Research map / 3D vision / Face and shape
SAFT: Shape and Appearance of Fabrics from Template via Differentiable Physical Simulations from Monocular Video
David Stotko, Reinhard Klein
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 energy regularization R_E (ablation) | Ours (SfT optimization) · phi-SfT dataset (mean) | L2 Chamfer distance: 23.59 -> 3.33 | Improved |
| add force regularization R_F (ablation) | Ours (SfT optimization) · phi-SfT dataset (mean) | L2 Chamfer distance: 5.96 -> 3.33 | Improved |
| add both regularizations R_E and R_F (ablation) | Ours (SfT optimization) · phi-SfT dataset (mean) | L2 Chamfer distance: 22.51 -> 3.33 | Improved |
| add silhouette loss (ablation) | Ours (SfT optimization) · phi-SfT dataset (mean) | L2 Chamfer distance: 3.96 -> 3.33 | No clear effect |
| add energy regularization R_E (ablation) | Ours · scene R5 | L2 Chamfer distance: 128.15 -> 7.12 | Improved |
| optimize physical stiffness parameters instead of fixing them (ablation) | Ours · phi-SfT dataset (mean) | L2 Chamfer distance: 4.73 -> 3.33 | Improved |
| fix physical parameters (with regularizations removed) (ablation) | Ours without R_E and R_F · phi-SfT dataset (mean) | L2 Chamfer distance: 22.51 -> 5.15 | Improved |
| add R_E when physical parameters are fixed (ablation) | Ours with fixed parameters · phi-SfT dataset (mean) | L2 Chamfer distance: 5.15 -> 6.37 | Hurt |
| use mass-spring physics SfT with regularizations instead of phi-SfT | Ours vs phi-SfT · phi-SfT dataset (mean) | L2 Chamfer distance: 8.80 -> 3.33 | Improved |
| use Ours instead of PG-SfT | Ours vs PG-SfT · phi-SfT dataset (mean) | L2 Chamfer distance: 9.71 -> 3.33 | Improved |
| use Ours instead of phi-SfT | Ours vs phi-SfT · phi-SfT dataset (mean) | geometry runtime (min): 1105 -> 29.9 | Improved |
| use Ours instead of PG-SfT | Ours vs PG-SfT · phi-SfT dataset (mean) | geometry runtime (min): 2.00 -> 29.9 | Hurt |
Closest papers by meaning
- Physics-guided Shape-from-Template: Monocular Video Perception through Neural Surrogate Models (CVPR 2024 · similarity 0.89)
- MonoPhysics: Estimating Geometry, Appearance, and Physical Parameters from Monocular Videos (NeurIPS 2026 · similarity 0.86)
- Vid2Sim: Generalizable, Video-based Reconstruction of Appearance, Geometry and Physics for Mesh-free Simulation (CVPR 2025 · similarity 0.86)
- MorpheuS: Neural Dynamic 360deg Surface Reconstruction from Monocular RGB-D Video (CVPR 2024 · similarity 0.85)
- Link to the Past: Temporal Propagation for Fast 3D Human Reconstruction from Monocular Video (CVPR 2025 · similarity 0.85)
- Thin-Shell-SfT: Fine-Grained Monocular Non-rigid 3D Surface Tracking with Neural Deformation Fields (CVPR 2025 · 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.
