Research map / Point clouds / Shape and surface completion
Vis2Mesh: Efficient Mesh Reconstruction From Unstructured Point Clouds of Large Scenes With Learned Virtual View Visibility
Shuang Song, Zhaopeng Cui, Rongjun Qin
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 depth completion module (VDVNet cascade) (ablation) | visibility network · BlendedMVS pixel-wise visibility | F1 (%): 93.01 -> 93.63 | Improved |
| add depth completion when partial convolution is absent (ablation) | visibility network · BlendedMVS pixel-wise visibility | F1 (%): 87.71 -> 92.21 | Improved |
| add partial convolution layers (ablation) | visibility network · BlendedMVS pixel-wise visibility | F1 (%): 92.21 -> 93.63 | Improved |
| add partial convolution to UNet (ablation) | UNet · BlendedMVS pixel-wise visibility | F1 (%): 88.68 -> 89.32 | Improved |
| add partial convolution layers in full pipeline (ablation) | visibility network · BlendedMVS pixel-wise visibility | AUC (%): 93.14 -> 94.17 | Improved |
| use learned VDVNet instead of geometric HPR | visibility estimation · BlendedMVS pixel-wise visibility | AUC (%): 82.22 -> 94.17 | Improved |
| add adaptive visibility weighting (gamma_av=1) on outdoor scenes (ablation) | Vis2Mesh · Outdoor 500K (average) | F-score: 0.9634 -> 0.9654 | Improved |
| add adaptive visibility weighting on outdoor scenes (ablation) | Vis2Mesh · Outdoor 500K (average) | Chamfer distance: 0.2890 -> 0.2770 | Improved |
| add adaptive visibility weighting on object scenes (ablation) | Vis2Mesh · Object 25K (average) | Chamfer distance: 0.1959 -> 0.1846 | Improved |
| add adaptive visibility weighting on BlendedMVS scenes (ablation) | Vis2Mesh · BlendedMVS MVS (average) | F-score: 0.7215 -> 0.7321 | Improved |
| use virtual view visibility instead of real per-point LiDAR visibility | Vis2Mesh · Mai City Dataset | visual quality | Improved |
| increase number of virtual views (ablation) | Vis2Mesh · qualitative scene | visual detail | Improved |
Closest papers by meaning
- Visibility-Aware Point-Based Multi-View Stereo Network (TPAMI 2020 · similarity 0.89)
- Point Scene Understanding via Disentangled Instance Mesh Reconstruction (ECCV 2022 · similarity 0.89)
- Meshing Point Clouds with Predicted Intrinsic-Extrinsic Ratio Guidance (ECCV 2020 · similarity 0.89)
- High-Fidelity Lightweight Mesh Reconstruction from Point Clouds (CVPR 2025 · similarity 0.86)
- SPAR3D: Stable Point-Aware Reconstruction of 3D Objects from Single Images (CVPR 2025 · similarity 0.86)
- Point2Real: Bridging the Gap between Point Cloud and Realistic Image for Open-World 3D Recognition (AAAI 2024 · 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.
