Research map / 3D vision / NeRF and radiance fields
FlipNeRF: Flipped Reflection Rays for Few-shot Novel View Synthesis
Seunghyeon Seo, Yeonjin Chang, Nojun Kwak
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 flipped rays alone (NLL on flipped rays) without UE and BFC losses (ablation) | FlipNeRF · DTU 3-view | PSNR: 17.44 -> 16.94 | Hurt |
| add UE loss on both original and flipped rays (ablation) | FlipNeRF · DTU 3-view | PSNR: 16.94 -> 17.89 | Improved |
| add BFC loss (ablation) | FlipNeRF · DTU 3-view | PSNR: 17.89 -> 19.01 | Improved |
| add orientation loss from Ref-NeRF (ablation) | FlipNeRF · DTU 3-view | PSNR: 19.01 -> 19.55 | Improved |
| use naive Emptiness loss instead of UE loss (ablation) | FlipNeRF · DTU 3-view | PSNR: 19.30 -> 18.62 (no emptiness loss vs naive Emptiness) | Hurt |
| replace Emptiness loss with uncertainty-aware emptiness (UE) loss (ablation) | FlipNeRF · DTU 3-view | NLL: 4.92 -> 2.56 | Improved |
| angle masking threshold tau 90 deg versus no masking (ablation) | flipped reflection rays · DTU 3-view | PSNR: 18.76 -> 19.55 | Improved |
| use narrower masking thresholds (30 or 60 deg) (ablation) | flipped reflection rays · DTU 3-view | PSNR: 19.55 -> 18.62 (30 deg), 18.12 (60 deg) | Hurt |
| apply BFC with MSE or cosine similarity instead of JSD (ablation) | BFC loss · DTU 3-view | PSNR: 18.79 -> 18.83 (MSE), 18.77 (cosine) | No clear effect |
| apply JSD BFC on direction-conditioned bottleneck feature instead of b (ablation) | BFC loss · DTU 3-view | PSNR: 19.55 -> 18.43 | Hurt |
| FlipNeRF versus MixNeRF on LLFF 3-view | FlipNeRF · LLFF 3-view | PSNR: 19.27 -> 19.34 | No clear effect |
Later papers that compared with this method
3 later accepted papers report FlipNeRF as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- SparseCraft: Few-Shot Neural Reconstruction through Stereopsis Guided Geometric Linearization (ECCV 2024 · better on 4 of 4 shared results)
- Global Latent Neural Rendering (CVPR 2024 · better on 2 of 2 shared results)
- ZeroRF: Fast Sparse View 360deg Reconstruction with Zero Pretraining (CVPR 2024 · better on 1 of 1 shared result)
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- MixNeRF — MixNeRF: Modeling a Ray With Mixture Density for Novel View Synthesis From Sparse Inputs (CVPR 2023) · better on 7 of 7 shared results
- Ref-NeRF — Ref-NeRF: Structured View-Dependent Appearance for Neural Radiance Fields (CVPR 2022) · better on 2 of 2 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
- Few-Shot Neural Radiance Fields under Unconstrained Illumination (AAAI 2024 · similarity 0.88)
- FrugalNeRF: Fast Convergence for Extreme Few-shot Novel View Synthesis without Learned Priors (CVPR 2025 · similarity 0.88)
- AltNeRF: Learning Robust Neural Radiance Field via Alternating Depth-Pose Optimization (AAAI 2024 · similarity 0.88)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.87)
- HybridNeRF: Efficient Neural Rendering via Adaptive Volumetric Surfaces (CVPR 2024 · similarity 0.87)
- ColNeRF: Collaboration for Generalizable Sparse Input Neural Radiance Field (AAAI 2024 · similarity 0.87)
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.
