Research map / Image generation / Super-resolution
RAW-Domain Degradation Models for Realistic Smartphone Super-Resolution
Ali Mosleh, Faraz Ali, Fengjia Zhang, Stavros Tsogkas, Junyong Lee, Michael S. Brown, Alex Levinshtein
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 calibrated camera-specific blur and noise instead of generic degradation pools | RRDBNet RAW-to-RGB 4x SR · S24U Tele 2 (camera-specific) | MTF50: 1.62 -> 2.00 (vs KernelGAN) | Improved |
| train on calibrated camera-specific degradations vs best baseline | RRDBNet RAW-to-RGB 4x SR · Pixel 9 Pro Tele (camera-specific) | PSNR: 33.32 vs 32.66 (KernelGAN), SSIM 0.889 ties Real-ESRGAN | Mixed |
| train on calibrated camera-specific degradations vs best baseline | RRDBNet RAW-to-RGB 4x SR · S23U Tele 1 (camera-specific) | PSNR: 33.47 -> 33.59 | Improved |
| train with degradation models from seven other cameras vs BSRAW/RAWSR/Real-ESRGAN pools | RRDBNet RAW-to-RGB 4x SR · Pixel 6 Main (held-out, cross-camera) | PSNR: 31.75 -> 32.41 | Improved |
| train with degradation models from seven other cameras vs baselines | RRDBNet RAW-to-RGB 4x SR · Mi 11 Main (held-out, cross-camera) | PSNR: 35.71 -> 36.11 | Improved |
| use calibrated kernels rather than bicubic downsampling kernels | RRDBNet RAW-to-RGB 4x SR · S23U Tele 1 (camera-specific) | MTF50: 0.05 -> 1.50 | Improved |
Closest papers by meaning
- Camera Lens Super-Resolution (CVPR 2019 · similarity 0.88)
- Unsupervised Diffusion-Based Degradation Modeling for Real-World Super-Resolution (AAAI 2025 · similarity 0.87)
- PocketSR: The Super-Resolution Expert in Your Pocket Mobiles (NeurIPS 2025 · similarity 0.86)
- Edge-SD-SR: Low Latency and Parameter Efficient On-device Super-Resolution with Stable Diffusion via Bidirectional Conditioning (CVPR 2025 · similarity 0.86)
- SEAL: A Framework for Systematic Evaluation of Real-World Super-Resolution (ICLR 2024 · similarity 0.85)
- Towards Realistic Data Generation for Real-World Super-Resolution (ICLR 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.
