mlsanityData to 7 Oct 2026

Convex Decomposition of Indoor Scenes

arXiv 2307.04246 ↗First posted 9 Jul 2023cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

We describe a method to parse a complex, cluttered indoor scene into primitives which offer a parsimonious abstraction of scene structure. Our primitives are simple convexes. Our method uses a learned regression procedure to parse a scene into a fixed number of convexes from RGBD input, and can optionally accept segmentations to improve the decomposition. The result is then polished with a descent method which adjusts the convexes to produce a very good fit, and greedily removes superfluous primitives. Because the entire scene is parsed, we can evaluate using traditional depth, normal, and segmentation error metrics. Our evaluation procedure demonstrates that the error from our primitive representation is comparable to that of predicting depth from a single image.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 28 of its 68 references that are in the map. First posted on arXiv 9 Jul 2023.

Builds on

28 of 68 references.

References · 28

  1. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  2. 2021Vision Transformers for Dense Predictioncited by 1,789
  3. 2019Towards Robust Monocular Depth Estimation: Mixing Datasets for Zero-shot Cross-dataset Transfercited by 1,484
  4. 2019From Big to Small: Multi-Scale Local Planar Guidance for Monocular Depth Estimationcited by 379
  5. 2019What Do Single-view 3D Reconstruction Networks Learn?cited by 259
  6. 2019BSP-Net: Generating Compact Meshes via Binary Space Partitioningcited by 205
  7. 2016Learning Shape Abstractions by Assembling Volumetric Primitivescited by 181
  8. 2019CvxNet: Learnable Convex Decompositioncited by 173
  9. 2019Superquadrics Revisited: Learning 3D Shape Parsing beyond Cuboidscited by 123
  10. 2021Estimating and Exploiting the Aleatoric Uncertainty in Surface Normal Estimationcited by 116
  11. 2018PlaneRCNN: 3D Plane Detection and Reconstruction from a Single Imagecited by 105
  12. 2018LayoutNet: Reconstructing the 3D Room Layout from a Single RGB Imagecited by 94
16 more
  1. 2018Supervised Fitting of Geometric Primitives to 3D Point Cloudscited by 90
  2. 20173D-PRNN: Generating Shape Primitives with Recurrent Neural Networkscited by 77
  3. 2018PlaneNet: Piece-wise Planar Reconstruction from a Single RGB Imagecited by 77
  4. 2021ShapeConv: Shape-aware Convolutional Layer for Indoor RGB-D Semantic Segmentationcited by 75
  5. 2022Approximate Convex Decomposition for 3D Meshes with Collision-Aware Concavity and Tree Searchcited by 61
  6. 2014Designing Deep Networks for Surface Normal Estimationcited by 59
  7. 2022Towards High-Fidelity Single-view Holistic Reconstruction of Indoor Scenescited by 43
  8. 2019Deep Parametric Shape Predictions using Distance Fieldscited by 30
  9. 2020Learning Generative Models of Shape Handlescited by 25
  10. 2020General 3D Room Layout from a Single View by Render-and-Comparecited by 21
  11. 2021Monte Carlo Scene Search for 3D Scene Understandingcited by 19
  12. 2021Cuboids Revisited: Learning Robust 3D Shape Fitting to Single RGB Imagescited by 17
  13. 2022MonteBoxFinder: Detecting and Filtering Primitives to Fit a Noisy Point Cloudcited by 11
  14. 2021LSD-StructureNet: Modeling Levels of Structural Detail in 3D Part Hierarchiescited by 8
  15. 2019Cerberus: A Multi-headed Derenderercited by 4
  16. 2023Blocks2World: Controlling Realistic Scenes with Editable Primitivescited by 4

Built on by

Cited by year

Citing papers per year: 2024: 3; 2025: 3.

6 papers.

  1. 2025Visual Jenga: Discovering Object Dependencies via Counterfactual Inpaintingcited by 11
  2. 2024Robust Shape Fitting for 3D Scene Abstractioncited by 3
  3. 2024Denoising Monte Carlo Renders with Diffusion Modelscited by 2
  4. 2025Generative Blocks World: Moving Things Around in Picturescited by 2
  5. 2025PrITTI: Primitive-based Generation of Controllable and Editable 3D Semantic Urban Scenescited by 1
  6. 2024Improved Convex Decomposition with Ensembling and Negative Primitivescited by 0