Research map / Image generation / Image restoration
Efficient Unified Demosaicing for Bayer and Non-Bayer Patterned Image Sensors
Haechang Lee, Dongwon Park, Wongi Jeong, Kijeong Kim, Hyunwoo Je, Dongil Ryu, Se Young Chun
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 two-stage knowledge learning (TKL) to baseline unified model (ablation) | Baseline UM · DF2K-CIS test | PSNR avg: 41.35 -> 41.40 | Improved |
| add adaptive discriminative filters (ADP) to baseline unified model (ablation) | Baseline UM · DF2K-CIS test | PSNR avg: 41.35 -> 41.46 | Improved |
| combine TKL and ADP (KLAP) (ablation) | Baseline UM · DF2K-CIS test | PSNR avg: 41.35 -> 41.71 | Improved |
| KLAP vs larger baseline UM-L with more blocks (ablation) | Baseline UM-L · DF2K-CIS test | PSNR avg: 41.40 -> 41.71 | Improved |
| select ADP filters by FAIG instead of randomly (ablation) | ADP · DF2K-CIS test | PSNR | Improved |
| increase ADP mask ratio from 1% to 15% (ablation) | KLAP · DF2K-CIS test | PSNR avg: 41.71 -> 41.87 | Improved |
| add meta-test learning (KLAP-M, 10 iterations) under strong noise (ablation) | KLAP · DF2K-CIS strong noise (Quad) | PSNR: 32.93 -> 35.41 | Improved |
| add meta-test learning on Bayer under strong noise (ablation) | KLAP · DF2K-CIS strong noise (Bayer) | PSNR: 32.98 -> 33.32 | Improved |
| fine-tune independent models after applying TKL (TKL-to-IM) (ablation) | Independent models · DF2K-CIS test | PSNR avg: 41.64 -> 41.86 | Improved |
Closest papers by meaning
- Super-Resolution Reconstruction from Bayer-Pattern Spike Streams (CVPR 2024 · similarity 0.83)
- Quad Bayer Joint Demosaicing and Denoising Based on Dual Encoder Network with Joint Residual Learning (AAAI 2024 · similarity 0.82)
- Joint Demosaicing and Denoising for Spike Camera (AAAI 2024 · similarity 0.82)
- Self-Supervised Learning for Color Spike Camera Reconstruction (CVPR 2025 · similarity 0.82)
- DiffRAW: Leveraging Diffusion Model to Generate DSLR-Comparable Perceptual Quality sRGB from Smartphone RAW Images (AAAI 2024 · similarity 0.81)
- Binarized Mamba-Transformer for Lightweight Quad Bayer HybridEVS Demosaicing (CVPR 2025 · similarity 0.81)
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.
