Research map / 3D vision / Face and shape
Topology-Preserving Shape Reconstruction and Registration via Neural Diffeomorphic Flow
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 |
|---|---|---|---|
| Replace unconstrained deformation with topology-preserving diffeomorphic flow | NDF · Unseen organ shape reconstruction | CD Mean: Pancreas: 0.63 -> 0.512; Liver: 0.509 -> 0.476; Lung: 0.669 -> 0.643 | Helped |
| Use quasi time-varying rather than stationary velocity fields (ablation) | NDF · Pancreas, unseen shape reconstruction | CD Mean: 1.42 -> 0.512 | Helped |
| Use quasi time-varying rather than time-varying velocity fields (ablation) | NDF · Pancreas, unseen shape reconstruction | CD Mean: 1.65 -> 0.512 | Helped |
| Add curriculum learning to four-step quasi time-varying flow (ablation) | NDF · Pancreas, unseen shape reconstruction | CD Mean: 0.516 -> 0.51 | Helped |
| Add curriculum learning to time-varying flow (ablation) | NDF · Pancreas, unseen shape reconstruction | CD Mean: 1.65 -> 125 | Hurt |
| Increase progressive flow steps from four to eight (ablation) | NDF · Pancreas, unseen shape reconstruction | CD Mean: 0.51 -> 0.51 | No clear effect |
| Remove point-pair smooth regularization (ablation) | NDF · Pancreas, unseen shape reconstruction | CD Mean: 0.51 -> 0.512 | No clear effect |
| Replace DIT deformation with NDF topology-preserving registration | NDF · D-FAUST chicken-wings sequence, test | L2: 0.0121 -> 0.0032 | Helped |
| Use NDF rather than DIT for label transfer | NDF · MMWHS label transfer | IOU: 89.94 -> 92.38 | Helped |
Closest papers by meaning
- Neural Template: Topology-Aware Reconstruction and Disentangled Generation of 3D Meshes (CVPR 2022 · similarity 0.89)
- Fast Symmetric Diffeomorphic Image Registration with Convolutional Neural Networks (CVPR 2020 · similarity 0.89)
- Deep Implicit Templates for 3D Shape Representation (CVPR 2021 · similarity 0.89)
- NODEO: A Neural Ordinary Differential Equation Based Optimization Framework for Deformable Image Registration (CVPR 2022 · similarity 0.88)
- Implicit Modeling of Non-rigid Objects with Cross-Category Signals (AAAI 2024 · similarity 0.87)
- $\texttt{NAISR}$: A 3D Neural Additive Model for Interpretable Shape Representation (ICLR 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.