Research map / 3D vision / NeRF and radiance fields
Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance Fields
Check 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 |
|---|---|---|---|
| Replace view-direction input with reflected-direction parameterization (ablation) | Ref-NeRF · Shiny Blender | PSNR: 29.47 -> 35.96 | Helped |
| Add the normal orientation loss (ablation) | Ref-NeRF · Shiny Blender | MAE°: 52.56 -> 18.38 | Helped |
| Use predicted normals to compute reflection directions (ablation) | Ref-NeRF · Shiny Blender | PSNR: 30.91 -> 35.96 | Helped |
| Do not append view direction to the directional MLP input (ablation) | Ref-NeRF · Shiny Blender | PSNR: 35.42 -> 35.96 | Helped |
| Include the normal-view dot product as an input instead of fixing the lobe (ablation) | Ref-NeRF · Shiny Blender | PSNR: 35.52 -> 35.96 | Helped |
| Add diffuse color as an explicit component (ablation) | Ref-NeRF · Shiny Blender | PSNR: 33.32 -> 35.96 | Helped |
| Add specular tint as an explicit component (ablation) | Ref-NeRF · Shiny Blender | PSNR: 35.45 -> 35.96 | Helped |
| Model material roughness explicitly (ablation) | Ref-NeRF · Shiny Blender | PSNR: 33.39 -> 35.96 | Helped |
| Use integrated directional encoding instead of positional encoding (ablation) | Ref-NeRF · Shiny Blender | PSNR: 35.90 -> 35.96 | Helped |
| Use Ref-NeRF rather than Mip-NeRF | novel-view rendering · Blender | PSNR: 33.09 -> 33.99 | Helped |
| Use Ref-NeRF rather than Mip-NeRF | normal estimation · Blender | MAE°: 38.30 -> 23.22 | Helped |
Closest papers by meaning
- NeRFReN: Neural Radiance Fields With Reflections (CVPR 2022 · similarity 0.94)
- RegNeRF: Regularizing Neural Radiance Fields for View Synthesis From Sparse Inputs (CVPR 2022 · similarity 0.93)
- NeRF On-the-go: Exploiting Uncertainty for Distractor-free NeRFs in the Wild (CVPR 2024 · similarity 0.92)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.91)
- InsertNeRF: Instilling Generalizability into NeRF with HyperNet Modules (ICLR 2024 · similarity 0.91)
- NeRF-HuGS: Improved Neural Radiance Fields in Non-static Scenes Using Heuristics-Guided Segmentation (CVPR 2024 · similarity 0.91)
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.