Research map / 3D vision / NeRF and radiance fields
Few-shot neural rendering architectures
Designs light-field, attention-based and reposable representations for view synthesis from few views.
53 accepted papers at ICML, NeurIPS, ICLR, CVPR and AAAI (2021–2026) · an approach in NeRF and radiance fields, keyword cluster “rendering · field · synthesis”
State History Tried Preprints Who Export
Where it stands
Not enough benchmark data Fewer than four papers report the same benchmark, so there is no record to follow.
35 papers in 2023–24 and 8 in 2025–26 (at ICML, NeurIPS, ICLR, CVPR and AAAI); its share of the topic shrank, from 13.2% to 8.9%.
How this is decided
A benchmark is tracked when at least four papers of this approach report their own method on the same dataset and metric in at least two different years. Its record is the best number reported up to 2024 compared with the best reported in 2025–26: still improving means at least 1% better, flat means at most 0.2% better with at least two papers from 2025–26 trying. Numbers on a different scale from the rest (a fraction among percentages) and far outliers are left out. Results come from different papers with different setups, so they are signals, not head-to-head comparisons. The share of papers is counted at ICML, NeurIPS, ICLR, CVPR and AAAI.
How it got here
Most cited paper of each year
- 2021 · NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in the Wild (NeurIPS 2021 · 51 citations)
- 2022 · NeRF in the Dark: High Dynamic Range View Synthesis From Noisy Raw Images (CVPR 2022 · 342 citations)
- 2023 · MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures (CVPR 2023 · 304 citations)
- 2024 · NeuRAD: Neural Rendering for Autonomous Driving (CVPR 2024 · 70 citations)
- 2025 · LookCloser: Frequency-aware Radiance Field for Tiny-Detail Scene (CVPR 2025 · 5 citations)
- 2026 · MAC-NeRF: Motion-Aware Curriculum Learning for Dynamic LiDAR NeRFs (ICML 2026)
Earliest papers here
- NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in the Wild (NeurIPS 2021 · 51 citations)
- Neural-PIL: Neural Pre-Integrated Lighting for Reflectance Decomposition (NeurIPS 2021 · 19 citations)
- NeRF in the Dark: High Dynamic Range View Synthesis From Noisy Raw Images (CVPR 2022 · 342 citations)
What has been tried
51 papers here report 518 measured changes (a component added, removed or swapped, compared with a baseline): helped 70%, no clear effect 6%, hurt 20%, mixed 3%. Papers mostly publish changes that helped, so "helped" is over-represented.
Ablations in recent papers
| What was changed | Measured on | Result | Effect | Paper |
|---|---|---|---|---|
| add CNN decoder | 20-sequence average · PSNR | 25.29 -> 27.26 | Helped | NeuRAD: Neural Rendering for Autonomous Driving (CVPR 2024 · 70 citations) |
| use quadratic warping instead of linear warping | NeRF rendering under large deformations · visual quality | Helped | PIE-NeRF: Physics-based Interactive Elastodynamics with NeRF (CVPR 2024 · 31 citations) | |
| add view dependence using spherical harmonics | All datasets · Mean SSIM | 0.022 ± 0.005 | Helped | DART: Implicit Doppler Tomography for Radar Novel View Synthesis (CVPR 2024 · 13 citations) |
| Add dynamic weighting to the decoupled-color loss | Glossy-object geometry reconstruction · Surface smoothness | Helped | Inverse Rendering of Glossy Objects via the Neural Plenoptic Function and Radiance Fields (CVPR 2024 · 13 citations) | |
| Replace the vanilla MLP with per-layer inputs incorporation (mi-MLP) | Blender, 8 views · PSNR | 14.73 -> 24.12 | Helped | Is Vanilla MLP in Neural Radiance Field Enough for Few-shot View Synthesis? (CVPR 2024 · 12 citations) |
| add patch sampling | KITTI image reconstruction · PSNR | 28.31 -> 27.04 | Hurt | PaReNeRF: Toward Fast Large-scale Dynamic NeRF with Patch-based Reference (CVPR 2024 · 8 citations) |
| Add the Frequency Grid instead of using a uniform frequency value | Music Room, normal-res(600×600) · PSNR | 31.95 -> 33.52 | Helped | LookCloser: Frequency-aware Radiance Field for Tiny-Detail Scene (CVPR 2025 · 5 citations) |
| add Conservation of Energy Loss | NeILF++ dataset (6-scene mean) · Albedo PSNR | albedo PSNR increasing by 2.34 | Helped | PBR-NeRF: Inverse Rendering with Physics-Based Neural Fields (CVPR 2025 · 5 citations) |
| Add distortion loss to DVGO | DTU · IMRC | 18.19 -> 18.47 | Helped | Evaluate Geometry of Radiance Fields with Low-Frequency Color Prior (AAAI 2024 · 4 citations) |
| use uniform medium temporal resolution (25) instead of motion-aware temporal sampling | Truck · PSNR | 27.49 -> 26.93 | Hurt | Gear-NeRF: Free-Viewpoint Rendering and Tracking with Motion-aware Spatio-Temporal Sampling (CVPR 2024 · 4 citations) |
| Add progressive refinement of emissive sources with reflection awareness | White-colored source identification · IoU | 0.40 -> 0.86 | Helped | ESR-NeRF: Emissive Source Reconstruction Using LDR Multi-view Images (CVPR 2024 · 3 citations) |
| Use photometric and kernel losses only, without motion regularizers | Cube, test views, overall · PSNR | 34.12 -> 34.70 | Mixed | DynaMoDe-NeRF: Motion-aware Deblurring Neural Radiance Field for Dynamic Scenes (CVPR 2025 · 2 citations) |
Recent preprints
Loading recent arXiv preprints…
Who publishes on it
Institutions on the most papers of this approach since 2023 (of 43 papers; from OpenAlex and the papers' affiliations).
- Carnegie Mellon University · 3 papers
- Google · 3 papers
- Zhejiang University · 3 papers
- University of Oxford · 2 papers
- SenseTime · 2 papers
- Northwestern Polytechnical University · 2 papers
- Seoul National University · 2 papers
- Simon Fraser University · 2 papers
- University of Toronto · 2 papers
- University of Bologna · 2 papers
Use it
Download BibTeX of 38 key papersOpen the topic in the interactive map
Get alerts. Describe your idea in scime atlas and save it (free scime account): you get a weekly email when new preprints land close to it.
Related approaches
Same keyword cluster
- Inverse rendering with neural fields (110 papers)
- 3D-aware generative radiance fields (20 papers)
Elsewhere in the topic
- Sparse-input NeRF regularization (106 papers)
- Dynamic and editable NeRF (88 papers)
- Single-view and streaming radiance fields (49 papers)
- NeRF for scene decomposition and understanding (30 papers)
Benchmark numbers, ablations and baseline numbers are extracted by AI from each paper's full text and can be wrong; the paper is the authority. Data: results index of 2026-11-02, citation counts from OpenAlex (2026-10-09).