mlsanityData to 14 Oct 2026

D2CSG: Unsupervised Learning of Compact CSG Trees with Dual Complements and Dropouts

arXiv 2301.11497 ↗First posted 27 Jan 2023cs.CV cs.GRcited by 40

Not on the map. Why?

Abstract (arXiv)

We present D2CSG, a neural model composed of two dual and complementary network branches, with dropouts, for unsupervised learning of compact constructive solid geometry (CSG) representations of 3D CAD shapes. Our network is trained to reconstruct a 3D shape by a fixed-order assembly of quadric primitives, with both branches producing a union of primitive intersections or inverses. A key difference between D2CSG and all prior neural CSG models is its dedicated residual branch to assemble the potentially complex shape complement, which is subtracted from an overall shape modeled by the cover branch. With the shape complements, our network is provably general, while the weight dropout further improves compactness of the CSG tree by removing redundant primitives. We demonstrate both quantitatively and qualitatively that D2CSG produces compact CSG reconstructions with superior quality and more natural primitives than all existing alternatives, especially over complex and high-genus CAD shapes.

Key facts. Cited by 40 papers in the map (data through 14 Oct 2026). Builds on 45 of its 57 references that are in the map. First posted on arXiv 27 Jan 2023.

Builds on

45 of 57 references.

References · 45

  1. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,897
  2. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,689
  3. 2018Occupancy Networks: Learning 3D Reconstruction in Function Spacecited by 2,041
  4. 2018Learning Implicit Fields for Generative Shape Modelingcited by 1,060
  5. 2021Volume Rendering of Neural Implicit Surfacescited by 996
  6. 2018PartNet: A Large-scale Benchmark for Fine-grained and Hierarchical Part-level 3D Object Understandingcited by 526
  7. 2021Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapescited by 426
  8. 2018ABC: A Big CAD Model Dataset For Geometric Deep Learningcited by 341
  9. 2021DeepCAD: A Deep Generative Network for Computer-Aided Design Modelscited by 271
  10. 2019Learning Shape Templates with Structured Implicit Functionscited by 242
  11. 2019BSP-Net: Generating Compact Meshes via Binary Space Partitioningcited by 206
  12. 2016Learning Shape Abstractions by Assembling Volumetric Primitivescited by 181
33 more
  1. 2019CvxNet: Learnable Convex Decompositioncited by 174
  2. 2022SkexGen: Autoregressive Generation of CAD Construction Sequences with Disentangled Codebookscited by 125
  3. 2019Superquadrics Revisited: Learning 3D Shape Parsing beyond Cuboidscited by 124
  4. 2018Deep Geometric Prior for Surface Reconstructioncited by 110
  5. 2017CSGNet: Neural Shape Parser for Constructive Solid Geometrycited by 106
  6. 2021Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networkscited by 103
  7. 2019Learning elementary structures for 3D shape generation and matchingcited by 98
  8. 2020ParSeNet: A Parametric Surface Fitting Network for 3D Point Cloudscited by 96
  9. 2021BRepNet: A topological message passing system for solid modelscited by 95
  10. 2019Write, Execute, Assess: Program Synthesis with a REPLcited by 93
  11. 2018Supervised Fitting of Geometric Primitives to 3D Point Cloudscited by 91
  12. 2020UCSG-Net -- Unsupervised Discovering of Constructive Solid Geometry Treecited by 78
  13. 2021JoinABLe: Learning Bottom-up Assembly of Parametric CAD Jointscited by 78
  14. 20173D-PRNN: Generating Shape Primitives with Recurrent Neural Networkscited by 77
  15. 2019Learning to Infer and Execute 3D Shape Programscited by 68
  16. 2021SketchGen: Generating Constrained CAD Sketchescited by 68
  17. 2021CAPRI-Net: Learning Compact CAD Shapes with Adaptive Primitive Assemblycited by 62
  18. 2021Point2Cyl: Reverse Engineering 3D Objects from Point Clouds to Extrusion Cylinderscited by 60
  19. 2018Im2Struct: Recovering 3D Shape Structure from a Single RGB Imagecited by 55
  20. 2020Learning 3D Part Assembly from a Single Imagecited by 52
  21. 2021CSG-Stump: A Learning Friendly CSG-Like Representation for Interpretable Shape Parsingcited by 51
  22. 2021Inferring CAD Modeling Sequences Using Zone Graphscited by 50
  23. 2021Unsupervised Learning for Cuboid Shape Abstraction via Joint Segmentation from Point Cloudscited by 44
  24. 2022Reconstructing editable prismatic CAD from rounded voxel modelscited by 41
  25. 2022ExtrudeNet: Unsupervised Inverse Sketch-and-Extrude for Shape Parsingcited by 38
  26. 2019Composite Shape Modeling via Latent Space Factorizationcited by 31
  27. 2020On the Effectiveness of Weight-Encoded Neural Implicit 3D Shapescited by 29
  28. 2021CPFN: Cascaded Primitive Fitting Networks for High-Resolution Point Cloudscited by 24
  29. 2020PLAD: Learning to Infer Shape Programs with Pseudo-Labels and Approximate Distributionscited by 19
  30. 2020Neural Star Domain as Primitive Representationcited by 17
  31. 2021Differentiable 3D CAD Programs for Bidirectional Editingcited by 14
  32. 2021Discovering 3D Parts from Image Collectionscited by 13
  33. 2018Physical Primitive Decompositioncited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 1; 2024: 18; 2025: 9; 2026: 12.

