Research map / 3D vision / Face and shape
ARAPReg: An As-Rigid-As Possible Regularization Loss for Learning Deformable Shape Generators
Qixing Huang, Xiangru Huang, Bo Sun, Zaiwei Zhang, Junfeng Jiang, Chandrajit Bajaj
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 ARAPReg to the auto-decoder (ablation) | Ours-AD · DFAUST | reconstruction error (mm): 5.17 -> 4.52 | Improved |
| add ARAPReg to the auto-decoder (ablation) | Ours-AD · SMAL | reconstruction error (mm): 8.74 -> 6.68 | Improved |
| add ARAPReg to the auto-decoder (ablation) | Ours-AD · Bone | reconstruction error (mm): 3.91 -> 3.76 | Improved |
| add ARAPReg to the VAE (ablation) | Ours-VAE · DFAUST | reconstruction error (mm): 5.45 -> 4.87 | Improved |
| add ARAPReg to the VAE (ablation) | Ours-VAE · SMAL | reconstruction error (mm): 9.11 -> 7.82 | Improved |
| add ARAPReg to the VAE (ablation) | Ours-VAE · Bone | reconstruction error (mm): 4.09 -> 3.85 | Improved |
| add ARAPReg to the 3DMM architecture (ablation) | 3DMM · DFAUST | reconstruction error (mm): 7.39 -> 6.72 | Improved |
| add ARAPReg to the COMA architecture (ablation) | COMA · DFAUST | reconstruction error (mm): 8.80 -> 4.87 | Improved |
| add ARAPReg to the MeshConv architecture (ablation) | MeshConv · DFAUST | reconstruction error (mm): 5.43 -> 5.02 | Improved |
| add shape and pose decoupling (pose directions penalized more) (ablation) | Ours-AD · DFAUST | reconstruction error (mm): 4.90 -> 4.52 | Improved |
| add shape and pose decoupling (pose directions penalized more) (ablation) | Ours-AD · SMAL | reconstruction error (mm): 7.23 -> 6.68 | Improved |
| add shape and pose decoupling (pose directions penalized more) (ablation) | Ours-AD · Bone | reconstruction error (mm): 3.82 -> 3.76 | Improved |
Closest papers by meaning
- Deforming Autoencoders: Unsupervised Disentangling of Shape and Appearance (ECCV 2018 · similarity 0.86)
- Unsupervised Learning of Dense Shape Correspondence (CVPR 2019 · similarity 0.86)
- Unsupervised Dense Deformation Embedding Network for Template-Free Shape Correspondence (ICCV 2021 · similarity 0.86)
- DiLO: Disentangled Latent Optimization for Learning Shape and Deformation in Grouped Deforming 3D Objects (AAAI 2026 · similarity 0.86)
- Enhancing Implicit Shape Generators Using Topological Regularizations (ICML 2024 · similarity 0.86)
- Neural Pose Representation Learning for Generating and Transferring Non-Rigid Object Poses (NeurIPS 2024 · 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.
