Research map / Point clouds / Shape and surface completion
Parametric Surface Constrained Upsampler Network for Point Cloud
Pingping Cai, Zhenyao Wu, Xinyi Wu, Song Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use the parametric surface constrained upsampler instead of PU-GCN | Point cloud upsampling · PU1K | CD (× 10^-4): 1.151 -> 0.886 | Improved |
| Use the parametric surface constrained upsampler instead of PU-GCN | Point cloud upsampling · PU1K | P2F (× 10^-3): 2.504 -> 1.091 | Improved |
| Use the parametric surface constrained upsampler instead of PU-GCN | Point cloud upsampling · PU1K | Uniformity (× 10^-3), 0.4%, 0.6%, 0.8%, 1.0%: 1.95 -> 1.40; 3.97 -> 2.85; 6.83 -> 5.06; 10.63 -> 7.95 | Improved |
| Use the parametric surface constrained upsampler instead of SnowFlake | Point cloud completion · ShapeNet-PCN | L1-CD (× 10^-3): 7.19 -> 7.04 | Improved |
| Use the parametric surface constrained upsampler instead of SnowFlake | Point cloud completion · ShapeNet-PCN | EMD (× 10^-3): 69.13 -> 66.57 | Hurt |
| Add spatial feature aggregation with a point transformer over nearest neighbors (ablation) | Parametric Surface Constrained Upsampler · PU1K | CD (× 10^-4): 1.293 -> 0.886 | Improved |
| Use explicit parametric surface functions instead of an MLP folding layer (ablation) | Parametric Surface Constrained Upsampler · PU1K | CD (× 10^-4): 0.967 -> 0.886 | Improved |
| Add displacement loss to select a better projection plane (ablation) | Parametric Surface Constrained Upsampler · PU1K | CD (× 10^-4): 0.893 -> 0.886 | Improved |
| Add Gaussian perturbation to input points | Parametric Surface Constrained Upsampler · PU1K | CD (× 10^-4): 0.886 -> 1.053 -> 1.280 -> 1.599 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- PU-GCN — PU-GCN: Point Cloud Upsampling Using Graph Convolutional Networks (CVPR 2021) · better on 6 of 6 shared results
- GR-Net — GRNet: Gridding Residual Network for Dense Point Cloud Completion (ECCV 2020) · better on 0 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- PU-Net: Point Cloud Upsampling Network (CVPR 2018 · similarity 0.91)
- SPU-IMR: Self-supervised Arbitrary-scale Point Cloud Upsampling via Iterative Mask-recovery Network (AAAI 2025 · similarity 0.89)
- Arbitrary-Scale Point Cloud Upsampling by Voxel-Based Network with Latent Geometric-Consistent Learning (AAAI 2024 · similarity 0.89)
- Learning Continuous Implicit Field with Local Distance Indicator for Arbitrary-Scale Point Cloud Upsampling (AAAI 2024 · similarity 0.88)
- SPU-PMD: Self-Supervised Point Cloud Upsampling via Progressive Mesh Deformation (CVPR 2024 · similarity 0.85)
- PLATYPUS: Progressive Local Surface Estimator for Arbitrary-Scale Point Cloud Upsampling (AAAI 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
