Research map / Learning theory and algorithms / Matrix and tensor methods
Training-Free Ultra Small Model for Universal Sparse Reconstruction in Compressed Sensing
Chaoqing Tang, Huanze Zhuang, Gui Yun Tian, Zhenli Zeng, Yijie Ding, Wenzhong Liu, Lin Lin, Xiang Bai
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 |
|---|---|---|---|
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · ECG, Sr=0.05 | Mean SSIM: 0.344 -> 0.549 | Improved |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · EEG, Sr=0.05 | Mean SSIM: 0.254 -> 0.735 | Improved |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · EEG, Sr=0.5 | Mean SSIM: 0.977 -> 0.980 | No clear effect |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · synthetic 1D, low Sr and high Kr | SSIM: 2 to 3 times SSIM | Improved |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · synthetic 1D, other regions | SSIM: similar accuracy to OMP | No clear effect |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · synthetic 1D, 1000 signals | time: 2 to 3 orders of magnitude less time | Improved |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · WSN, Sr=0.05 | Time (s): IHT-GPU 0.191 vs CLOMP 0.319 | Hurt |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · UHD 3840x2160, Sr=0.1 | Mean SSIM: 0.219 -> 0.707 | Improved |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · 4K image dataset | speedup: 117 and 29 times faster than OMP and IHT at Sr=0.1 | Improved |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · 4K image dataset, Sr=0.5 | speedup: 552 and 23 times faster | Improved |
| replace OMP pseudo-inverse with learnable coefficients (CLOMP) | OMP · FHD 1920x1080, Sr=0.5 | Time (s): 159.825 -> 0.268 | Improved |
| raise sampling rate Sr | CLOMP · 8 image datasets | Mean SSIM improvement over baselines: 292%, 98%, 45% at Sr=0.1, 0.3, 0.5 | Hurt |
Closest papers by meaning
- Modeling Sparse Deviations for Compressed Sensing using Generative Models (ICML 2018 · similarity 0.89)
- Deep Compressed Sensing (ICML 2019 · similarity 0.88)
- Practical Compact Deep Compressed Sensing (TPAMI 2024 · similarity 0.88)
- ISTA-Net: Interpretable Optimization-Inspired Deep Network for Image Compressive Sensing (CVPR 2018 · similarity 0.88)
- Learning Sparse and Low-Rank Priors for Image Recovery via Iterative Reweighted Least Squares Minimization (ICLR 2023 · similarity 0.88)
- The Price of Sparsity: Sufficient Conditions for Sparse Recovery using Sparse and Sparsified Measurements (NeurIPS 2025 · similarity 0.87)
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.
