Research map / 3D vision / NeRF and radiance fields
Generalizable Radio-Frequency Radiance Fields for Spatial Spectrum Synthesis
Kang Yang, Yuning Chen, Wan Du
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 geometry-aware cross-attention instead of dot-product attention (ablation) | GRaF · Unseen scenes | PSNR: 19.37 -> 20.96 | Improved |
| add neural ray tracing instead of NeRF2-style ray tracing (ablation) | GRaF · Unseen scenes | PSNR: 16.79 -> 20.96 | Improved |
| use GRaF instead of NeRF2 in unseen layouts | GRaF · Unseen layouts | PSNR: 12.76 -> 17.81 | Improved |
| use GRaF instead of NeRF2 in single scene | GRaF · Single-scene | PSNR: 19.93 -> 21.94 | Improved |
| train on 2.412 GHz and test on 5.805 GHz without retraining (ablation) | GRaF · Cross-frequency | PSNR: 24.91 -> 11.61 | Hurt |
| train on 2.412 GHz and test on 928 MHz without retraining (ablation) | GRaF · Cross-frequency | PSNR: 25.70 -> 14.84 | Hurt |
| augment 50% GT AoA training data with GRaF-synthesized spectra | AANN · AoA estimation | AoA error: -61.6% | Improved |
| use GRaF- instead of NeRF2-synthesized spectra for AoA training augmentation | AANN · AoA estimation | AoA error: -25.8% | Improved |
Closest papers by meaning
- RxGS: Receiver-Generalizable 3D Gaussian Splatting for Radio-Frequency Data Synthesis (NeurIPS 2026 · similarity 0.89)
- Radio Frequency Ray Tracing with Neural Object Representation for Enhanced RF Modeling (CVPR 2025 · similarity 0.88)
- GSRF: Complex-Valued 3D Gaussian Splatting for Efficient Radio-Frequency Data Synthesis (NeurIPS 2025 · similarity 0.86)
- SpectralNeRF: Physically Based Spectral Rendering with Neural Radiance Field (AAAI 2024 · similarity 0.86)
- Can NeRFs "See" without Cameras? (NeurIPS 2025 · similarity 0.85)
- GeRaF: Neural Geometry Reconstruction from Radio Frequency Signals (NeurIPS 2025 · similarity 0.85)
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.
