Research map / Point clouds / Shape and surface completion
PU-EVA: An Edge-Vector Based Approximation Solution for Flexible-Scale Point Cloud Upsampling
Luqing Luo, Lulu Tang, Wanyi Zhou, Shizheng Wang, Zhi-Xin Yang
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 |
|---|---|---|---|
| add second-order error term of Taylor's expansion (ablation) | PU-EVA · 4x, Sparse (256) input | CD / HD (x1e-3): CD 2.318 -> 1.784; HD 18.713 -> 13.939 | Improved |
| add second-order error term of Taylor's expansion (ablation) | PU-EVA · 4x, Medium (2,048) input | CD / HD (x1e-3): CD 0.312 -> 0.266; HD 4.622 -> 3.070 | Improved |
| add second-order error term of Taylor's expansion (ablation) | PU-EVA · 4x, Dense (4,096) input | CD / HD (x1e-3): CD 0.157 -> 0.123; HD 5.535 -> 1.394 | Improved |
| use EVA upsampling unit instead of feature duplication (ablation) | PU-EVA · 4x, Sparse (256) input | CD / HD (x1e-3): CD 2.94 -> 1.78; HD 24.67 -> 13.93 | Improved |
| use EVA upsampling unit instead of feature duplication (ablation) | PU-EVA · 4x, Medium (2,048) input | CD / HD (x1e-3): CD 0.38 -> 0.26; HD 3.99 -> 3.07 | Improved |
| use EVA upsampling unit instead of feature duplication (ablation) | PU-EVA · 4x, Dense (4,096) input | CD / HD (x1e-3): CD 0.13 -> 0.12; HD 4.47 -> 1.39 | Improved |
| use EVA upsampling unit instead of NodeShuffle (ablation) | PU-EVA · 4x, Sparse (256) input | CD / HD (x1e-3): CD 1.86 -> 1.78; HD 15.56 -> 13.93 | Improved |
| use EVA upsampling unit instead of NodeShuffle (ablation) | PU-EVA · 4x, Medium (2,048) input | CD / HD (x1e-3): CD 0.27 -> 0.26; HD 4.66 -> 3.07 | Improved |
| use EVA upsampling unit instead of NodeShuffle (ablation) | PU-EVA · 4x, Dense (4,096) input | CD / HD (x1e-3): CD 0.12 -> 0.12; HD 3.11 -> 1.39 | Improved |
| use PU-EVA instead of PU-GAN under input noise | PU-EVA · Gaussian noise sigma=0.02 | CD / HD (x1e-3): CD 0.912 -> 0.864; HD 11.326 -> 9.178 | Improved |
| use PU-EVA instead of PU-GAN on dense input | PU-EVA · 4x, Dense (4,096) input | HD (x1e-3): HD 1.284 -> 1.394 | Hurt |
| use PU-EVA instead of PU-GAN on sparse input | PU-EVA · 4x, Sparse (256) input | HD / P2F (x1e-3): CD 2.072 -> 1.784; HD 16.592 -> 13.939; P2F 8.055 -> 8.727 | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- PU-GAN — PU-GAN: A Point Cloud Upsampling Adversarial Network (ICCV 2019) · better on 14 of 17 shared results
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
- Arbitrary-Scale Point Cloud Upsampling by Voxel-Based Network with Latent Geometric-Consistent Learning (AAAI 2024 · similarity 0.90)
- SPU-IMR: Self-supervised Arbitrary-scale Point Cloud Upsampling via Iterative Mask-recovery Network (AAAI 2025 · similarity 0.89)
- A Conditional Denoising Diffusion Probabilistic Model for Point Cloud Upsampling (CVPR 2024 · similarity 0.88)
- Learning Continuous Implicit Field with Local Distance Indicator for Arbitrary-Scale Point Cloud Upsampling (AAAI 2024 · similarity 0.88)
- PLATYPUS: Progressive Local Surface Estimator for Arbitrary-Scale Point Cloud Upsampling (AAAI 2025 · similarity 0.86)
- EdgeMovingNet: Edge-preserving Point Cloud Reconstruction via Joint Geometry Features (CVPR 2025 · similarity 0.86)
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.
