Research map / Image generation / Image restoration
Transformed Low-Rank Model for Line Pattern Noise Removal
Yi Chang, Luxin Yan, Sheng Zhong
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 the directional total variation prior to the transformed low-rank decomposition (ablation) | TLR · Mixed random and stripe noise removal (visual example) | stripe suppression (qualitative) | Improved |
| add the nonlocal low-rank prior to the image layer (ablation) | TLR · Mixed random and stripe noise removal (visual example) | random noise removal (qualitative) | Improved |
| combine directional TV and low-rank priors (ablation) | TLR · Mixed random and stripe noise removal (visual example) | mixed noise removal (qualitative) | Improved |
| apply low-rank prior to the rotated rain layer instead of the rotated image layer (ablation) | TLR · Low-rank matrix construction, Fig 3 | approximate rank of constructed matrix: approximately rank 30 -> approximately rank 1 | Improved |
| rotate the rain patch rather than use rain patch redundancy (ablation) | TLR · Low-rank matrix construction, Fig 3 | approximate rank of constructed matrix: approximately rank 8 -> approximately rank 1 | Improved |
| TLR versus GMM on dataset Fig 5 | TLR · Simulated rain streak, dataset of Li (Fig 5, light dense rain) | ILNIQE: 26.03 -> 25.12 | Improved |
| TLR versus GMM on dataset Fig 6 | TLR · Simulated rain streak, dataset of Li (Fig 6, sparse bright rain) | ILNIQE: 19.26 -> 18.72 | Improved |
| TLR versus GMM on SSIM, Fig 5 and 6 | TLR · Simulated rain streak, datasets of Li | SSIM: 0.77 -> 0.73; 0.92 -> 0.89 | Hurt |
| single-image TLR versus multi-band ASSTV | TLR · Simulated Pavia University hyperspectral (mixed random+stripe) | PSNR: 24.07 -> 30.52 | Improved |
| single-image TLR versus multi-band MoG | TLR · Simulated Pavia University hyperspectral (mixed random+stripe) | PSNR / SSIM: 30.78 -> 30.52 PSNR; 0.85 -> 0.89 SSIM | Mixed |
Closest papers by meaning
- Non-Local Meets Global: An Integrated Paradigm for Hyperspectral Denoising (CVPR 2019 · similarity 0.86)
- Joint Bi-Layer Optimization for Single-Image Rain Streak Removal (ICCV 2017 · similarity 0.86)
- Non-local Meets Global: An Integrated Paradigm for Hyperspectral Image Restoration (TPAMI 2020 · similarity 0.85)
- Simultaneous Fidelity and Regularization Learning for Image Restoration (TPAMI 2019 · similarity 0.85)
- FourierMamba: Fourier Learning Integration with State Space Models for Image Deraining (ICML 2025 · similarity 0.82)
- APR-RD: Complemental Two Steps for Self-Supervised Real Image Denoising (AAAI 2025 · similarity 0.82)
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.
