Research map / Point clouds / Shape and surface completion
SIGMA: Scale-Invariant Global Sparse Shape Matching
Maolin Gao, Paul Roetzer, Marvin Eisenberger, Zorah Lähner, Michael Moeller, Daniel Cremers, Florian Bernard
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 |
|---|---|---|---|
| replace MINA's rotation-matrix extrinsic deformation model with the projected LBO formulation | SIGMA · TOSCA (71 pairs) | pairs certified globally optimal: 8 -> 56 of 71 | Improved |
| add orientation-aware term to disambiguate intrinsic symmetries | SIGMA · TOSCA, SMAL, SHREC20, DT4D-M PCK | PCK AUC | Improved |
| add orientation-aware term on DT4D-M | SIGMA · DT4D-M | PCK AUC | No clear effect |
| use projected LBO instead of standard LBO as deformation prior | SIGMA shape reconstruction · reconstruction qualitative | reconstruction quality | Improved |
| use conservative pruning k=11 instead of aggressive k=5 | SIGMA · TOSCA | matching accuracy | Improved |
| reduce pruning to k=5 for faster optimisation (ablation) | SIGMA · TOSCA (71 pairs) | runtime to global optimum: all 71 pairs within 15 minutes | Improved |
| replace PMSDP's default AGD pruning with the paper's pruning | PMSDP · TOSCA | accuracy | Improved |
| apply method to partial shapes | SIGMA · partial matching | global optimality within time limit | Hurt |
| rescale one shape by factors 0.1 to 10 | SIGMA and PMSDP · SHREC20 (5 pairs) | mean geodesic error | No clear effect |
Closest papers by meaning
- MINA: Convex Mixed-Integer Programming for Non-Rigid Shape Alignment (CVPR 2020 · similarity 0.89)
- Probabilistic Deformation Consistency for Unsupervised Shape Matching (AAAI 2026 · similarity 0.88)
- SE(n)-Invariant Flow Matching: A General Framework with Application to Object Reassembly (ICML 2026 · similarity 0.88)
- Fast Globally Optimal and Geometrically Consistent 3D Shape Matching (ICCV 2025 · similarity 0.88)
- SpiderMatch: 3D Shape Matching with Global Optimality and Geometric Consistency (CVPR 2024 · similarity 0.86)
- Spectral Meets Spatial: Harmonising 3D Shape Matching and Interpolation (CVPR 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.
