Research map / 3D vision / NeRF and radiance fields
Volumetric Rendering with Baked Quadrature Fields
Gopal Sharma, Daniel Rebain, Kwang Moo Yi, Andrea Tagliasacchi
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 baked quadrature-field mesh instead of MobileNeRF | Shelly dataset rendering · Shelly dataset | avg PSNR: 31.62 -> 35.13 | Improved |
| use quadrature field instead of MobileNeRF on a transparent object | Objaverse glassy object · Objaverse | PSNR: 24.6 -> 30.6 | Improved |
| use quadrature-field mesh instead of Adaptive shells on Mip-NeRF 360 | Mip-NeRF 360 · Mip-NeRF 360 | avg samples per ray: 17.41 -> 5.15 | Improved |
| add fine-tuning to align mesh with NeRF density field (ablation) | Quadrature field rendering · Mic scene, NeRF synthetic | PSNR: 31.05 -> 34.06 | Improved |
| use mesh from quadrature field instead of density field only (ablation) | Quadrature field rendering · Mic scene | PSNR: 30.37 -> 33.40 | Improved |
| increase spherical Gaussian lobes from 3 to 6 (ablation) | Quadrature field rendering · Mic scene | PSNR: 33.68 -> 34.06 | Improved |
| bake features with quantization (ablation) | Quadrature field rendering · Mic scene | PSNR: 34.06 -> 33.70 | Hurt |
| use 2x super-sampling instead of 1x (ablation) | Quadrature field rendering · Mic scene | PSNR and FPS: PSNR 32.05 -> 33.70; FPS 750 -> 210 | Mixed |
| increase mesh size from 880k to 3.3M faces (ablation) | Quadrature field rendering · Mic scene | PSNR and FPS: PSNR 33.70 -> 33.92; FPS 210 -> 180 | Mixed |
| remove continuous rendering loss (ablation) | Quadrature field fine-tuning · Mic scene | PSNR: 34.06 -> 34.00 | Hurt |
| use quadrature mesh instead of Adaptive-shells on NeRF synthetic | NeRF synthetic · NeRF synthetic | avg PSNR: 31.84 -> 31.00 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- MobileNeRF — MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures (CVPR 2023) · better on 4 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
- NeRF Revisited: Fixing Quadrature Instability in Volume Rendering (NeurIPS 2023 · similarity 0.92)
- Dynamic Mesh-Aware Radiance Fields (ICCV 2023 · similarity 0.90)
- NeRFix: fixing subtle mistakes in the quadrature of the NeRF volumetric integral (NeurIPS 2026 · similarity 0.90)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.89)
- GL-NeRF: Gauss-Laguerre Quadrature Enables Training-Free NeRF Acceleration (NeurIPS 2024 · similarity 0.88)
- Compact 3D Gaussian Representation for Radiance Field (CVPR 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.
