Research map / 3D vision / NeRF and radiance fields
High-Fidelity and Transferable NeRF Editing by Frequency Decomposition
Yisheng He, Weihao Yuan, Siyu Zhu, Zilong Dong, Liefeng Bo, Qixing Huang
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 frequency decomposition with high-frequency blending (ablation) | Freditor · NeRF editing | Sharpness: 193.13 -> 255.03 | Improved |
| add frequency decomposition with high-frequency blending (ablation) | Freditor · NeRF editing | BRISQUE: 20.01 -> 17.15 | Improved |
| add frequency decomposition with high-frequency blending (ablation) | Freditor · NeRF editing, long-term | Long-RMSE: 0.069 -> 0.062 | Improved |
| blend high-frequency details from original scene back into edited low-frequency result (ablation) | Freditor · NeRF editing, short-term | LPIPS: 0.086 -> 0.092 | Hurt |
| use Freditor instead of Instruct-NeRF2NeRF for per-scene editing | NeRF editing · NeRF editing, image quality | Sharpness: 40.61 -> 255.03 | Improved |
| use Freditor instead of Instruct-NeRF2NeRF for per-scene editing | NeRF editing · NeRF editing, image quality | BRISQUE: 38.99 -> 17.15 | Improved |
| use Freditor instead of Instruct-NeRF2NeRF | NeRF editing · NeRF editing, long-term | LPIPS: 0.414 -> 0.388 | Improved |
| lift 2D edits per view independently (Instruct-Pix2Pix) instead of 3D fusion | NeRF editing · NeRF editing, long-term | LPIPS: 0.414 -> 0.588 | Hurt |
| set low-frequency filter downsampling below 1/10 resolution (ablation) | Freditor · CLIP score and BRISQUE sweep | editing effectiveness | Hurt |
| set low-frequency filter resolution above 1/4 (ablation) | Freditor · CLIP score and BRISQUE sweep | artifacts | Hurt |
Closest papers by meaning
- ED-NeRF: Efficient Text-Guided Editing of 3D Scene With Latent Space NeRF (ICLR 2024 · similarity 0.87)
- NeRFDeformer: NeRF Transformation from a Single View via 3D Scene Flows (CVPR 2024 · similarity 0.86)
- FDS: Frequency-Aware Denoising Score for Text-Guided Latent Diffusion Image Editing (CVPR 2025 · similarity 0.85)
- EditBridge: Towards Faithful and Efficient Ultra-High-Resolution Image Editing (NeurIPS 2026 · similarity 0.85)
- DynVideo-E: Harnessing Dynamic NeRF for Large-Scale Motion- and View-Change Human-Centric Video Editing (CVPR 2024 · similarity 0.85)
- LuSh-NeRF: Lighting up and Sharpening NeRFs for Low-light Scenes (NeurIPS 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.
