Research map / 3D vision / Gaussian splatting
Plug-and-Play Learned Proximal Trajectory for 3D Sparse-View X-Ray Computed Tomography
Romain Vo, Julie Escoda, Caroline Vienne, Etienne Decenciere
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 |
|---|---|---|---|
| train restoration network on samples from the optimization trajectory instead of only the initial FDK image (ablation) | LPT · Walnut-CBCT validation | PSNR: 25.67 -> 28.30 | Improved |
| train restoration network on optimization trajectory samples (ablation) | LPT · Cork-CBCT validation | PSNR: 33.69 -> 33.74 | Improved |
| add iteration step conditioning to the restoration network (ablation) | LPT · Walnut-CBCT validation | PSNR: 28.30 -> 28.44 | Improved |
| add iteration step conditioning to the restoration network (ablation) | LPT · Cork-CBCT validation | PSNR: 33.74 -> 34.41 | Improved |
| add Lipschitz constraint on the restoration network (ablation) | LPT · Walnut-CBCT validation | PSNR: 28.44 -> 26.18 | Hurt |
| add Lipschitz constraint on the restoration network (ablation) | LPT · Cork-CBCT validation | PSNR: 34.41 -> 33.06 | Hurt |
| replace Gaussian denoiser with trajectory-trained restoration network in PnP | PnP-PGD · Walnut-CBCT (30 views) | PSNR: 22.94 -> 25.26 | Improved |
| replace Gaussian denoiser with trajectory-trained restoration network in PnP | PnP-PGD · Cork-CBCT (30 views) | PSNR: 32.01 -> 32.11 | Mixed |
| use LPT instead of one-pass post-processing FDK-UNet | FDK-UNet · Walnut-CBCT (30 views) | data fidelity l/m: 7.82e−4 -> 6.69e−5 | Improved |
| use LPT instead of FDK-UNet | FDK-UNet · Walnut-CBCT (30 views) | SSIM: 0.767 -> 0.763 | Hurt |
Closest papers by meaning
- Proximal Algorithm Unrolling: Flexible and Efficient Reconstruction Networks for Single-Pixel Imaging (CVPR 2025 · similarity 0.87)
- Plug-and-Play image restoration with Stochastic deNOising REgularization (ICML 2024 · similarity 0.87)
- QN-Mixer: A Quasi-Newton MLP-Mixer Model for Sparse-View CT Reconstruction (CVPR 2024 · similarity 0.87)
- Provably Robust Score-Based Diffusion Posterior Sampling for Plug-and-Play Image Reconstruction (NeurIPS 2024 · similarity 0.87)
- PnP-Flow: Plug-and-Play Image Restoration with Flow Matching (ICLR 2025 · similarity 0.86)
- FiRe: Fixed-points of Restoration Priors for Solving Inverse Problems (CVPR 2025 · similarity 0.86)
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.