40 papers.

Sort:

By area:

  1. 2024CAD-MLLM: Unifying Multimodality-Conditioned CAD Generation With MLLMcited by 81
  2. 2024CAD-Recode: Reverse Engineering CAD Code from Point Cloudscited by 66
  3. 2024CAD-SIGNet: CAD Language Inference from Point Clouds using Layer-wise Sketch Instance Guided Attentioncited by 64
  4. 2025Text-to-CAD Generation Through Infusing Visual Feedback in Large Language Modelscited by 47
  5. 2024Img2CAD: Conditioned 3D CAD Model Generation from Single Image with Structured Visual Geometrycited by 37
  6. 2024Img2CAD: Reverse Engineering 3D CAD Models from Images through VLM-Assisted Conditional Factorizationcited by 35
  7. 2024FlexCAD: Unified and Versatile Controllable CAD Generation with Fine-tuned Large Language Modelscited by 34
  8. 2025HoLa: B-Rep Generation using a Holistic Latent Representationcited by 30
  9. 2024TransCAD: A Hierarchical Transformer for CAD Sequence Inference from Point Cloudscited by 29
  10. 2024Split-and-Fit: Learning B-Reps via Structure-Aware Voronoi Partitioningcited by 18
  11. 2025AutoBrep: Autoregressive B-Rep Generation with Unified Topology and Geometrycited by 17
  12. 2025cadrille: Multi-modal CAD Reconstruction with Reinforcement Learningcited by 16
28 more
  1. 2024From 2D CAD Drawings to 3D Parametric Models: A Vision-Language Approachcited by 15
  2. 2023Doodle Your 3D: From Abstract Freehand Sketches to Precise 3D Shapescited by 14
  3. 2025BrepGPT: Autoregressive B-rep Generation with Voronoi Half-Patchcited by 13
  4. 2026Pointer-CAD: Unifying B-Rep and Command Sequences via Pointer-based Edges & Faces Selectioncited by 11
  5. 2025CAD-Judge: Toward Efficient Morphological Grading and Verification for Text-to-CAD Generationcited by 10
  6. 2024Learning to Edit Visual Programs with Self-Supervisioncited by 7
  7. 2026CADReasoner: Iterative Program Editing for CAD Reverse Engineeringcited by 6
  8. 2024DPA-Net: Structured 3D Abstraction from Sparse Views via Differentiable Primitive Assemblycited by 5
  9. 2024ContrastCAD: Contrastive Learning-based Representation Learning for Computer-Aided Design Modelscited by 5
  10. 2024SweepNet: Unsupervised Learning Shape Abstraction via Neural Sweeperscited by 5
  11. 2025GeoCAD: Local Geometry-Controllable CAD Generation with Large Language Modelscited by 5
  12. 2025MiCADangelo: Fine-Grained Reconstruction of Constrained CAD Models from 3D Scanscited by 5
  13. 2024Concise Plane Arrangements for Low-Poly Surface and Volume Modellingcited by 4
  14. 2026Can AI Understand the Language of Origami?cited by 4
  15. 2024MV2Cyl: Reconstructing 3D Extrusion Cylinders from Multi-View Imagescited by 3
  16. 2024CropCraft: Complete Structural Characterization of Crop Plants From Imagescited by 3
  17. 2025CADKnitter: Compositional CAD Generation from Text and Geometry Guidancecited by 2
  18. 2026Autoregressive B-Rep Shape Generation with Parametric Surfacescited by 2
  19. 2024PS-CAD: Local Geometry Guidance via Prompting and Selection for CAD Reconstructioncited by 1
  20. 2026CADENA: Stepwise CAD Reverse Engineeringcited by 1
  21. 2024NESI: Shape Representation via Neural Explicit Surface Intersectioncited by 0
  22. 2026AutoRegressive Generation with B-rep Holistic Token Sequence Representationcited by 0
  23. 2026Flatten The Complex: Joint B-Rep Generation via Compositional k-Cell Particlescited by 0
  24. 2026SharpNet: Enhancing MLPs to Represent Functions with Controlled Non-differentiabilitycited by 0
  25. 2026DualPrim: Compact 3D Reconstruction with Positive and Negative Primitivescited by 0
  26. 2026GeoFusion-CAD: Structure-Aware Diffusion with Geometric State Space for Parametric 3D Designcited by 0
  27. 2026HiFi-BRep: High-Fidelity Latent Representation for Robust B-Rep Generationcited by 0
  28. 2026Rethinking Human Intent-to-CAD: Parametric CAD Model Generation via Cooperative Multi-Task Alignment and Spatial-Aware Reinforcement Learningcited by 0