Research map / Point clouds / Shape and surface completion
Dynamic Point Fields
Sergey Prokudin, Qianli Ma, Maxime Raafat, Julien Valentin, Siyu Tang
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 as-isometric-as-possible regularisation L_iso (ablation) | Dynamic Point Fields (unsupervised) · DeformingThings4D | strict accuracy Acc_S: 27.2 -> 59.8 | Improved |
| add as-isometric-as-possible regularisation L_iso (ablation) | Dynamic Point Fields · DeformingThings4D | Chamfer distance: 1.045 -> 0.786 | Improved |
| add as-isometric-as-possible regularisation L_iso (ablation) | Dynamic Point Fields · Owlii | Chamfer distance: 0.111 -> 0.123 | Mixed |
| add as-isometric-as-possible regularisation L_iso (ablation) | Dynamic Point Fields · Lego deformation | Chamfer distance: 0.376 -> 0.166 | Improved |
| add keypoint supervision from Lepard matcher (ablation) | Dynamic Point Fields · DeformingThings4D | strict accuracy Acc_S: 59.8 -> 72.6 | Improved |
| add SMPL-X vertex guidance (ablation) | Dynamic Point Fields · ReSynth | Chamfer distance: 0.309 -> 0.165 | Improved |
| replace implicit SDF/hash-grid canonical surface with explicit point set | static surface reconstruction · Lego excavator static reconstruction | Chamfer distance (x1e-4): 0.361 -> 0.246 | Improved |
| reduce number of points to 10^4 (ablation) | static point reconstruction · Lego excavator static reconstruction | Chamfer distance (x1e-4): 0.246 -> 3.362 | Hurt |
| use point deformation fields instead of LBS-based avatar models | clothed avatar animation · ReSynth user study | user choice %: 85.8% | Improved |
Closest papers by meaning
- AutoAvatar: Autoregressive Neural Fields for Dynamic Avatar Modeling (ECCV 2022 · similarity 0.90)
- NSF: Neural Surface Fields for Human Modeling from Monocular Depth (ICCV 2023 · similarity 0.89)
- Implicit Neural Surface Deformation with Explicit Velocity Fields (ICLR 2025 · similarity 0.89)
- Degrees of Freedom Matter: Inferring Dynamics from Point Trajectories (CVPR 2024 · similarity 0.88)
- 4Deform: Neural Surface Deformation for Robust Shape Interpolation (CVPR 2025 · similarity 0.87)
- Implicit-ARAP: Efficient Handle-Guided Neural Field Deformation via Local Patch Meshing (NeurIPS 2025 · similarity 0.87)
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.
