Research map / Point clouds / Shape and surface completion
DiGS: Divergence Guided Shape Implicit Neural Representation for Unoriented Point Clouds
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 |
|---|---|---|---|
| add the divergence-guided loss to SIREN without normal supervision | SIREN · SRB | Chamfer distance: 0.42 -> 0.19 | Helped |
| add the divergence-guided loss to SIREN without normal supervision | SIREN · SRB | Hausdorff distance: 7.67 -> 3.52 | Helped |
| add divergence guidance without normal supervision | ShapeNet reconstruction · ShapeNet (260 shapes, no normals) | IoU mean: 0.3085 -> 0.9390 | Helped |
| add normal supervision (ablation) | DiGS · ShapeNet (260 shapes) | squared Chamfer median: 2.55e-5 -> 2.32e-5 | No clear effect |
| add normal supervision (ablation) | DiGS · ShapeNet (260 shapes) | IoU mean: 0.9390 -> 0.9200 | Hurt |
| add the divergence constraint (ablation) | A+C2 toy SDF training · Toy problem (200x200 grid) | mean Dirichlet energy: 3.58 ± 3.55 -> 1.37 ± 1.35 | Helped |
| add the divergence constraint (ablation) | A+C2 toy SDF training · Toy problem (200x200 grid) | mean gradient norm: 1.51 ± 0.54 -> 1.04 ± 0.39 | Helped |
| omit the divergence constraint (ablation) | A+C2 toy SDF training · Toy problem | stability of sign and contour spacing | Hurt |
| use MFGI rather than geometric initialization alone (ablation) | DiGS · Koch's snowflake (2D) | reconstruction quality | Helped |
| anneal the divergence constraint rather than keeping it constant (ablation) | DiGS · Koch's snowflake (2D) | reconstruction quality | Helped |
| remove normal supervision (ablation) | IGR · DFaust | one-sided Chamfer (reg to recon): 1.053 -> 3.745 | Hurt |
| add the divergence-guided method without normals | IGR without normals · DFaust | one-sided Chamfer (reg to recon): 3.745 -> 0.856 | Helped |
Closest papers by meaning
- Small Steps and Level Sets: Fitting Neural Surface Models with Point Guidance (CVPR 2024 · similarity 0.88)
- 4Deform: Neural Surface Deformation for Robust Shape Interpolation (CVPR 2025 · similarity 0.88)
- NeuralIndicator: Implicit Surface Reconstruction from Neural Indicator Priors (ICML 2024 · similarity 0.88)
- LISR: Learning Linear 3D Implicit Surface Representation Using Compactly Supported Radial Basis Functions (AAAI 2024 · similarity 0.87)
- Learning Bijective Surface Parameterization for Inferring Signed Distance Functions from Sparse Point Clouds with Grid Deformation (CVPR 2025 · similarity 0.86)
- A Lightweight UDF Learning Framework for 3D Reconstruction Based on Local Shape Functions (CVPR 2025 · 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.