Research map / 3D vision / Face and shape
Deformable Surface Tracking by Graph Matching
Tao Wang, Haibin Ling, Congyan Lang, Songhe Feng, Xiaohui Hou
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 |
|---|---|---|---|
| increase keypoint number N from 1000 to 2000 (ablation) | graph-matching tracker · DeSurT average | tracking error (mm): 40.55 -> 31.15 | Improved |
| increase keypoint number N from 1000 to 2000 (ablation) | graph-matching tracker · TSO average | tracking error (mm): 5.62 -> 4.44 | Improved |
| increase keypoint number N from 1000 to 2000 (ablation) | graph-matching tracker · DeSurT average | computational time (s): 1.32 -> 5.24 | Hurt |
| increase keypoint number N from 1000 to 2000 (ablation) | graph-matching tracker · TSO average | computational time (s): 1.25 -> 4.17 | Hurt |
| increase keypoint number N beyond 1500 (ablation) | graph-matching tracker · DeSurT | tracking error | No clear effect |
| fuse structure information with alpha between 0.3 and 0.8 instead of alpha=0 (ablation) | graph-matching tracker · DeSurT | tracking error | Improved |
| fuse structure information (alpha) on occluded surfaces (ablation) | graph-matching tracker · TSO | tracking error | No clear effect |
| use graph matching approach (N=1000) instead of FSD | graph-matching tracker · DeSurT average | tracking error (mm): 93.43 -> 40.55 | Improved |
| use graph matching approach (N=1000) instead of DIR on occluded surfaces | graph-matching tracker · TSO average | tracking error (mm): 4.85 -> 5.62 | Hurt |
| use graph matching approach (N=2000) instead of DIR on occluded surfaces | graph-matching tracker · TSO average | tracking error (mm): 4.85 -> 4.44 | Improved |
| use graph matching approach (N=1000) instead of DIR | graph-matching tracker · DeSurT average | computational time (s): 16.72 -> 1.32 | Improved |
Closest papers by meaning
- Textureless Deformable Object Tracking With Invisible Markers (TPAMI 2024 · similarity 0.88)
- Higher-Order Graph Principles towards Non-Rigid Surface Registration (TPAMI 2016 · similarity 0.87)
- Neural Deformation Graphs for Globally-Consistent Non-Rigid Reconstruction (CVPR 2021 · similarity 0.85)
- Iterative Corresponding Geometry: Fusing Region and Depth for Highly Efficient 3D Tracking of Textureless Objects (CVPR 2022 · similarity 0.85)
- Thin-Shell-SfT: Fine-Grained Monocular Non-rigid 3D Surface Tracking with Neural Deformation Fields (CVPR 2025 · similarity 0.84)
- Spectral Meets Spatial: Harmonising 3D Shape Matching and Interpolation (CVPR 2024 · similarity 0.84)
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.
