Research map / Architectures and efficiency / RNNs and normalization
Learned D-AMP: Principled Neural Network based Compressive Image Recovery
Chris Metzler, Ali Mousavi, Richard Baraniuk
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 |
|---|---|---|---|
| unroll D-AMP with learned DnCNN denoisers and retain the Onsager correction | Image compressive recovery · Gaussian measurements, m/n=0.10–0.25 | PSNR: 23.1 -> 23.7; 25.1 -> 25.7; 26.6 -> 27.2; 27.9 -> 28.5 | Improved |
| use unrolled D-AMP with learned DnCNN denoisers | Image compressive recovery · Coded diffraction, m/n=0.10–0.25 | PSNR: 24.0 -> 25.3; 25.7 -> 27.4; 27.5 -> 28.9; 29.1 -> 30.5 | Improved |
| add the Onsager correction to D-IT (ablation) | Learned D-AMP image recovery · Gaussian measurements, m/n=0.10–0.25 | PSNR: 20.1 -> 23.7; 20.7 -> 25.7; 21.1 -> 27.2; 21.7 -> 28.5 | Improved |
| train LDAMP layers sequentially instead of end-to-end (ablation) | LDAMP · Gaussian measurements, trained and tested at m/n=0.20 | PSNR: 33.1 -> 33.1 | No clear effect |
| train denoisers separately on AWGN denoising problems (ablation) | LDAMP · Gaussian measurements, trained and tested at m/n=0.20 | PSNR: 33.1 -> 31.6 | Hurt |
| train denoisers separately without tying training to a sampling rate (ablation) | LDAMP · Gaussian measurements, trained at m/n=0.20 and tested at m/n=0.05 | PSNR: 18.7 -> 25.9 | Improved |
Closest papers by meaning
- Training Deep Learning Based Image Denoisers From Undersampled Measurements Without Ground Truth and Without Image Prior (CVPR 2019 · similarity 0.88)
- Practical Compact Deep Compressed Sensing (TPAMI 2024 · similarity 0.88)
- Compressive sensing with un-trained neural networks: Gradient descent finds a smooth approximation (ICML 2020 · similarity 0.88)
- Untrained Neural Nets for Snapshot Compressive Imaging: Theory and Algorithms (NeurIPS 2024 · similarity 0.86)
- Multi-Cross Sampling and Frequency-Division Reconstruction for Image Compressed Sensing (AAAI 2024 · similarity 0.86)
- Using Powerful Prior Knowledge of Diffusion Model in Deep Unfolding Networks for Image Compressive Sensing (CVPR 2025 · similarity 0.85)
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.
