Research map / 3D vision / NeRF and radiance fields
Template-free Articulated Neural Point Clouds for Reposable View Synthesis
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| random initialization instead of TiNeuVox-pretrained features (ablation) | Articulated point-cloud NeRF · Robots | PSNR: 30.19 -> 29.97 | No clear effect |
| skip fine-tuning of feature points (ablation) | Articulated point-cloud NeRF · Blender | PSNR: 29.10 -> 29.16 | No clear effect |
| condition feature network on learned skeletal poses (Ours-pose) | Articulated point-cloud NeRF · ZJU-MoCap | PSNR: 29.60 -> 32.05 | Helped |
| replace learned skeleton with ground-truth SMPL skeleton | Ours-pose · ZJU-MoCap | PSNR: 32.05 -> 32.01 | No clear effect |
| use our method instead of WIM at 10h training | Articulated point-cloud NeRF vs WIM · Robots | PSNR: 27.61 -> 30.19 | Helped |
| use our method instead of WIM at 2h training | Articulated point-cloud NeRF vs WIM · Robots | PSNR: 25.05 -> 28.81 | Helped |
| use our method instead of WIM | Articulated point-cloud NeRF vs WIM · Blender | PSNR: 23.81 -> 29.10 | Helped |
| use base Ours instead of WIM | Articulated point-cloud NeRF vs WIM · ZJU-MoCap | PSNR: 31.08 -> 29.60 | Hurt |
| only 100 iterations of TiNeuVox pre-training (ablation) | Articulated point-cloud NeRF · Blender Jumping Jacks | PSNR: over 30 dB | No clear effect |
| add L_tranf regularization (ablation) | Articulated point-cloud NeRF · Atlas (Robots) | removed components after simplification: 49 -> 62 | Helped |
| add regularization terms (ARAP, smooth, skel, tranf) (ablation) | Articulated point-cloud NeRF · Robots | image quality | No clear effect |
| add L_ARAP (ablation) | Articulated point-cloud NeRF · Robots | rigidity of deformation | Helped |
Closest papers by meaning
- REACTO: Reconstructing Articulated Objects from a Single Video (CVPR 2024 · similarity 0.89)
- Joint Optimization of Neural Radiance Fields and Continuous Camera Motion from a Monocular Video (CVPR 2025 · similarity 0.89)
- 4D3R: Motion-Aware Neural Reconstruction and Rendering of Dynamic Scenes from Monocular Videos (NeurIPS 2025 · similarity 0.88)
- TFS-NeRF: Template-Free NeRF for Semantic 3D Reconstruction of Dynamic Scene (NeurIPS 2024 · similarity 0.88)
- ICON: Incremental CONfidence for Joint Pose and Radiance Field Optimization (CVPR 2024 · similarity 0.88)
- CF-NeRF: Camera Parameter Free Neural Radiance Fields with Incremental Learning (AAAI 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.