Research map / 3D vision / NeRF and radiance fields
Animatable Neural Radiance Fields for Modeling Dynamic Human Bodies
Sida Peng, Junting Dong, Qianqian Wang, Shangzhan Zhang, Qing Shuai, Xiaowei Zhou, Hujun Bao
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 learned residual to SMPL blend weight field (neural blend weight field) (ablation) | Animatable NeRF · H36M S9 novel pose | PSNR: 21.65 -> 23.72 | Improved |
| add learned residual to SMPL blend weight field (neural blend weight field) (ablation) | Animatable NeRF · H36M S9 novel pose | SSIM: 0.850 -> 0.886 | Improved |
| use accurate marker-based poses instead of marker-less estimates (ablation) | Animatable NeRF · H36M S9 novel pose | PSNR: 22.27 -> 23.72 | Improved |
| train on 100 frames instead of 1 frame (ablation) | Animatable NeRF · H36M S9 | PSNR: 20.29 -> 23.40 | Improved |
| increase training frames from 200 to 800 (ablation) | Animatable NeRF · H36M S9 | PSNR: 23.69 -> 23.16 | Hurt |
| increase training views from 1 to 3 (ablation) | Animatable NeRF · H36M S9 | PSNR: 23.81 -> 23.72 | No clear effect |
| increase training views from 1 to 3 (ablation) | Animatable NeRF · H36M S9 | SSIM: 0.877 -> 0.886 | Improved |
| use Animatable NeRF instead of Neural Textures and NHR | Animatable NeRF · H36M novel view | PSNR: 20.93 -> 23.00 | Improved |
| use Animatable NeRF instead of Neural Body | Animatable NeRF · ZJU-MoCap novel view | PSNR: 28.90 -> 27.10 | Hurt |
| use Animatable NeRF instead of Neural Body | Animatable NeRF · ZJU-MoCap novel pose | PSNR: 23.06 -> 23.16 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 on most shared results.
- Neural Body — Neural Body: Implicit Neural Representations With Structured Latent Codes for Novel View Synthesis of Dynamic Humans (CVPR 2021) · better on 2 of 4 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
- Animatable Implicit Neural Representations for Creating Realistic Avatars From Videos (TPAMI 2024 · similarity 0.97)
- NDF: Neural Deformable Fields for Dynamic Human Modelling (ECCV 2022 · similarity 0.91)
- BANMo: Building Animatable 3D Neural Models From Many Casual Videos (CVPR 2022 · similarity 0.90)
- H-NeRF: Neural Radiance Fields for Rendering and Temporal Reconstruction of Humans in Motion (NeurIPS 2021 · similarity 0.90)
- REACTO: Reconstructing Articulated Objects from a Single Video (CVPR 2024 · similarity 0.89)
- Pose Modulated Avatars from Video (ICLR 2024 · similarity 0.88)
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.
