Research map / 3D vision / Face and shape
Splat-Based Metal Artifact Reduction in Cone-Beam CT via Compact Attenuation Modeling
Kiseok Choi, Jaemin Cho, Inchul Kim, Min H. Kim
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 polychromatic forward model (ablation) | Gaussian splat CBCT baseline · Ablation (synthetic CBCT) | PSNR3D: 23.22 -> 27.97 | Improved |
| add polychromatic forward model (ablation) | Gaussian splat CBCT baseline · Ablation (synthetic CBCT) | SSIM3D: 0.984 -> 0.994 | Improved |
| add TV regularization (ablation) | Baseline + Poly · Ablation (synthetic CBCT) | PSNR3D: 27.97 -> 28.04 | Improved |
| add TV regularization (ablation) | Baseline + Poly · Ablation (synthetic CBCT) | SSIM3D: 0.994 -> 0.995 | Improved |
| increase spectral components from 15 to 63 (ablation) | Ours · Spectral components sweep | PSNR3D: 28.04 -> 27.99 | No clear effect |
| increase spectral components from 7 to 15 (ablation) | Ours · Spectral components sweep | PSNR3D: 28.01 -> 28.04 | No clear effect |
| use compact polychromatic splat method instead of Polyner | Ours · Synthetic CBCT Lung | PSNR3D: 20.63 -> 28.96 | Improved |
| use compact polychromatic splat method instead of Park et al. | Ours · Synthetic CBCT Broccoli | PSNR3D: 0.07 -> 27.76 | Improved |
| use splat-based compact material model instead of Polyner | Ours · Computation time, Synthetic Broccoli | computation time: 1h 50m 02s -> 28m 57s | Improved |
Closest papers by meaning
- Unsupervised Polychromatic Neural Representation for CT Metal Artifact Reduction (NeurIPS 2023 · similarity 0.92)
- Revisiting Pose Sensitivity in Splat-based Computed Tomography under Sparse-view Reconstruction (CVPR 2026 · similarity 0.88)
- X-Field: A Physically Informed Representation for 3D X-ray Reconstruction (NeurIPS 2025 · similarity 0.86)
- R$^2$-Gaussian: Rectifying Radiative Gaussian Splatting for Tomographic Reconstruction (NeurIPS 2024 · similarity 0.84)
- Structure-Aware Sparse-View X-ray 3D Reconstruction (CVPR 2024 · similarity 0.84)
- C^2RV: Cross-Regional and Cross-View Learning for Sparse-View CBCT Reconstruction (CVPR 2024 · 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.
