Research map / Image generation / Super-resolution
CuNeRF: Cube-Based Neural Radiance Field for Zero-Shot Medical Image Arbitrary-Scale Super Resolution
Zixuan Chen, Lingxiao Yang, Jian-Huang Lai, Xiaohua Xie
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 cube-based sampling instead of ray sampling (ablation) | NeRF-dagger · MSD MRI 3D MISR x2 | PSNR: 29.33 -> 35.82 | Improved |
| Add isotropic volume rendering (ablation) | CuNeRF with CuS · MSD MRI 3D MISR x2 | PSNR: 35.82 -> 38.15 | Improved |
| Add adaptive rendering loss (ablation) | CuNeRF with CuS and IVR · MSD MRI 3D MISR x2 | PSNR: 38.15 -> 39.62 | Improved |
| Use l-infinity distance instead of l2 in cube rendering (ablation) | CuNeRF · MSD MRI 3D MISR x2 | PSNR: 39.62 -> 35.78 | Hurt |
| Vary cube edge length from 1 to 0.5 or 2 (ablation) | CuNeRF · MSD MRI 3D MISR | PSNR: 39.62 vs 38.17 (l=0.5), 39.65 (l=2) at x2 | No clear effect |
| Use zero-shot CuNeRF in place of supervised ArSSR on MRI | CuNeRF · MSD MRI 3D MISR | PSNR: 36.98 -> 39.62 at x2 | Improved |
| Use zero-shot CuNeRF instead of supervised SAINT/TVSRN on CT | CuNeRF · KiTS19 volumetric MISR | PSNR: x2: 39.47 -> 38.33; x8: 32.10 -> 32.44 | Mixed |
Closest papers by meaning
- MedGSSR: Generalizable Medical Image Super-Resolution 3D Reconstruction via Hierarchical Feed-forward Gaussian Splatting (ECCV 2026 · similarity 0.87)
- Rethinking Multi-Contrast MRI Super-Resolution: Rectangle-Window Cross-Attention Transformer and Arbitrary-Scale Upsampling (ICCV 2023 · similarity 0.87)
- CycleINR: Cycle Implicit Neural Representation for Arbitrary-Scale Volumetric Super-Resolution of Medical Data (CVPR 2024 · similarity 0.85)
- ColNeRF: Collaboration for Generalizable Sparse Input Neural Radiance Field (AAAI 2024 · similarity 0.84)
- Sparse Voxels Rasterization: Real-time High-fidelity Radiance Field Rendering (CVPR 2025 · similarity 0.83)
- FrugalNeRF: Fast Convergence for Extreme Few-shot Novel View Synthesis without Learned Priors (CVPR 2025 · similarity 0.83)
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.
