Research map / Optimization / Optimizers
Stochastic Solutions for Linear Inverse Problems using the Prior Implicit in a Denoiser
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Average over 10 sampled reconstructions (ablation) | denoiser-prior sampler · Super-resolution Set5 4:1 | PSNR (dB): 29.47 -> 31.20 | Helped |
| Average over 10 sampled reconstructions (ablation) | denoiser-prior sampler · Compressive sensing Set68 25% | PSNR (dB): 29.16 -> 29.74 | Helped |
| Single-sample sampler instead of DIP/DeepRED | denoiser-prior sampler · Super-resolution Set5 4:1 | PSNR (dB): 30.22 -> 29.47 | Hurt |
| Single-sample sampler instead of DIP/DeepRED | denoiser-prior sampler · Deblurring Set5 5% | PSNR (dB): 29.91 -> 29.22 | Hurt |
| Sampler instead of BNN/ISTA-Net | denoiser-prior sampler · Compressive sensing Set68 10% | PSNR (dB): 25.24 -> 25.47 | Helped |
| Sampler instead of BNN | denoiser-prior sampler · Compressive sensing Set68 4% | PSNR (dB): 22.52 -> 22.07 | Hurt |
| Use sampler instead of DeepRED for speed | denoiser-prior sampler · Super-resolution Set14 | runtime (sec): 1,584 -> 9 | Helped |
| Inject noise (beta < 1) rather than none (ablation) | sampler · constrained linear inverse problems | convergence | Mixed |
| Vary h0 and beta (ablation) | sampler · convergence on 50 images | iterations to converge | No clear effect |
Closest papers by meaning
- Exact Posterior Score Estimation for Solving Linear Inverse Problems (NeurIPS 2026 · similarity 0.91)
- Stochastic Deep Restoration Priors for Imaging Inverse Problems (ICML 2025 · similarity 0.91)
- A Restoration Network as an Implicit Prior (ICLR 2024 · similarity 0.90)
- Unleashing the Denoising Capability of Diffusion Prior for Solving Inverse Problems (NeurIPS 2024 · similarity 0.90)
- A Variational Perspective on Solving Inverse Problems with Diffusion Models (ICLR 2024 · similarity 0.90)
- Plug-and-Play image restoration with Stochastic deNOising REgularization (ICML 2024 · similarity 0.89)
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.