Research map / 3D vision / NeRF and radiance fields
Editing Conditional Radiance Fields
Steven Liu, Xiuming Zhang, Zhoutong Zhang, Richard Zhang, Jun-Yan Zhu, Bryan Russell
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 per-instance latent codes (ablation) | Single NeRF · PhotoShapes view reconstruction | PSNR: 17.81 -> 36.50 | Improved |
| add separate shape and color codes (ablation) | NeRF with learned latent codes · Aubry chairs view reconstruction | PSNR: 20.93 -> 21.54 | Improved |
| add shared/instance shape network (ablation) | NeRF with separate codes · PhotoShapes view reconstruction | PSNR: 36.88 -> 37.67 | Improved |
| add shared/instance shape network (ablation) | NeRF with separate codes · Aubry chairs view reconstruction | LPIPS: 0.153 -> 0.141 | Improved |
| add shared shape branch when training on all instances (ablation) | conditional radiance field · PhotoShape (all 626 instances) | PSNR: 33.91 -> 35.79 | Improved |
| conditional model versus per-instance NeRFs | conditional radiance field · Aubry chairs view reconstruction | PSNR: 24.15 -> 21.78 | Hurt |
| update later layers (hybrid) instead of only the color code (ablation) | color editing · PhotoShapes color editing | PSNR: 26.29 -> 35.25 | Improved |
| update later layers (hybrid) instead of all weights (ablation) | color editing · PhotoShapes color editing | PSNR: 31.00 -> 35.25 | Improved |
| edit conditional field instead of single-instance NeRF | color editing · PhotoShapes color editing | PSNR: 29.53 -> 35.25 | Improved |
| update F_fuse and F_dens instead of all weights (ablation) | shape editing · Aubry chairs shape editing | PSNR: 20.31 -> 24.57 | Improved |
| update F_fuse and F_dens instead of only F_dens (ablation) | shape editing · Aubry chairs shape editing | PSNR: 21.84 -> 24.57 | Improved |
| update F_fuse and F_dens instead of all weights (time) (ablation) | shape editing · Aubry chairs shape editing | edit time (s): 66.4 -> 37.4 | Improved |
Closest papers by meaning
- NeRF-Editing: Geometry Editing of Neural Radiance Fields (CVPR 2022 · similarity 0.92)
- KaRF: Weakly-Supervised Kolmogorov-Arnold Networks-based Radiance Fields for Local Color Editing (NeurIPS 2025 · similarity 0.92)
- ViCA-NeRF: View-Consistency-Aware 3D Editing of Neural Radiance Fields (NeurIPS 2023 · similarity 0.91)
- IReNe: Instant Recoloring of Neural Radiance Fields (CVPR 2024 · similarity 0.91)
- LAENeRF: Local Appearance Editing for Neural Radiance Fields (CVPR 2024 · similarity 0.90)
- Language-driven Object Fusion into Neural Radiance Fields with Pose-Conditioned Dataset Updates (CVPR 2024 · similarity 0.90)
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.
