Research map / 3D vision / Face and shape
TAVA: Template-Free Animatable Volumetric Actors
Ruilong Li, Julian Tanke, Minh Vo, Michael Zollhöfer, Jürgen Gall, Angjoo Kanazawa, Christoph Lassner
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 |
|---|---|---|---|
| use forward LBS skinning in canonical space instead of pose-conditioned inverse deformation (TAVA vs NARF) | TAVA · Animal subjects, novel pose OOD | PSNR: 30.92 -> 35.77 | Improved |
| TAVA vs A-NeRF | TAVA · Animal subjects, novel pose OOD | PSNR: 25.66 -> 35.77 | Improved |
| TAVA vs NARF | TAVA · Animal subjects, novel pose OOD | P2P error: 8.46 -> 3.38 | Improved |
| TAVA vs NARF | TAVA · Animal subjects, novel view | PSNR: 36.55 -> 37.30 | Improved |
| TAVA vs NARF | TAVA · ZJU-Mocap, novel view | PSNR: 32.94 -> 33.11 | Improved |
| TAVA vs NARF | TAVA · ZJU-Mocap, novel pose OOD | PSNR: 30.60 -> 30.69 | Improved |
| TAVA vs NeuralBody (template-free vs SMPL template) | TAVA · ZJU-Mocap, novel view | PSNR: 33.91 vs 33.11 | No clear effect |
| TAVA vs NeuralBody on OOD poses | TAVA · ZJU-Mocap, novel pose OOD | PSNR: 30.33 -> 30.69 | Improved |
| shift from in-distribution to out-of-distribution poses | TAVA · Animal subjects | PSNR: 37.45 -> 35.77 | Hurt |
| shift from in-distribution to out-of-distribution poses | TAVA, baselines · ZJU-Mocap | PSNR: 0.67db to 2.66db drop | Hurt |
| replace failed root-finding points' zero density with neighbor interpolation (ablation) | TAVA · supplemental | PSNR: slightly better performance | Improved |
Closest papers by meaning
- Make-It-Animatable: An Efficient Framework for Authoring Animation-Ready 3D Characters (CVPR 2025 · similarity 0.88)
- ARAH: Animatable Volume Rendering of Articulated Human SDFs (ECCV 2022 · similarity 0.87)
- Animatable Gaussians: Learning Pose-dependent Gaussian Maps for High-fidelity Human Avatar Modeling (CVPR 2024 · similarity 0.85)
- 3D²-Actor: Learning Pose-Conditioned 3D-Aware Denoiser for Realistic Gaussian Avatar Modeling (AAAI 2025 · similarity 0.85)
- AniGS: Animatable Gaussian Avatar from a Single Image with Inconsistent Gaussian Reconstruction (CVPR 2025 · similarity 0.85)
- NECA: Neural Customizable Human Avatar (CVPR 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.
