Research map / Learning theory and algorithms / Matrix and tensor methods
TV: A Sparse Optimization Method for Impulse Noise Image Restoration
Ganzhao Yuan, Bernard Ghanem
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 |
|---|---|---|---|
| use l0-norm data fidelity (PADMM) instead of l1TV-SBM | l0TV-PADMM · Random-value impulse, walkbridge (90% noise) | SNR0 (%): 32 -> 57 | Improved |
| use l0-norm fidelity PADMM instead of l02TV-AOP | l0TV-PADMM · Salt-and-pepper impulse, walkbridge (90% noise) | SNR0 (%): 36 -> 74 | Improved |
| use l0TV-PADMM instead of two-stage method TSM | l0TV-PADMM · Mixed impulse, walkbridge (90% noise) | SNR0 (%): 34 -> 63 | Improved |
| use l0TV-PADMM instead of BM3D | l0TV-PADMM · Mixed impulse, walkbridge (50% noise) | SNR0 (%): 55 -> 87 | Improved |
| use l0TV-PADMM instead of l1TV-SBM at low noise | l0TV-PADMM · Random-value impulse, walkbridge (10% noise) | SNR0 (%): 95 -> 97 | Improved |
| use proximal ADMM (PADMM) instead of PDA to solve l0TV | l0TV-PADMM · Random-value impulse, pepper (90% noise) | SNR0 (%): 39 -> 76 | Improved |
| use PADMM instead of PDA | l0TV-PADMM · Salt-and-pepper impulse, lake (90% noise) | SNR0 (%): 51 -> 84 | Improved |
| use l0-norm fidelity instead of lp (p=1/2) fidelity with ADMM | l0TV-PADMM · Random-value impulse, jetplane (90% noise) | SNR0 (%): 14 -> 74 | Improved |
| use l0TV-PADMM instead of MFM median filter at low pepper noise | l0TV-PADMM · Random-value impulse, pepper (10% noise) | SNR0 (%): 99 -> 99 | No clear effect |
| use l0TV-PADMM instead of TSM | l0TV-PADMM · Salt-and-pepper impulse, walkbridge (70% noise) | SNR0 (%): 87 -> 87 | No clear effect |
| use l0TV-PADMM instead of PDA at mild noise | l0TV-PADMM · Random-value impulse, pepper (10% noise) | SNR1: 20.2 -> 18.0 | Hurt |
Closest papers by meaning
- Modular Proximal Optimization for Multidimensional Total-Variation Regularization (JMLR 2018 · similarity 0.88)
- Adaptive Rates for Total Variation Image Denoising (JMLR 2020 · similarity 0.86)
- Learning to solve TV regularised problems with unrolled algorithms (NeurIPS 2020 · similarity 0.86)
- Estimating Sparse Signals With Smooth Support via Convex Programming and Block Sparsity (CVPR 2016 · similarity 0.86)
- Denoising as Adaptation: Noise-Space Domain Adaptation for Image Restoration (ICLR 2025 · similarity 0.85)
- The Perception-Robustness Tradeoff in Deterministic Image Restoration (ICML 2024 · 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.
