用神经网络直接学习NURBS参数，提升3D建模效率与精度
NeuroNURBS: Learning Efficient Surface Representations for 3D Solids
- 用神经网络直接编码NURBS表面参数，替代传统采样网格
- 训练时显存降低79.9%，存储3D模型内存减少79.9%
- 生成的3D模型更平滑，FID从30.04降至27.24，适合工业设计
在计算机辅助设计中，边界表示（B-Rep）是3D实体的标准表示方式，由一组NURBS（非均匀有理B样条）曲面构成封闭体积。现有方法通常采用UV网格对曲面进行均匀采样，但该方法效率低，且常缺乏精度与规则性。本文提出NeuroNURBS，一种直接学习NURBS曲面参数的表示学习方法。在实体生成与分割任务中的评估表明，NeuroNURBS性能与UV网格相当甚至更优，但效率显著提升：训练表面自编码器时GPU消耗减少86.7%；存储3D实体时内存需求降低79.9%。此外，将BrepGen用于实体生成时结合NeuroNURBS，FID从30.04降至27.24，并解决了生成曲面的波浪状失真问题。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。