Research map / Image generation / Image restoration
RAW Image Reconstruction Using a Self-Contained sRGB-JPEG Image With Only 64 KB Overhead
Rang M. H. Nguyen, Michael S. Brown
Check 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 breadth-first octree partitioning to select gamut-mapping control points instead of uniform partitioning (ablation) | RAW reconstruction · RAW-sRGB image pairs | RMSE: 0.0037 -> 0.0023 | Improved |
| Use breadth-first octree partitioning to select gamut-mapping control points instead of k-means (ablation) | RAW reconstruction · RAW-sRGB image pairs | RMSE: 0.0026 -> 0.0023 | Improved |
| Use breadth-first octree partitioning instead of k-means for control-point selection (ablation) | RAW reconstruction · RAW-sRGB image pairs | Time (s): 101.70 -> 11.18 | Improved |
| Use breadth-first octree partitioning instead of uniform partitioning for control-point selection (ablation) | RAW reconstruction · RAW-sRGB image pairs | Time (s): 0.53 -> 11.18 | Hurt |
| Reconstruct RAW with the proposed method instead of Yuan and Sun using half-resolution RAW | RAW reconstruction · Canon 1Ds; Canon 600D; Nikon D5200; Sony α57 | RMSE: 0.0049 -> 0.0018; 0.0085 -> 0.0038; 0.0078 -> 0.0033; 0.0055 -> 0.0020 | Improved |
| Reconstruct RAW with the proposed method instead of Yuan and Sun using 100 × 90 RAW | RAW reconstruction · Canon 1Ds; Canon 600D; Nikon D5200; Sony α57 | RMSE: 0.0135 -> 0.0018; 0.0191 -> 0.0038; 0.0173 -> 0.0033; 0.0150 -> 0.0020 | Improved |
| Lower JPEG quality from fine to normal or basic (ablation) | Proposed RAW reconstruction · Nikon D5200 image | RMSE: 0.0020 -> 0.0021; 0.0020 -> 0.0022 | Hurt |
| Correct white balance using reconstructed RAW rather than directly on sRGB | White-balance correction · Canon 1Ds Mark III; Nikon D5200; Sony α57 | White-balance correction quality | Improved |
| Deblur using reconstructed RAW rather than sRGB | Image deblurring · Sony A200; Canon 1Ds Mark III | SNR: 28.88 dB vs 24.00 dB; 19.14 dB vs 13.90 dB | Improved |
Closest papers by meaning
- RAW Image Reconstruction Using a Self-contained sRGB–JPEG Image with Small Memory Overhead (IJCV 2017 · similarity 0.98)
- Leveraging Frame Affinity for sRGB-to-RAW Video De-rendering (CVPR 2024 · similarity 0.87)
- Compress Clean Signal from Noisy Raw Image: A Self-Supervised Approach (ICML 2024 · similarity 0.84)
- DiffRAW: Leveraging Diffusion Model to Generate DSLR-Comparable Perceptual Quality sRGB from Smartphone RAW Images (AAAI 2024 · similarity 0.84)
- ReRAW: RGB-to-RAW Image Reconstruction via Stratified Sampling for Efficient Object Detection on the Edge (CVPR 2025 · similarity 0.84)
- ISPDiffuser: Learning RAW-to-sRGB Mappings with Texture-Aware Diffusion Models and Histogram-Guided Color Consistency (AAAI 2025 · similarity 0.83)
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.
