mlsanityData to 7 Oct 2026

APC2Mesh: Bridging the gap from occluded building faccades to full 3D models

arXiv 2404.02391 ↗First posted 3 Apr 2024cs.CVcited by 2

Not on the map. Why?

Abstract (arXiv)

The benefits of having digital twins of urban buildings are numerous. However, a major difficulty encountered in their creation from airborne LiDAR point clouds is the effective means of accurately reconstructing significant occlusions amidst point density variations and noise. To bridge the noise/sparsity/occlusion gap and generate high fidelity 3D building models, we propose APC2Mesh which integrates point completion into a 3D reconstruction pipeline, enabling the learning of dense geometrically accurate representation of buildings. Specifically, we leveraged complete points generated from occluded ones as input to a linearized skip attention-based deformation network for 3D mesh reconstruction. In our experiments, conducted on 3 different scenes, we demonstrate that: (1) APC2Mesh delivers comparatively superior results, indicating its efficacy in handling the challenges of occluded airborne building points of diverse styles and complexities. (2) The combination of point completion with typical deep learning-based 3D point cloud reconstruction methods offers a direct and effective solution for reconstructing significantly occluded airborne building points. As such, this neural integration holds promise for advancing the creation of digital twins for urban buildings with greater accuracy and fidelity.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 35 of its 60 references that are in the map. First posted on arXiv 3 Apr 2024.

Builds on

35 of 60 references.

References · 35

  1. 2017Attention Is All You Needcited by 59,932
  2. 2019DeepSDF: Learning Continuous Signed Distance Functions for Shape Representationcited by 2,682
  3. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  4. 2020Convolutional Occupancy Networkscited by 760
  5. 2018PointConv: Deep Convolutional Networks on 3D Point Cloudscited by 656
  6. 2018AtlasNet: A Papier-Mach'e Approach to Learning 3D Surface Generationcited by 539
  7. 2019SAL: Sign Agnostic Learning of Shapes from Raw Datacited by 389
  8. 2018PCN: Point Completion Networkcited by 382
  9. 2020Deep Local Shapes: Learning Local SDF Priors for Detailed 3D Reconstructioncited by 321
  10. 2020Neural Unsigned Distance Fields for Implicit Function Learningcited by 265
  11. 2017Octree Generating Networks: Efficient Convolutional Architectures for High-resolution 3D Outputscited by 189
  12. 2017Learning Efficient Point Cloud Generation for Dense 3D Object Reconstructioncited by 133
23 more
  1. 2020Points2Surf: Learning Implicit Surfaces from Point Cloud Patchescited by 122
  2. 2020SALD: Sign Agnostic Learning with Derivativescited by 115
  3. 2020Point Cloud Completion by Skip-attention Network with Hierarchical Foldingcited by 106
  4. 2022ShapeFormer: Transformer-based Shape Completion via Sparse Representationcited by 106
  5. 2021Variational Relational Point Completion Networkcited by 103
  6. 2022POCO: Point Convolution for Surface Reconstructioncited by 91
  7. 2020Deep Implicit Templates for 3D Shape Representationcited by 86
  8. 2020Learning Deformable Tetrahedral Meshes for 3D Reconstructioncited by 84
  9. 2022SeedFormer: Patch Seeds based Point Cloud Completion with Upsample Transformercited by 83
  10. 2020DualSDF: Semantic Shape Manipulation using a Two-Level Representationcited by 82
  11. 2021PointMixer: MLP-Mixer for Point Cloud Understandingcited by 71
  12. 2018MeshCNN: A Network with an Edgecited by 69
  13. 2020PIE-NET: Parametric Inference of Point Cloud Edgescited by 67
  14. 2019RL-GAN-Net: A Reinforcement Learning Agent Controlled GAN Network for Real-Time Point Cloud Shape Completioncited by 58
  15. 2021Neural Fields as Learnable Kernels for 3D Reconstructioncited by 58
  16. 2021PC2WF: 3D Wireframe Reconstruction from Raw Point Cloudscited by 33
  17. 2020Sign-Agnostic Implicit Learning of Surface Self-Similarities for Shape Modeling and Reconstruction from Raw Point Cloudscited by 30
  18. 2022LAKe-Net: Topology-Aware Point Cloud Completion by Localizing Aligned Keypointscited by 27
  19. 2023Building3D: An Urban-Scale Dataset and Benchmarks for Learning Roof Structures from Point Cloudscited by 24
  20. 2022Few 'Zero Level Set'-Shot Learning of Shape Signed Distance Functions in Feature Spacecited by 22
  21. 2021Reconstructing Compact Building Models from Point Clouds Using Deep Implicit Fieldscited by 20
  22. 2022ImpliCity: City Modeling from Satellite Images with Deep Implicit Occupancy Fieldscited by 9
  23. 2023Feature-assisted interactive geometry reconstruction in 3D point clouds using incremental region growingcited by 1

Built on by

Cited by year

Citing papers per year: 2025: 2.

2 papers.

  1. 2025Explaining Human Preferences via Metrics for Structured 3D Reconstructioncited by 3
  2. 2025OCCDiff: Occupancy Diffusion Model for High-Fidelity 3D Building Reconstruction from Noisy Point Cloudscited by 0