Research map / 3D vision / Face and shape
Learning Compositional Representation for 4D Captures With Neural ODE
Boyan Jiang, Yinda Zhang, Xingkui Wei, Xiangyang Xue, Yanwei Fu
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 the compositional latent ODE method rather than OFlow | 4D reconstruction · D-FAUST seen individuals | IoU: 79.9% -> 81.8% | Improved |
| Use the compositional latent ODE method rather than OFlow | 4D reconstruction · D-FAUST unseen individual | IoU: 69.6% -> 68.2% | Hurt |
| Use the compositional latent ODE method rather than OFlow | 4D reconstruction · Warping Cars | IoU: 70.0% -> 70.9% | Improved |
| Use the compositional representation for motion transfer | D-FAUST motion transfer · D-FAUST | IoU: 26.7% -> 85.0% | Improved |
| Use the compositional method rather than OFlow | D-FAUST completion · D-FAUST temporal completion | IoU: 85.1% -> 86.4% | Improved |
| Use the compositional method rather than OFlow | D-FAUST completion · D-FAUST spatial completion | IoU: 86.0% -> 87.2% | Improved |
| Use the compositional method rather than OFlow | D-FAUST future prediction · D-FAUST | IoU: 75.5% -> 80.8% | Improved |
| Replace the conditional latent ODE with an MLP pose predictor (ablation) | Proposed 4D reconstruction method · D-FAUST | IoU: 81.8% -> 80.4% | Hurt |
| Replace the conditional latent ODE with an MLP pose predictor (ablation) | Proposed 4D reconstruction method · D-FAUST | Chamfer Distance: 0.068 -> 0.073 | Hurt |
| Add identity exchange training at 50% rather than no exchange (ablation) | Proposed method · D-FAUST motion transfer | IoU: 65.3% -> 85.0% | Improved |
| Add identity exchange training at 50% rather than no exchange (ablation) | Proposed method · D-FAUST 4D reconstruction | IoU: 83.3% -> 81.8% | Hurt |
| Vary the identity exchange rate from 25% to 50% (ablation) | Proposed method · D-FAUST motion transfer | IoU: 84.1% -> 85.0% | Improved |
Closest papers by meaning
- Learning Neural Deformation Representation for 4D Dynamic Shape Generation (ECCV 2024 · similarity 0.87)
- Mesh4D: 4D Mesh Reconstruction and Tracking from Monocular Video (CVPR 2026 · similarity 0.87)
- Motion2VecSets: 4D Latent Vector Set Diffusion for Non-rigid Shape Reconstruction and Tracking (CVPR 2024 · similarity 0.87)
- DNF: Unconditional 4D Generation with Dictionary-based Neural Fields (CVPR 2025 · similarity 0.86)
- 4Deform: Neural Surface Deformation for Robust Shape Interpolation (CVPR 2025 · similarity 0.85)
- Category-Agnostic Neural Object Rigging (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.
