Research map / Point clouds / Shape and surface completion
SSRNet: Scalable 3D Surface Reconstruction Network
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 local geometry-aware vertex features and the SSRNet reconstruction pipeline instead of ONet | Surface reconstruction · ShapeNet test | IoU: 0.778 -> 0.957 | Helped |
| Use local geometry-aware vertex features and the SSRNet reconstruction pipeline instead of ONet | Surface reconstruction · ShapeNet test | Chamfer-L1: 0.079 -> 0.024 | Helped |
| Use local geometry-aware vertex features and the SSRNet reconstruction pipeline instead of ONet | Surface reconstruction · ShapeNet test | Normal Consistency: 0.895 -> 0.967 | Helped |
| Use SSRNet rather than PSR trim 8 | Surface reconstruction · DTU test scenes | DTU Accuracy Mean: 0.473 -> 0.321 | Helped |
| Use SSRNet rather than PSR trim 8 | Surface reconstruction · DTU test scenes | DTU Completeness Mean: 0.327 -> 0.304 | Helped |
| Use SSRNet rather than PSR trim 9.5 | Surface reconstruction · DTU test scenes | DTU Completeness Mean: 0.345 -> 0.304 | Helped |
| Use SSRNet rather than PSR trim 9.5 | Surface reconstruction · DTU test scenes | CD Mean: 1.17 -> 1.46 | Hurt |
| Apply an SSRNet trained on ShapeNet directly to Stanford 3D | SSRNet · Stanford 3D Armadillo | CD Mean: 93.46 -> 0.028 | Helped |
| Use four GPUs for prediction instead of one | SSRNet · Large-scale point-cloud reconstruction | Prediction time: 133 -> 50.5 | Helped |
| Train on ShapeNet and transfer directly to Stanford 3D | SSRNet · Stanford 3D | Octree vertex classification accuracy: 98.1% | Helped |
Closest papers by meaning
- Scalable Surface Reconstruction From Point Clouds With Extreme Scale and Density Diversity (CVPR 2017 · similarity 0.88)
- EdgeMovingNet: Edge-preserving Point Cloud Reconstruction via Joint Geometry Features (CVPR 2025 · similarity 0.86)
- Arbitrary-Scale Point Cloud Upsampling by Voxel-Based Network with Latent Geometric-Consistent Learning (AAAI 2024 · similarity 0.86)
- LinNet: Linear Network for Efficient Point Cloud Representation Learning (NeurIPS 2024 · similarity 0.85)
- Learning Bijective Surface Parameterization for Inferring Signed Distance Functions from Sparse Point Clouds with Grid Deformation (CVPR 2025 · similarity 0.85)
- OffsetOPT: Explicit Surface Reconstruction without Normals (CVPR 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.