Research map / 3D vision / Face and shape
MorphGS: Morphology-Adaptive Articulated Motion Transfer from Videos
Taeyeon Kim, Youngju Na, Jumin Lee, Sebin Lee, Minhyuk Sung, Sung-eui Yoon
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 |
|---|---|---|---|
| anchor canonical vertices to the skeleton with local offsets instead of free global vertices (ablation) | MorphGS · Mixamo-Humanoids | PMD: 8.45 -> 1.91 | Improved |
| anchor canonical vertices to the skeleton with local offsets instead of free global vertices (ablation) | MorphGS · Mixamo-Humanoids | FVMD: 17.54 -> 8.82 | Improved |
| add morphology adaptation (bone lengths, scale, offsets) (ablation) | MorphGS · Mixamo-Humanoids | PMD: 3.72 -> 1.91 | Improved |
| add dense 2D-3D keypoint supervision to rendering losses (ablation) | MorphGS · Mixamo-Humanoids | PMD: 4.52 -> 2.96 | Improved |
| add synthesized-view guidance (4 views) (ablation) | MorphGS · Mixamo-Humanoids | PMD: 2.96 -> 1.91 | Improved |
| add synthesized-view guidance (4 views) (ablation) | MorphGS · Mixamo-Humanoids | FVMD: 11.93 -> 8.82 | Improved |
| increase synthesized views from 4 to 8 (ablation) | MorphGS · Mixamo-Humanoids | PMD: 1.91 -> 1.85 | Improved |
| use Geo-Aware features instead of DINOv2 for correspondences (ablation) | MorphGS · Mixamo-Humanoids | PMD: 2.90 -> 1.91 | Improved |
| use Geo-Aware features instead of DINOv3 for correspondences (ablation) | MorphGS · Mixamo-Humanoids | PMD: 2.19 -> 1.91 | Improved |
| reduce retained keypoint correspondences from 100% to 10% (ablation) | MorphGS · Mixamo-Humanoids | PMD: 1.91 -> 1.96 | No clear effect |
| use time-conditioned pose network instead of per-frame pose optimization (ablation) | MorphGS · Mixamo-Humanoids | PMD: 2.60 -> 1.91 | Improved |
| use target-driven morphology optimization instead of Pinocchio rigging | MorphGS · Mannequin to Ortiz skeleton | MPJPE: 0.1013 -> 0.0736 | Improved |
Closest papers by meaning
- RigGS: Rigging of 3D Gaussians for Modeling Articulated Objects in Videos (CVPR 2025 · similarity 0.87)
- S3O: A Dual-Phase Approach for Reconstructing Dynamic Shape and Skeleton of Articulated Objects from Single Monocular Video (ICML 2024 · similarity 0.87)
- Learning Implicit Representation for Reconstructing Articulated Objects (ICLR 2024 · similarity 0.87)
- Articulation in Prime: Primitive-Based Articulated Object Understanding from a Single Casual Video (NeurIPS 2026 · similarity 0.86)
- ArticulatedGS: Self-supervised Digital Twin Modeling of Articulated Objects using 3D Gaussian Splatting (CVPR 2025 · similarity 0.86)
- REACTO: Reconstructing Articulated Objects from a Single Video (CVPR 2024 · 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.
