Research map / 3D vision / Face and shape
MetaAvatar: Learning Animatable Clothed Human Models from Few Depth Images
Shaofei Wang, Marko Mihajlovic, Qianli Ma, Andreas Geiger, Siyu Tang
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 |
|---|---|---|---|
| fine-tune MetaAvatar from 2.5D depth vs baselines trained on full meshes | MetaAvatar · CAPE interpolation subj 00122/00215 | D_p (cm): 0.454 -> 0.450 vs LEAP | Improved |
| MetaAvatar vs LEAP on subjects with unseen action trials | MetaAvatar · CAPE interpolation subj 00134/03375 | D_p (cm): 0.483 (LEAP) vs 0.518 | Hurt |
| add hypernetwork predicting SDF residuals over meta-SDF (vs plain MLP with pose input) (ablation) | MetaAvatar · CAPE interpolation subj 00122/00215 | D_p (cm): 3.278 -> 0.461 | Improved |
| add positional encoding to MLP baseline (ablation) | MLP · CAPE interpolation subj 00122/00215 | D_p (cm): 3.278 -> 1.806 | Improved |
| add hypernetwork (BoneEnc) over meta-learned SIREN conditioned on pose (ablation) | MetaAvatar · CAPE interpolation subj 00122/00215 | D_p (cm): 0.472 -> 0.461 | Improved |
| hierarchical bone encoder instead of unit quaternions (ablation) | MetaAvatar hypernetwork · CAPE interpolation subj 00122/00215 | D_p (cm): 0.460 -> 0.461 | No clear effect |
| hierarchical bone encoder instead of unit quaternions (ablation) | MetaAvatar hypernetwork · CAPE interpolation subj 00134/03375 | D_p (cm): 0.526 -> 0.523 | Improved |
| reduce fine-tuning data from 50% to 10% (ablation) | MetaAvatar · CAPE interpolation subj 00122/00215 | D_p (cm): 0.450 -> 0.512 | Hurt |
| reduce fine-tuning data to <1% (8-20 frames) (ablation) | MetaAvatar · CAPE interpolation subj 00122/00215 | D_p (cm): 0.450 -> 0.592 | Hurt |
| reduce fine-tuning data to <1% on extrapolation perceptual score (ablation) | MetaAvatar · CAPE extrapolation | perceptual score: 0.5 -> 0.510 (00122/00215); 0.5 -> 0.387 (00134/03375) | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- LEAP — LEAP: Learning Articulated Occupancy of People (CVPR 2021) · better on 2 of 3 shared results
- SCANimate — SCANimate: Weakly Supervised Learning of Skinned Clothed Avatar Networks (CVPR 2021) · better on 2 of 2 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
- SCANimate: Weakly Supervised Learning of Skinned Clothed Avatar Networks (CVPR 2021 · similarity 0.91)
- Structured Local Radiance Fields for Human Avatar Modeling (CVPR 2022 · similarity 0.90)
- FreeCloth: Free-form Generation Enhances Challenging Clothed Human Modeling (CVPR 2025 · similarity 0.90)
- Pose Modulated Avatars from Video (ICLR 2024 · similarity 0.90)
- SCULPT: Shape-Conditioned Unpaired Learning of Pose-dependent Clothed and Textured Human Meshes (CVPR 2024 · similarity 0.89)
- 3DGS-Avatar: Animatable Avatars via Deformable 3D Gaussian Splatting (CVPR 2024 · similarity 0.89)
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.
