Research map / Point clouds / Shape and surface completion
Balanced Chamfer Distance as a Comprehensive Metric for Point Cloud Completion
Tong Wu, Liang Pan, Junzhe Zhang, Tai WANG, Ziwei Liu, Dahua Lin
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 |
|---|---|---|---|
| train with DCD loss instead of CD loss (ablation) | VRCNet · MVP completion | CD x10^4: 6.09 -> 5.85 | Improved |
| train with DCD loss instead of CD loss (ablation) | VRCNet · MVP completion | EMD x10^2: 5.27 -> 5.14 | Improved |
| train with DCD loss instead of CD loss (ablation) | PCN · MVP completion | EMD x10^2: 6.80 -> 4.71 | Improved |
| train with DCD loss instead of CD loss (ablation) | PCN · MVP completion | CD x10^4: 9.78 -> 9.36 | Improved |
| train with EMD loss instead of CD loss | PCN · MVP completion | CD x10^4: 9.78 -> 10.70 | Hurt |
| train with EMD loss instead of CD loss | VRCNet · MVP completion | CD x10^4: 6.09 -> 7.87 | Hurt |
| add additional point up-sampling (PU) to VRCNet with DCD loss (ablation) | VRCNet + DCD loss · MVP completion | EMD x10^2: 5.14 -> 3.68 | Improved |
| add additional point up-sampling (PU) (ablation) | VRCNet + DCD loss · MVP completion | CD x10^4: 5.85 -> 6.65 | Hurt |
| add guided down-sampling (GPD) with point discriminator (ablation) | VRCNet + DCD loss + PU · MVP completion | DCD: 0.422 -> 0.420 | Improved |
| add GPD (ablation) | VRCNet + DCD loss + PU · MVP completion | CD x10^4: 6.65 -> 6.51 | Improved |
Closest papers by meaning
- Fully Dynamic Algorithms for Chamfer Distance (NeurIPS 2025 · similarity 0.88)
- CMG-Net: Robust Normal Estimation for Point Clouds via Chamfer Normal Distance and Multi-Scale Geometry (AAAI 2024 · similarity 0.86)
- APML: Adaptive Probabilistic Matching Loss for Robust 3D Point Cloud Reconstruction (NeurIPS 2025 · similarity 0.85)
- SymmCompletion: High-Fidelity and High-Consistency Point Cloud Completion with Symmetry Guidance (AAAI 2025 · similarity 0.85)
- An in depth look at the Procrustes-Wasserstein distance: properties and barycenters (ICML 2025 · similarity 0.85)
- Beyond Chamfer Distance: Granular Order-aware Evaluation Metric For Online Mapping (NeurIPS 2026 · 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.
