Research map / 3D vision / Face and shape
TetraSDF: Analytic Isosurface Extraction with Multi-resolution Tetrahedral Grid
Seonghun Oh, Youngjung Uh, Jin-Hwa Kim
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 input preconditioner A* (ablation) | TetraSDF · Thingi10K (Large) | Chamfer Distance | Improved |
| add encoder-derived preconditioner A* (ablation) | HashGrid · Thingi10K (Large) | Chamfer Distance | Hurt |
| replace ReLU MLP (AM) with tetrahedral encoder + small ReLU MLP | analytic extraction · Stanford (Large) | Chamfer Distance x1e-6: 5475 -> 1718 | Improved |
| replace trilinear HashGrid encoder (TropicalNeRF) with tetrahedral barycentric encoder | analytic extraction · Stanford (Small) | Chamfer Distance x1e-6: 5367 -> 4210 | Improved |
| replace trilinear HashGrid encoder (TropicalNeRF) with tetrahedral barycentric encoder | analytic extraction · Stanford (Large) | Chamfer Distance x1e-6: 1737 -> 1718 | Improved |
| replace TropicalNeRF extraction with exact analytic extraction | TetraSDF · Thingi10K | SSDF x1e-6: 144.1 -> 0.078 | Improved |
| analytic extraction instead of Marching Cubes at 1024 | TetraSDF network · Thingi10K | SSDF x1e-6: 17.6 -> 0.078 | Improved |
| increase Marching Cubes resolution from 256 to 1024 | TetraSDF network · Thingi10K | SSDF x1e-6: 217.0 -> 17.6 | Improved |
| use tetrahedral encoder instead of HashGrid | SDF fitting · Stanford (Large) | GT-mesh SDF MAE: 0.53 -> 0.62 | No clear effect |
| use tetrahedral encoder instead of HashGrid | SDF fitting · Stanford (Small) | GT-mesh SDF MAE: 3.24 -> 2.82 | Improved |
| enlarge plain ReLU MLP for Analytic Marching from H10/W90 to H20/W200 | AM · Stanford | Chamfer Distance x1e-6: 5475 -> 5062 | No clear effect |
| use TetraSDF instead of adaptive octree extractors | extraction on Large network · Stanford | SSDF x1e-6: 8.85 -> 0.076 | Improved |
Closest papers by meaning
- Polyhedral Complex Derivation from Piecewise Trilinear Networks (NeurIPS 2024 · similarity 0.87)
- Details Enhancement in Unsigned Distance Field Learning for High-fidelity 3D Surface Reconstruction (AAAI 2025 · similarity 0.85)
- Neural Signed Distance Function Inference through Splatting 3D Gaussians Pulled on Zero-Level Set (NeurIPS 2024 · similarity 0.85)
- DUDF: Differentiable Unsigned Distance Fields with Hyperbolic Scaling (CVPR 2024 · similarity 0.85)
- Small Steps and Level Sets: Fitting Neural Surface Models with Point Guidance (CVPR 2024 · similarity 0.85)
- High Resolution UDF Meshing via Iterative Networks (NeurIPS 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.
