2411.10848 |