Research map / 3D vision / Face and shape
Implicit Field Supervision for Robust Non-rigid Shape Matching
Ramana Sundararaman, Gautam Pai, Maks Ovsjanikov
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 SDFNet implicit surface supervision of the latent embedding (ablation) | Ours · SHREC'19 | mean geodesic error (cm): 11.6 -> 6.5 | Improved |
| add SDFNet implicit surface supervision of the latent embedding (ablation) | Ours · FAUST | mean geodesic error (cm): 14.8 -> 2.6 | Improved |
| add off-surface field regularisations (ablation) | Ours · SHREC'19 | mean geodesic error (cm): 7.3 -> 6.5 | Improved |
| add off-surface field regularisations (ablation) | Ours · FAUST | mean geodesic error (cm): 4.9 -> 2.6 | Improved |
| add signed distance regularisation L_SDR at train and test (ablation) | Ours · SHREC'19 | mean geodesic error (cm): 7.5 -> 6.5 | Improved |
| add signed distance regularisation L_SDR at test time (ablation) | Ours · FAUST | mean geodesic error (cm): 3.6 -> 2.6 | Improved |
| add surface loss L_surf (ablation) | Ours · FAUST | mean geodesic error (cm): 26.8 -> 2.6 | Improved |
| add surface loss L_surf (ablation) | Ours · SHREC'19 | mean geodesic error (cm): 17.0 -> 6.5 | Improved |
| add Chamfer distance optimization at inference (ablation) | Ours · SHREC'19 | mean geodesic error (cm): 10.8 -> 6.5 | Improved |
| train on more shapes (100 to 2000) from SURREAL (ablation) | Ours vs 3D-CODED · SHREC'19 scenarios | geodesic error | No clear effect |
| use 2k training shapes versus Oracle baselines trained on 230k with augmentation | Ours · noisy and cluttered SHREC'19/FAUST | geodesic accuracy | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- 3D-CODED — 3D-CODED: 3D Correspondences by Deep Deformation (ECCV 2018) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Reduced Representation of Deformation Fields for Effective Non-rigid Shape Matching (NeurIPS 2022 · similarity 0.85)
- Shape Non-rigid Kinematics (SNK): A Zero-Shot Method for Non-Rigid Shape Matching via Unsupervised Functional Map Regularized Reconstruction (NeurIPS 2023 · similarity 0.84)
- MINA: Convex Mixed-Integer Programming for Non-Rigid Shape Alignment (CVPR 2020 · similarity 0.84)
- Locality Optimization Refinement with Deformation for Shape Matching via Functional Maps (IJCV 2026 · similarity 0.83)
- Probabilistic Deformation Consistency for Unsupervised Shape Matching (AAAI 2026 · similarity 0.83)
- MINIMA: Modality Invariant Image Matching (CVPR 2025 · similarity 0.81)
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.
