Shape and surface completion: research map
558 papers on Shape and surface completion in Point clouds, from ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR (2016–2026), grouped into 2 clusters and 6 approaches. The busiest year so far is 2025.
What it is
Scans are usually incomplete and noisy. This topic completes missing parts of shapes, upsamples sparse points, estimates normals and reconstructs smooth surfaces from points, and generates point clouds.
Terms you will see on this page:
- Shape completion
- predicting the missing parts of a partial scan.
- Surface reconstruction
- turning points into a continuous surface or mesh.
- Normal estimation
- estimating the surface direction at each point.
Start reading with the most cited papers here:
- A Point Set Generation Network for 3D Object Reconstruction From a Single Image (CVPR 2017 · 2,524 citations)
- PU-Net: Point Cloud Upsampling Network (CVPR 2018 · 689 citations)
Within Point clouds, its share grew from 19.5% in 2023–24 to 22.2% in 2025–26 (89 → 113 papers at ICML, NeurIPS, ICLR, CVPR and AAAI, the venues with data for all four years).
Explore Shape and surface completion in the interactive map
Working on something in this topic? Describe your idea in scime atlas to see which approach it falls under, the closest papers by meaning and how crowded the spot has become.
What changed recently
Approaches whose share of the topic moved most between 2023–24 and 2025–26 (ICML, NeurIPS, ICLR, CVPR and AAAI):
- Point cloud diffusion generation rising: 12% → 21% of the topic (11 → 24 papers)
- Point cloud completion networks rising: 12% → 17% of the topic (11 → 19 papers)
- Point cloud upsampling and normals rising: 21% → 26% of the topic (19 → 29 papers)
- Distance-function surface reconstruction falling: 20% → 11% of the topic (18 → 12 papers)
- Unsupervised shape correspondence falling: 24% → 19% of the topic (21 → 22 papers)
Papers from 2025–2026 that are already cited most:
- SPAR3D: Stable Point-Aware Reconstruction of 3D Objects from Single Images (CVPR 2025 · 21 citations)
- EdgeDiff: Edge-aware Diffusion Network for Building Reconstruction from Point Clouds (CVPR 2025 · 12 citations)
- SymmCompletion: High-Fidelity and High-Consistency Point Cloud Completion with Symmetry Guidance (AAAI 2025 · 11 citations)
Approaches and key papers
Surface reconstruction and generation · 446 papers
Keywords: reconstruction, implicit, meshes
Approaches in this cluster:
- Unsupervised shape correspondence (182 papers)
Learn dense shape correspondence and implicit structure from raw point clouds. - Point cloud upsampling and normals (120 papers)
Upsample point clouds and estimate normals with learned surface refinement. - Distance-function surface reconstruction (74 papers)
Reconstruct surfaces by learning signed or unsigned distance functions from point clouds. - Point cloud diffusion generation (70 papers)
Generate 3D shapes with diffusion models over point clouds and latent spaces.
Most cited and most cited since 2024:
- A Point Set Generation Network for 3D Object Reconstruction From a Single Image (CVPR 2017 · 2,524 citations)
- PU-Net: Point Cloud Upsampling Network (CVPR 2018 · 689 citations)
- Point2CAD: Reverse Engineering CAD Models from 3D Point Clouds (CVPR 2024 · 42 citations)
- A Conditional Denoising Diffusion Probabilistic Model for Point Cloud Upsampling (CVPR 2024 · 32 citations)
Point cloud completion · 112 papers
Keywords: cloud completion, complete, shape completion
Approaches in this cluster:
- Point cloud completion networks (67 papers)
Complete partial point clouds with refinement, folding and cycle-based unpaired learning. - Structure-guided completion (45 papers)
Complete shapes using skip-attention, structure priors, diffusion refinement and CLIP guidance.
Most cited and most cited since 2024:
- TopNet: Structural Point Cloud Decoder (CVPR 2019 · 532 citations)
- PF-Net: Point Fractal Network for 3D Point Cloud Completion (CVPR 2020 · 510 citations)
- PointAttN: You Only Need Attention for Point Cloud Completion (AAAI 2024 · 78 citations)
- Orthogonal Dictionary Guided Shape Completion Network for Point Cloud (AAAI 2024 · 31 citations)
Related topics in Point clouds
- Point cloud classification (1,097)
- Registration and matching (272)
- Segmentation and detection (567)
