Research map / 3D vision / Face and shape
KillingFusion: Non-Rigid 3D Reconstruction Without Correspondences
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 |
|---|---|---|---|
| use SDF-based registration without explicit correspondence search | KillingFusion · Duck sequence | average reconstruction error (mm): 5.362 mm -> 3.896 mm | Helped |
| use SDF-based registration without explicit correspondence search | KillingFusion · Snoopy sequence | average reconstruction error (mm): 4.205 mm -> 3.543 mm | Helped |
| remove level set property preservation (ablation) | KillingFusion · qualitative reconstruction sequence | visual reconstruction quality | Hurt |
| remove motion regularization (ablation) | KillingFusion · qualitative reconstruction sequence | visual reconstruction quality | Hurt |
| replace the damped Killing regularizer with conventional motion smoothness (ablation) | KillingFusion · qualitative reconstruction sequence | visual reconstruction quality | Hurt |
| set γ = 1, removing damping of the Killing condition (ablation) | KillingFusion · qualitative reconstruction sequence | preservation of non-rigid parts | Hurt |
| use a damped Killing regularizer rather than standard motion smoothness alone (ablation) | KillingFusion · qualitative reconstruction sequence | geometric consistency | Helped |
| use SDF level set evolution instead of mesh-based correspondence estimation | KillingFusion · topological changes and rapid motion | ability to recover changing geometry | Helped |
Closest papers by meaning
- 4Deform: Neural Surface Deformation for Robust Shape Interpolation (CVPR 2025 · similarity 0.86)
- Correspondence-Free Non-Rigid Point Set Registration Using Unsupervised Clustering Analysis (CVPR 2024 · similarity 0.86)
- Small Steps and Level Sets: Fitting Neural Surface Models with Point Guidance (CVPR 2024 · similarity 0.86)
- Learning Bijective Surface Parameterization for Inferring Signed Distance Functions from Sparse Point Clouds with Grid Deformation (CVPR 2025 · similarity 0.85)
- Diffeomorphic Mesh Deformation via Efficient Optimal Transport for Cortical Surface Reconstruction (ICLR 2024 · similarity 0.85)
- Spectral Meets Spatial: Harmonising 3D Shape Matching and Interpolation (CVPR 2024 · 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.