mlsanityData to 7 Oct 2026

FSC: Few-point Shape Completion

arXiv 2403.07359 ↗First posted 12 Mar 2024cs.CVcited by 6

Not on the map. Why?

Abstract (arXiv)

While previous studies have demonstrated successful 3D object shape completion with a sufficient number of points, they often fail in scenarios when a few points, e.g. tens of points, are observed. Surprisingly, via entropy analysis, we find that even a few points, e.g. 64 points, could retain substantial information to help recover the 3D shape of the object. To address the challenge of shape completion with very sparse point clouds, we then propose Few-point Shape Completion (FSC) model, which contains a novel dual-branch feature extractor for handling extremely sparse inputs, coupled with an extensive branch for maximal point utilization with a saliency branch for dynamic importance assignment. This model is further bolstered by a two-stage revision network that refines both the extracted features and the decoder output, enhancing the detail and authenticity of the completed point cloud. Our experiments demonstrate the feasibility of recovering 3D shapes from a few points. The proposed Few-point Shape Completion (FSC) model outperforms previous methods on both few-point inputs and many-point inputs, and shows good generalizability to different object categories.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 49 of its 77 references that are in the map. First posted on arXiv 12 Mar 2024.

Builds on

49 of 77 references.

References · 49

  1. 2016PointNet: Deep Learning on Point Sets for 3D Classification and Segmentationcited by 5,223
  2. 2017PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Spacecited by 4,384
  3. 2015ShapeNet: An Information-Rich 3D Model Repositorycited by 2,880
  4. 2018Dynamic Graph CNN for Learning on Point Cloudscited by 2,220
  5. 2020Point Transformercited by 1,462
  6. 2016A Point Set Generation Network for 3D Object Reconstruction from a Single Imagecited by 972
  7. 2020PCT: Point cloud transformercited by 730
  8. 2018AtlasNet: A Papier-Mach'e Approach to Learning 3D Surface Generationcited by 539
  9. 2022Masked Autoencoders for Point Cloud Self-supervised Learningcited by 456
  10. 2022PointNeXt: Revisiting PointNet++ with Improved Training and Scaling Strategiescited by 432
  11. 2018PCN: Point Completion Networkcited by 382
  12. 2021PoinTr: Diverse Point Cloud Completion with Geometry-Aware Transformerscited by 284
37 more
  1. 2016Shape Completion using 3D-Encoder-Predictor CNNs and Shape Synthesiscited by 219
  2. 2020GRNet: Gridding Residual Network for Dense Point Cloud Completioncited by 174
  3. 2021SnowflakeNet: Point Cloud Completion by Snowflake Point Deconvolution with Skip-Transformercited by 162
  4. 2020PF-Net: Point Fractal Network for 3D Point Cloud Completioncited by 146
  5. 2021Revisiting Point Cloud Shape Classification with a Simple and Effective Baselinecited by 145
  6. 2019Morphing and Sampling Network for Dense Point Cloud Completioncited by 132
  7. 2021Beyond Self-attention: External Attention using Two Linear Layers for Visual Taskscited by 109
  8. 2020Point Cloud Completion by Skip-attention Network with Hierarchical Foldingcited by 106
  9. 2020PointMixup: Augmentation for Point Cloudscited by 104
  10. 2021Variational Relational Point Completion Networkcited by 103
  11. 2020PMP-Net: Point Cloud Completion by Learning Multi-step Point Moving Pathscited by 97
  12. 2022SeedFormer: Patch Seeds based Point Cloud Completion with Upsample Transformercited by 83
  13. 2021PointMixer: MLP-Mixer for Point Cloud Understandingcited by 71
  14. 2020Detail Preserved Point Cloud Completion via Separated Feature Aggregationcited by 70
  15. 2021Unsupervised 3D Shape Completion through GAN Inversioncited by 70
  16. 2022Benchmarking and Analyzing Point Cloud Classification under Corruptionscited by 70
  17. 2023AdaPoinTr: Diverse Point Cloud Completion with Adaptive Geometry-Aware Transformerscited by 65
  18. 2021Regularization Strategy for Point Cloud via Rigidly Mixed Samplecited by 59
  19. 2022PMP-Net++: Point Cloud Completion by Transformer-Enhanced Multi-step Point Moving Pathscited by 57
  20. 2021Cycle4Completion: Unpaired Point Cloud Completion using Cycle Transformation with Missing Region Codingcited by 50
  21. 2021View-Guided Point Cloud Completioncited by 46
  22. 2021Style-based Point Generator with Adversarial Rendering for Point Cloud Completioncited by 45
  23. 2022PointAttN: You Only Need Attention for Point Cloud Completioncited by 43
  24. 2021Point Cloud Augmentation with Weighted Local Transformationscited by 39
  25. 2023SVDFormer: Complementing Point Cloud via Self-view Augmentation and Self-structure Dual-generatorcited by 34
  26. 2020SoftPoolNet: Shape Descriptor for Point Cloud Completion and Classificationcited by 33
  27. 2021Voxel-based Network for Shape Completion by Leveraging Edge Generationcited by 32
  28. 2021ASFM-Net: Asymmetrical Siamese Feature Matching Network for Point Completioncited by 28
  29. 2023Unsupervised Inference of Signed Distance Functions from Single Sparse Point Clouds without Learning Priorscited by 28
  30. 2021PC-HMR: Pose Calibration for 3D Human Mesh Recovery from 2D Images/Videoscited by 25
  31. 2020Skeleton-bridged Point Completion: From Global Inference to Local Adjustmentcited by 23
  32. 2020Cascaded Refinement Network for Point Cloud Completion with Self-supervisioncited by 20
  33. 2023High Fidelity 3D Hand Shape Reconstruction via Scalable Graph Frequency Decompositioncited by 18
  34. 2020Point Cloud Completion by Learning Shape Priorscited by 14
  35. 2024Spectrum AUC Difference (SAUCD): Human-aligned 3D Shape Evaluationcited by 9
  36. 2023EPiC: Ensemble of Partial Point Clouds for Robust Classificationcited by 5
  37. 2019Latent-Space Laplacian Pyramids for Adversarial Representation Learning with 3D Point Cloudscited by 4

Built on by

Cited by year

Citing papers per year: 2024: 2; 2025: 4.

6 papers.

  1. 2025LightGen: Efficient Image Generation through Knowledge Distillation and Direct Preference Optimizationcited by 20
  2. 2024MH-pFLID: Model Heterogeneous personalized Federated Learning via Injection and Distillation for Medical Data Analysiscited by 14
  3. 2025Temporal Regularization Makes Your Video Generator Strongercited by 10
  4. 2024Deep Learning for 3D Point Cloud Enhancement: A Surveycited by 4
  5. 2025Niagara: Normal-Integrated Geometric Affine Field for Scene Reconstruction from a Single Viewcited by 3
  6. 2025HeartFormer: Semantic-Aware Dual-Structure Transformers for 3D Four-Chamber Cardiac Point Cloud Reconstructioncited by 1