Research map / 3D vision / NeRF and radiance fields
A View-consistent Sampling Method for Regularized Training of Neural Radiance Fields
Aoxiang Fan, Corentin Dumery, Nicolas Talabot, Pascal Fua
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 |
|---|---|---|---|
| add view-consistent sampling and depth-pushing loss | Nerfacto · Mip-NeRF360 (9 scenes, 10-110 views averaged) | PSNR: 19.05 -> 21.40 | Improved |
| add view-consistent sampling and depth-pushing loss | Nerfacto · Tanks&Temples Intermediate (8 scenes, 10-110 views averaged) | PSNR: 18.29 -> 19.45 | Improved |
| add view-consistent sampling only (ablation) | Nerfacto · Mip-NeRF360 (50 views ablation) | PSNR: 18.88 -> 19.34 | Improved |
| add depth-pushing loss only (ablation) | Nerfacto · Mip-NeRF360 (50 views ablation) | PSNR: 18.88 -> 19.10 | Improved |
| add depth-pushing loss to view-consistent sampling (ablation) | VS-NeRF · Mip-NeRF360 (50 views ablation) | PSNR: 19.34 -> 21.57 | Improved |
| use distilled features in addition to color features (ablation) | VS-NeRF · Mip-NeRF360 (50 views ablation) | PSNR: 19.42 -> 21.57 | Improved |
| use color features in addition to distilled features (ablation) | VS-NeRF · Mip-NeRF360 (50 views ablation) | PSNR: 19.96 -> 21.57 | Improved |
| reduce distilled feature dimension from 32 to 16 (ablation) | VS-NeRF · Mip-NeRF360 (50 views ablation) | PSNR: 21.57 -> 20.04 | Hurt |
| increase distilled feature dimension from 32 to 64 (ablation) | VS-NeRF · Mip-NeRF360 (50 views ablation) | PSNR: 21.57 -> 21.65 | No clear effect |
| use monocular depth regularization instead of none | Nerfacto · Tanks&Temples Intermediate (8 scenes, 10-110 views averaged) | PSNR: 18.29 -> 13.81 | Hurt |
| add multi-view depth regularization | Nerfacto · Mip-NeRF360 (9 scenes, 10-110 views averaged) | PSNR: 19.05 -> 20.15 | Improved |
| add view-consistent sampling to Mip-NeRF360 | Mip-NeRF360 · Mip-NeRF360 | PSNR: 22.54 -> 23.37 | Improved |
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.
- Mip-NeRF360 — 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
- RegNeRF: Regularizing Neural Radiance Fields for View Synthesis From Sparse Inputs (CVPR 2022 · similarity 0.94)
- Reducing Shape-Radiance Ambiguity in Radiance Fields with a Closed-Form Color Estimation Method (NeurIPS 2023 · similarity 0.92)
- Dense Depth Priors for Neural Radiance Fields From Sparse Input Views (CVPR 2022 · similarity 0.92)
- AltNeRF: Learning Robust Neural Radiance Field via Alternating Depth-Pose Optimization (AAAI 2024 · similarity 0.91)
- Joint Optimization of Neural Radiance Fields and Continuous Camera Motion from a Monocular Video (CVPR 2025 · similarity 0.90)
- Learning Robust Generalizable Radiance Field with Visibility and Feature Augmented Point Representation (ICLR 2024 · 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.
