Research map / 3D vision / NeRF and radiance fields
Volume Feature Rendering for Fast Neural Radiance Field Reconstruction
Kang Han, Wei Xiang, Lu Yu
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 |
|---|---|---|---|
| use GELU instead of ReLU (ablation) | Instant-NGP · NeRF synthetic | PSNR: 33.11 -> 33.43 | Improved |
| use GELU instead of ReLU (ablation) | VFR · NeRF synthetic | PSNR: 33.17 -> 33.61 | Improved |
| use GELU instead of ReLU (ablation) | VFR · Real-world 360 | PSNR: 28.20 -> 28.26 | Improved |
| use GELU instead of ReLU (ablation) | Instant-NGP · Real-world 360 | PSNR: 28.39 -> 28.43 | No clear effect |
| enlarge MLP from 4x64 to 4x128 (ablation) | Instant-NGP · NeRF synthetic | PSNR: 33.43 -> 33.92 | Improved |
| enlarge MLP from 4x64 to 4x128 (training time) (ablation) | Instant-NGP · NeRF synthetic | training time: 4.4 -> 5.6 | Hurt |
| enlarge VFR MLP hidden neurons 64 -> 128 -> 256 (ablation) | VFR · NeRF synthetic | PSNR: 34.05 -> 34.33 -> 34.62 | Improved |
| enlarge VFR MLP hidden neurons 64 -> 128 -> 256 (ablation) | VFR · Real-world 360 | PSNR: 29.06 -> 29.28 -> 29.48 | Improved |
| replace SH encoding with SH feature encoding (ablation) | VFR · NeRF synthetic | PSNR: 34.46 -> 34.62 | Improved |
| replace SH encoding with SH feature encoding (ablation) | VFR · Real-world 360 | PSNR: 28.91 -> 29.48 | Improved |
| use VFR single network evaluation instead of standard volume rendering (ablation) | Instant-NGP · NeRF synthetic | training time: 4.4 -> 3.0 | Improved |
| use VFR instead of standard rendering with same small MLP (ablation) | Instant-NGP GELU 4x64 · Real-world 360 | PSNR: 28.43 -> 28.26 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 on most shared results.
- Zip-NeRF — Zip-NeRF: Anti-Aliased Grid-Based Neural Radiance Fields (ICCV 2023) · better on 1 of 2 shared results
- Mip-NeRF — Mip-NeRF: A Multiscale Representation for Anti-Aliasing Neural Radiance Fields (ICCV 2021) · better on 2 of 2 shared results
- Mip-NeRF 360 — Mip-NeRF 360: Unbounded Anti-Aliased Neural Radiance Fields (CVPR 2022) · better on 1 of 1 shared result
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
- NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis (ECCV 2020 · similarity 0.91)
- NeRFLight: Fast and Light Neural Radiance Fields Using a Shared Feature Grid (CVPR 2023 · similarity 0.91)
- Learning Robust Generalizable Radiance Field with Visibility and Feature Augmented Point Representation (ICLR 2024 · similarity 0.90)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.89)
- Neural NeRF Compression (ICML 2024 · similarity 0.89)
- NeRF-VPT: Learning Novel View Representations with Neural Radiance Fields via View Prompt Tuning (AAAI 2024 · similarity 0.89)
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.
