Research map / 3D vision / NeRF and radiance fields
NeRF Revisited: Fixing Quadrature Instability in Volume Rendering
Read the paperCheck 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 piecewise constant opacity with piecewise linear opacity | NeRF · Blender | PSNR: 30.61 -> 31.10 | Helped |
| replace piecewise constant opacity with piecewise linear opacity | NeRF · LLFF | LPIPS (×10²): 7.37 -> 6.06 | Helped |
| use piecewise linear opacity for volume rendering | Mip-NeRF · Blender | PSNR: 31.76 -> 32.48 | Helped |
| use piecewise linear opacity | Vanilla NeRF · Blender, close-up test at Dist 0.25x | PSNR: 20.18 -> 20.34 | Helped |
| use piecewise linear opacity | Vanilla NeRF with random-distance training cameras · Blender, close-up test at Dist 0.25x | LPIPS (×10²): 45.7 -> 41.1 | Helped |
| use piecewise linear opacity with precise importance sampling for depth supervision | Vanilla NeRF · Blender, random-distance cameras with depth supervision | RMSE: 0.178 -> 0.147 | Helped |
| use piecewise linear opacity | Vanilla NeRF · DTU (15 scenes) | PSNR: 27.96 -> 28.43 | Helped |
| use piecewise linear opacity | Vanilla NeRF · Blender geometry extraction | Chamfer distance (×10³): 10.43 -> 10.10 | Helped |
| use piecewise linear opacity for volume rendering | DIVeR · Blender | PSNR: 30.78 -> 30.88 | Helped |
| use precise importance sampling instead of the surrogate CDF function (ablation) | PL-NeRF · Blender, hemisphere | PSNR: 30.25 -> 30.87 | Helped |
| use 128 coarse and 64 fine samples rather than 64 coarse and 128 fine samples, keeping total samples the same (ablation) | PL-NeRF with precise importance sampling · Blender, hemisphere | PSNR: 30.87 -> 31.10 | Helped |
| reduce both coarse and fine samples to 64 (ablation) | PL-NeRF compared with Vanilla NeRF · Blender | PSNR: 29.86 -> 30.09 | Helped |
Closest papers by meaning
- GL-NeRF: Gauss-Laguerre Quadrature Enables Training-Free NeRF Acceleration (NeurIPS 2024 · similarity 0.92)
- Rad-NeRF: Ray-decoupled Training of Neural Radiance Field (NeurIPS 2024 · similarity 0.90)
- L0-Sampler: An L0 Model Guided Volume Sampling for NeRF (CVPR 2024 · similarity 0.89)
- CVT-xRF: Contrastive In-Voxel Transformer for 3D Consistent Radiance Fields from Sparse Inputs (CVPR 2024 · similarity 0.88)
- PBR-NeRF: Inverse Rendering with Physics-Based Neural Fields (CVPR 2025 · similarity 0.88)
- Ced-NeRF: A Compact and Efficient Method for Dynamic Neural Radiance Fields (AAAI 2024 · similarity 0.88)
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.