Research map / 3D vision / NeRF and radiance fields
NeRFrac: Neural Radiance Fields through Refractive Surface
Yifan Zhan, Shohei Nobuhara, Ko Nishino, Yinqiang Zheng
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 |
|---|---|---|---|
| model refraction explicitly with a Refractive Field and Snell's law instead of a deformation layer | NeRFrac vs LB-NeRF · Synthetic data (primary sine) | PSNR: 22.59 -> 34.38 | Improved |
| explicit refractive ray bending instead of straight-ray sampling | NeRF · Synthetic data (second sine) | PSNR: 25.93 -> 34.98 | Improved |
| NeRFrac vs best baseline on real plant scene | NeRFrac vs Tensorf · Real data (plant) | PSNR: 28.23 -> 28.29 | No clear effect |
| remove learned Refractive Field to render water-free views | NeRFrac vs LB-NeRF with deformation removed · Synthetic water-free restoration (second sine) | PSNR: 16.26 -> 29.38 | Improved |
| replace direct Refractive Field with global depth map network requiring intersection sampling (ablation) | NeRFrac · synthetic and real scenes | render quality and surface shape | Hurt |
Later papers that compared with this method
1 later accepted paper reports NeRFrac as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- RefracGS: Novel View Synthesis Through Refractive Water Surfaces with 3D Gaussian Ray Tracing (ECCV 2026 · better on 5 of 5 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
- Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance Fields (CVPR 2022 · similarity 0.93)
- NeRFReN: Neural Radiance Fields With Reflections (CVPR 2022 · similarity 0.92)
- MS-NeRF: Multi-Space Neural Radiance Fields (TPAMI 2025 · similarity 0.91)
- Multi-Space Neural Radiance Fields (CVPR 2023 · similarity 0.91)
- NeRFLight: Fast and Light Neural Radiance Fields Using a Shared Feature Grid (CVPR 2023 · similarity 0.91)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · 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.
