Research map / 3D vision / Face and shape
MoDA: Modeling Deformable 3D Objects from Casual Videos
Chaoyue Song, Jiacheng Wei, Tianyi Chen, Yiwen Chen, Chuan-Sheng Foo, Fayao Liu, Guosheng Lin
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 neural dual quaternion blend skinning instead of LBS (ablation) | MoDA · AMA-Swing (multiple videos) | Chamfer distance (cm): 7.6 -> 7.0 | Improved |
| add NeuDBS instead of LBS (ablation) | MoDA · AMA-Samba (multiple videos) | Chamfer distance (cm): 6.7 -> 5.6 | Improved |
| add NeuDBS instead of displacement field (ablation) | MoDA · AMA-Swing (multiple videos) | Chamfer distance (cm): 11.3 -> 7.0 | Improved |
| add NeuDBS instead of SE(3) field (ablation) | MoDA · AMA-Swing (multiple videos) | Chamfer distance (cm): 10.4 -> 7.0 | Improved |
| learn quaternions directly, with direct quaternion blending instead of LBS (ablation) | MoDA · AMA-Swing (multiple videos) | Chamfer distance (cm): 7.6 -> 7.5 | No clear effect |
| use BANMo-style rotation parametrization with direct quaternion blending instead of LBS (ablation) | MoDA · AMA-Swing (multiple videos) | Chamfer distance (cm): 7.6 -> 7.8 | Hurt |
| add NeuDBS instead of RAC-style DBS (ablation) | MoDA · AMA-Swing (multiple videos) | Chamfer distance (cm): 7.1 -> 7.0 | Improved |
| add optimal transport 2D-3D matching instead of soft-argmax (ablation) | MoDA · AMA-Swing (multiple videos) | Chamfer distance (cm): 7.2 -> 7.0 | Improved |
| add optimal transport 2D-3D matching instead of soft-argmax (ablation) | MoDA · AMA-Samba (multiple videos) | F-score@2% (%): 73.1 -> 75.5 | Improved |
| add optimal transport 2D-3D matching (ablation) | MoDA · casual-cat (qualitative) | mask refinement | Improved |
| add texture filtering (ablation) | MoDA and BANMo · casual-adult / casual-human (qualitative) | texture noise | Improved |
| train on a single video instead of multiple videos (ablation) | MoDA · AMA-Swing | Chamfer distance (cm): 7.0 -> 8.5 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- BANMo — BANMo: Building Animatable 3D Neural Models From Many Casual Videos (CVPR 2022) · better on 16 of 16 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
- REACTO: Reconstructing Articulated Objects from a Single Video (CVPR 2024 · similarity 0.89)
- DreamMesh4D: Video-to-4D Generation with Sparse-Controlled Gaussian-Mesh Hybrid Representation (NeurIPS 2024 · similarity 0.87)
- Thin-Shell-SfT: Fine-Grained Monocular Non-rigid 3D Surface Tracking with Neural Deformation Fields (CVPR 2025 · similarity 0.87)
- MorpheuS: Neural Dynamic 360deg Surface Reconstruction from Monocular RGB-D Video (CVPR 2024 · similarity 0.87)
- DeformMaster: An Interactive Physics-Neural World Model for Deformable Objects from Videos (NeurIPS 2026 · similarity 0.87)
- TFS-NeRF: Template-Free NeRF for Semantic 3D Reconstruction of Dynamic Scene (NeurIPS 2024 · similarity 0.87)
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.
