mlsanityData to 7 Oct 2026

3D-PRNN: Generating Shape Primitives with Recurrent Neural Networks

arXiv 1708.01648 ↗First posted 4 Aug 2017cs.CV cs.AI cs.LGcited by 77

Not on the map. Why?

Abstract (arXiv)

The success of various applications including robotics, digital content creation, and visualization demand a structured and abstract representation of the 3D world from limited sensor data. Inspired by the nature of human perception of 3D shapes as a collection of simple parts, we explore such an abstract shape representation based on primitives. Given a single depth image of an object, we present 3D-PRNN, a generative recurrent neural network that synthesizes multiple plausible shapes composed of a set of primitives. Our generative model encodes symmetry characteristics of common man-made objects, preserves long-range structural coherence, and describes objects of varying complexity with a compact representation. We also propose a method based on Gaussian Fields to generate a large scale dataset of primitive-based shape representations to train our network. We evaluate our approach on a wide range of examples and show that it outperforms nearest-neighbor based shape retrieval methods and is on-par with voxel-based generative models while using a significantly reduced parameter space.

Key facts. Cited by 77 papers in the map (data through 7 Oct 2026). First posted on arXiv 4 Aug 2017.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 23; 2022: 13; 2023: 19; 2024: 11; 2025: 8; 2026: 3.

77 papers.

Sort:

By area:

  1. 2021Im2Vec: Synthesizing Vector Graphics without Vector Supervisioncited by 114
  2. 2021A-SDF: Learning Disentangled Signed Distance Functions for Articulated Shape Representationcited by 113
  3. 2021Deep Generative Models in Engineering Design: A Reviewcited by 103
  4. 2021Neural Parts: Learning Expressive 3D Shape Abstractions with Invertible Neural Networkscited by 101
  5. 2022DM-NeRF: 3D Scene Geometry Decomposition and Manipulation from 2D Imagescited by 69
  6. 2021Deep Implicit Moving Least-Squares Functions for 3D Reconstructioncited by 67
  7. 2021CAPRI-Net: Learning Compact CAD Shapes with Adaptive Primitive Assemblycited by 62
  8. 2022Approximate Convex Decomposition for 3D Meshes with Collision-Aware Concavity and Tree Searchcited by 61
  9. 2022Unsupervised Point Cloud Representation Learning with Deep Neural Networks: A Surveycited by 54
  10. 2021Unsupervised Learning for Cuboid Shape Abstraction via Joint Segmentation from Point Cloudscited by 44
  11. 2021ARAPReg: An As-Rigid-As Possible Regularization Loss for Learning Deformable Shape Generatorscited by 40
  12. 2024MeshArt: Generating Articulated Meshes with Structure-Guided Transformerscited by 40
38 more
  1. 2023D2CSG: Unsupervised Learning of Compact CSG Trees with Dual Complements and Dropoutscited by 39
  2. 2021SAIL-VOS 3D: A Synthetic Dataset and Baselines for Object Detection and 3D Mesh Reconstruction from Video Datacited by 38
  3. 2023PartNeRF: Generating Part-Aware Editable 3D Shapes without 3D Supervisioncited by 36
  4. 2021HPNet: Deep Primitive Segmentation Using Hybrid Representationscited by 35
  5. 2022Point Primitive Transformer for Long-Term 4D Point Cloud Video Understandingcited by 35
  6. 2022Neural Template: Topology-aware Reconstruction and Disentangled Generation of 3D Meshescited by 34
  7. 2024Img2CAD: Reverse Engineering 3D CAD Models from Images through VLM-Assisted Conditional Factorizationcited by 34
  8. 2021PC2WF: 3D Wireframe Reconstruction from Raw Point Cloudscited by 33
  9. 2021Scene Synthesis via Uncertainty-Driven Attribute Synchronizationcited by 29
  10. 2021Robust and Accurate Superquadric Recovery: a Probabilistic Approachcited by 29
  11. 2023NVFi: Neural Velocity Fields for 3D Physics Learning from Dynamic Videoscited by 28
  12. 2021CPFN: Cascaded Primitive Fitting Networks for High-Resolution Point Cloudscited by 24
  13. 2022Few 'Zero Level Set'-Shot Learning of Shape Signed Distance Functions in Feature Spacecited by 22
  14. 2021Learning 3D Dense Correspondence via Canonical Point Autoencodercited by 21
  15. 2022RIM-Net: Recursive Implicit Fields for Unsupervised Learning of Hierarchical Shape Structurescited by 18
  16. 2023Marching-Primitives: Shape Abstraction from Signed Distance Functioncited by 18
  17. 2021Cuboids Revisited: Learning Robust 3D Shape Fitting to Single RGB Imagescited by 17
  18. 2025GaussianArt: Unified Modeling of Geometry and Motion for Articulated Objectscited by 17
  19. 2022RangeUDF: Semantic Surface Reconstruction from 3D Point Cloudscited by 16
  20. 2022ANISE: Assembly-based Neural Implicit Surface rEconstructioncited by 16
  21. 2023Single-Shot Implicit Morphable Faces with Consistent Texture Parameterizationcited by 16
  22. 2024Unsupervised Occupancy Learning from Sparse Point Cloudcited by 15
  23. 2022Primitive-based Shape Abstraction via Nonparametric Bayesian Inferencecited by 14
  24. 2024PASTA: Controllable Part-Aware Shape Generation with Autoregressive Transformerscited by 14
  25. 2025FreeGave: 3D Physics Learning from Dynamic Videos by Gaussian Velocitycited by 13
  26. 2023RayDF: Neural Ray-surface Distance Fields with Multi-view Consistencycited by 12
  27. 2023Mixing-Denoising Generalizable Occupancy Networkscited by 12
  28. 2025PrimitiveAnything: Human-Crafted 3D Primitive Assembly Generation with Auto-Regressive Transformercited by 12
  29. 2023DeFormer: Integrating Transformers with Deformable Models for 3D Shape Abstraction from a Single Imagecited by 11
  30. 2023DPF-Net: Combining Explicit Shape Priors in Deformable Primitive Field for Unsupervised Structural Reconstruction of 3D Objectscited by 10
  31. 2021Fostering Generalization in Single-view 3D Reconstruction by Learning a Hierarchy of Local and Global Shape Priorscited by 9
  32. 2025PartSDF: Part-Based Implicit Neural Representation for Composite 3D Shape Parametrization and Optimizationcited by 8
  33. 2021The Shape Part Slot Machine: Contact-based Reasoning for Generating 3D Shapes from Partscited by 7
  34. 2023Convex Decomposition of Indoor Scenescited by 6
  35. 2021Learning Mesh Representations via Binary Space Partitioning Tree Networkscited by 5
  36. 2021HybridSDF: Combining Deep Implicit Shapes and Geometric Primitives for 3D Shape Representation and Manipulationcited by 5
  37. 2023Learnable Earth Parser: Discovering 3D Prototypes in Aerial Scanscited by 5
  38. 2023Deep Deformable Models: Learning 3D Shape Abstractions with Part Consistencycited by 5

50 most cited of 77 All 77 (JSON)