Research map / 3D vision / NeRF and radiance fields
Is Attention All That NeRF Needs?
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 |
|---|---|---|---|
| Use epipolar-geometry-constrained view aggregation and learned ray rendering | GNT · LLFF (cross-scene generalization) | Avg: 0.055 -> 0.047 | Helped |
| Use epipolar-geometry-constrained view aggregation and learned ray rendering | GNT · NeRF Synthetic (cross-scene generalization) | Avg: 0.032 -> 0.029 | Helped |
| Use cross-scene-trained GNT for complex optics | GNT · Shiny-6 | PSNR: 24.12 -> 27.10 | Helped |
| Use GNT rather than SRT | Novel-view synthesis · NMR | PSNR: 27.87 -> 32.12 | Helped |
| Use one transformer stage instead of separate view and ray aggregation (ablation) | GNT · Drums (Blender) | PSNR: 29.41 -> 21.80 | Hurt |
| Aggregate epipolar features before aggregating views (ablation) | GNT · Drums (Blender) | PSNR: 29.41 -> 21.57 | Hurt |
| Use dot-product attention rather than subtraction attention in the view transformer (ablation) | GNT · Drums (Blender) | PSNR: 29.41 -> 26.98 | Hurt |
| Use subtraction attention rather than dot-product attention in the view transformer (ablation) | GNT · Drums (Blender) | SSIM: 0.953 -> 0.931 | No clear effect |
| Replace learned ray rendering with classical volumetric rendering (ablation) | GNT · Drums (Blender) | PSNR: 29.41 -> 24.24 | Hurt |
| Add autoregressive ray decoding (ablation) | GNT · Orchids (LLFF) | PSNR: 20.67 -> 21.05 | Helped |
| Add autoregressive ray decoding (ablation) | GNT · Horns (LLFF) | PSNR: 29.62 -> 28.20 | Hurt |
| Add attention-guided fine sampling (ablation) | GNT · Orchids (LLFF) | PSNR: 20.67 -> 20.69 | No clear effect |
Closest papers by meaning
- To View Transform or Not to View Transform: NeRF-based Pre-training Perspective (ICLR 2026 · similarity 0.91)
- Rad-NeRF: Ray-decoupled Training of Neural Radiance Field (NeurIPS 2024 · similarity 0.90)
- NeRF-VPT: Learning Novel View Representations with Neural Radiance Fields via View Prompt Tuning (AAAI 2024 · similarity 0.90)
- NeRF On-the-go: Exploiting Uncertainty for Distractor-free NeRFs in the Wild (CVPR 2024 · similarity 0.90)
- A General Implicit Framework for Fast NeRF Composition and Rendering (AAAI 2024 · similarity 0.90)
- GoLF-NRT: Integrating Global Context and Local Geometry for Few-Shot View Synthesis (CVPR 2025 · 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.