Research map / Probabilistic and scientific modeling / Variational and generative sampling
Variational Graph-based Normal Integration
Lixiong Chen, Bohan Yu, Victor Adrian Prisacariu, Imari Sato
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 per-point triplet sampling (edge re-weighting) on scattered points (ablation) | GNI · DiLiGenT-MV front view, scattered, cow | MADE: 35.70 -> 18.72 | Improved |
| add triplet sampling on scattered points (ablation) | GNI · DiLiGenT-MV front view, scattered, reading | MADE: 10.14 -> 5.08 | Improved |
| add triplet sampling on scattered points (ablation) | GNI · DiLiGenT-MV side view, scattered, reading | MADE: 31.23 -> 17.62 | Improved |
| add triplet sampling on scattered points (ablation) | GNI · DiLiGenT-MV side view, scattered, pot | MADE: 19.35 -> 14.44 | Improved |
| use GNI point-to-plane graph formulation instead of BiNI on grid data (front view) | GNI · DiLiGenT-MV synthesized front, grid | MADE: cow 0.65 -> 5.62; reading 14.73 -> 2.42 | Mixed |
| use GNI instead of BiNI on grid data (side view) | GNI · DiLiGenT-MV synthesized side, grid | MADE: cow 9.84 -> 0.36; buddha 3.81 -> 4.04 | Mixed |
| use GNI instead of BiNI under perspective projection | GNI · Perspective DiLiGenT | point-to-plane distance: goblet 8.51 -> 9.75; reading 0.32 -> 0.26 | Mixed |
| switch from grid to scattered points (Halton sampled, about 20% kept) | GNI · DiLiGenT-MV front, cow | MADE: 5.62 -> 18.72 | Hurt |
Closest papers by meaning
- Discontinuity-preserving Normal Integration with Auxiliary Edges (CVPR 2024 · similarity 0.89)
- Bilateral Normal Integration (ECCV 2022 · similarity 0.89)
- Learning Normals of Noisy Points by Local Gradient-Aware Surface Filtering (ICCV 2025 · similarity 0.87)
- Flatten Anything: Unsupervised Neural Surface Parameterization (NeurIPS 2024 · similarity 0.86)
- Feature-Preserving Mesh Decimation for Normal Integration (CVPR 2025 · similarity 0.86)
- Learning Bijective Surface Parameterization for Inferring Signed Distance Functions from Sparse Point Clouds with Grid Deformation (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.
