Research map / 3D vision / Face and shape
DeepMetaHandles: Learning Deformation Meta-Handles of 3D Meshes With Biharmonic Coordinates
Minghua Liu, Minhyuk Sung, Radomir Mech, Hao Su
Check 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 |
|---|---|---|---|
| Use DeepMetaHandles instead of Neural Cages for deformation generation | ShapeNet chairs · Chair | COV: 54.4% -> 64.6% | Improved |
| Use DeepMetaHandles instead of Neural Cages for deformation generation | ShapeNet cars · Car | COV: 66.6% -> 76.5% | Improved |
| Use DeepMetaHandles instead of Neural Cages for deformation generation | ShapeNet tables · Table | COV: 44.7% -> 54.9% | Improved |
| Use DeepMetaHandles instead of Neural Cages for deformation generation | ShapeNet chairs · Chair | MMD (×100): 4.23 -> 4.28 | Hurt |
| Use DeepMetaHandles instead of Neural Cages for deformation generation | ShapeNet cars · Car | MMD (×100): 2.65 -> 2.97 | Hurt |
| Use DeepMetaHandles instead of Neural Cages for deformation generation | ShapeNet tables · Table | MMD (×100): 3.85 -> 3.70 | Improved |
| Add learned meta-handles rather than deform directly with control-point handles (ablation) | DeepMetaHandles with adversarial loss · Chair target-driven deformation | Cotangent Laplacian (×10): 7.69 -> 5.75 | Improved |
| Add adversarial regularization (ablation) | DeepMetaHandles · Chair target-driven deformation | Cotangent Laplacian (×10): 8.61 -> 5.75 | Improved |
| Add adversarial regularization (ablation) | DeepMetaHandles · Chair target-driven deformation | Chamfer distance (×100): 5.76 -> 6.28 | Hurt |
| Add adversarial loss (ablation) | Chair deformation generation · Chair | COV: 56.3% -> 64.6% | Improved |
| Add disentanglement regularization (ablation) | DeepMetaHandles · Chair deformation generation | COV: 64.1% -> 64.6% | No clear effect |
| Increase control points from 50 to 100 (ablation) | DeepMetaHandles · Chair target-driven deformation | Chamfer distance: 0.0628 -> 0.0715 | Hurt |
Later papers that compared with this method
1 later accepted paper reports DeepMetaHandles as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Few-Shot Physically-Aware Articulated Mesh Generation via Hierarchical Deformation (ICCV 2023 · better on 5 of 5 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Neural Shape Deformation Priors (NeurIPS 2022 · similarity 0.89)
- 3D-CODED: 3D Correspondences by Deep Deformation (ECCV 2018 · similarity 0.88)
- Neural Template: Topology-Aware Reconstruction and Disentangled Generation of 3D Meshes (CVPR 2022 · similarity 0.88)
- Variational Autoencoders for Deforming 3D Mesh Models (CVPR 2018 · similarity 0.87)
- Deep Feature Deformation Weights (CVPR 2026 · similarity 0.87)
- Implicit-ARAP: Efficient Handle-Guided Neural Field Deformation via Local Patch Meshing (NeurIPS 2025 · 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.
