Research map / Image generation / Image restoration
Physics-Aware Accelerated Unrolling Model for Sparse-View CT Reconstruction
Shaojie Guo, Yingying Fang, Junkang Zhang, Yan Wang
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 global extrapolation (ablation) | PAUM · NBIA (64 views) ablation | PSNR: 39.46 -> 42.44 | Improved |
| use physics-aware dual-domain extrapolation instead of global (ablation) | PAUM · NBIA (64 views) ablation | PSNR: 42.44 -> 44.49 | Improved |
| swap row-wise and pixel-wise schemes between domains (ablation) | PAUM · NBIA (64 views) ablation | PSNR: 44.49 -> 43.78 | Hurt |
| increase inner extrapolation steps from 8 to 12 (ablation) | PAUM · NBIA (64 views) ablation | PSNR: 44.49 -> 44.60 | No clear effect |
| increase inner extrapolation steps from 0 to 8 (ablation) | PAUM · NBIA (64 views) ablation | PSNR: 39.46 -> 44.49 | Improved |
| add multi-scale conv to MLP-Mixer baseline (ablation) | BDRN · NBIA (64 views) ablation | PSNR: 40.51 -> 42.36 | Improved |
| add block-wise attention (ablation) | BDRN · NBIA (64 views) ablation | PSNR: 42.36 -> 43.88 | Improved |
| replace standard conv with deformable conv (ablation) | BDRN · NBIA (64 views) ablation | PSNR: 43.88 -> 44.49 | Improved |
| use DAT regularizer instead of BDRN (ablation) | PAUM · NBIA (64 views) ablation | PSNR: 44.49 -> 42.66 | Hurt |
| use PAUM vs LAMA | PAUM · NBIA 64 views | PSNR: 43.31 -> 44.49 | Improved |
Closest papers by meaning
- Deep Physics-Guided Unrolling Generalization for Compressed Sensing (IJCV 2023 · similarity 0.88)
- Extremely Sparse-View Computed Tomography from 2D Projections via Pose-Aware Diffusion Priors (NeurIPS 2026 · similarity 0.87)
- Near-Exact Recovery for Tomographic Inverse Problems via Deep Learning (ICML 2022 · similarity 0.87)
- QN-Mixer: A Quasi-Newton MLP-Mixer Model for Sparse-View CT Reconstruction (CVPR 2024 · similarity 0.86)
- Patient-Level Anatomy Meets Scanning-Level Physics: Personalized Federated Low-Dose CT Denoising Empowered by Large Language Model (CVPR 2025 · similarity 0.85)
- Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction Networks for Single-Pixel Imaging (CVPR 2025 · similarity 0.84)
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.
