mlsanityData to 13 Oct 2026

NeuroNURBS: Learning Efficient Surface Representations for 3D Solids

arXiv 2411.10848 ↗First posted 16 Nov 2024cs.CV cs.CEcited by 8

Not on the map. Why?

Abstract (arXiv)

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.

Key facts. Cited by 8 papers in the map (data through 13 Oct 2026). Builds on 28 of its 50 references that are in the map. First posted on arXiv 16 Nov 2024.

Builds on

28 of 50 references.

References · 28

  1. 2017Attention Is All You Needcited by 60,106
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,304
  3. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 30,043
  4. 2020Denoising Diffusion Probabilistic Modelscited by 20,952
  5. 2020Denoising Diffusion Implicit Modelscited by 9,718
  6. 2017Graph Attention Networkscited by 8,888
  7. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,231
  8. 2022LION: Latent Point Diffusion Models for 3D Shape Generationcited by 553
  9. 20213D Shape Generation and Completion through Point-Voxel Diffusioncited by 516
  10. 2014Predicting Depth, Surface Normals and Semantic Labels with a Common Multi-Scale Convolutional Architecturecited by 483
  11. 20233DShape2VecSet: A 3D Shape Representation for Neural Fields and Generative Diffusion Modelscited by 452
  12. 2022MultiMAE: Multi-modal Multi-task Masked Autoencoderscited by 289
16 more
  1. 2021DeepCAD: A Deep Generative Network for Computer-Aided Design Modelscited by 270
  2. 2023Multimodal Deep Learningcited by 257
  3. 2023MeshDiffusion: Score-based Generative 3D Mesh Modelingcited by 199
  4. 2022Neural Wavelet-domain Diffusion for 3D Shape Generationcited by 132
  5. 2022Diffusion-SDF: Text-to-Shape via Voxelized Diffusioncited by 128
  6. 2022SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebookscited by 124
  7. 2024BrepGen: A B-rep Generative Diffusion Model with Structured Latent Geometrycited by 103
  8. 2022SolidGen: An Autoregressive Model for Direct B-rep Synthesiscited by 80
  9. 2021JoinABLe: Learning Bottom-up Assembly of Parametric CAD Jointscited by 77
  10. 2022Reconstructing editable prismatic CAD from rounded voxel modelscited by 41
  11. 2024OctFusion: Octree-based Diffusion Models for 3D Shape Generationcited by 41
  12. 2024Geometric Deep Learning for Computer-Aided Design: A Surveycited by 14
  13. 2021NURBS-Diff: A Differentiable Programming Module for NURBScited by 11
  14. 2023On the Noise Scheduling for Generating Plausible Designs with Diffusion Modelscited by 6
  15. 2023Adversarial Latent Autoencoder with Self-Attention for Structural Image Synthesiscited by 5
  16. 2021UVStyle-Net: Unsupervised Few-shot Learning of 3D Style Similarity Measure for B-Repscited by 3

Built on by

Cited by year (2026 to 13 Oct)

Citing papers per year: 2025: 2; 2026: 6.

8 papers.

  1. 2025AutoBrep: Autoregressive B-Rep Generation with Unified Topology and Geometrycited by 17
  2. 2026FoV-Net: Rotation-Invariant CAD B-rep Learning via Field-of-View Ray Castingcited by 3
  3. 2026Autoregressive B-Rep Shape Generation with Parametric Surfacescited by 2
  4. 2025BrepLLM: Enabling Large Language Models to Understand Boundary Representationscited by 1
  5. 2026Draw it like Euclid: Teaching transformer models to generate CAD profiles using ruler and compass construction stepscited by 1
  6. 2026AutoRegressive Generation with B-rep Holistic Token Sequence Representationcited by 0
  7. 2026NURBS Splatting: A Unified Differentiable Rendering Framework for Vector Graphicscited by 0
  8. 2026CANTO: CAD-Native Transformer Operators for AI-Aided Engineeringcited by 0