Research map / Image generation / Image restoration
Dual Pixel Exploration: Simultaneous Depth Estimation and Image Restoration
Liyuan Pan, Shah Chowdhury, Richard Hartley, Miaomiao Liu, Hongguang Zhang, Hongdong Li
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 |
|---|---|---|---|
| Add the DP-model-based reblur loss (ablation) | DDDNet · Our-syn | abs_rel: 0.091 -> 0.083 | Improved |
| Add the DP-model-based reblur loss (ablation) | DDDNet · Our-syn | PSNR: 32.171 -> 33.218 | Improved |
| Add the DP-model-based reblur loss (ablation) | DDDNet · DPD-blur | PSNR: 26.15/26.76 | Improved |
| Add the DP-model-based reblur loss (ablation) | DDDNet · Our-real | SSIM: 0.826/0.850 | Improved |
| Jointly optimize depth estimation and deblurring rather than use coarse DepthNet depth alone (ablation) | DDDNet on Our-syn · Our-syn | abs_rel: 0.149 -> 0.091 | Improved |
| Fine-tune on test images using reblur loss without ground-truth depth (ablation) | DDDNet trained on Our-syn · DPD-disp | Geometric Mean: 0.1207 -> 0.1098 | Improved |
| Train on simulated DP data, then fine-tune on DPD-blur | DDDNet · DPD-blur | PSNR: 20.28dB -> 26.92dB | Improved |
| Fine-tune the simulated-data model with half the DPD-blur images (ablation) | DDDNet · DPD-blur | PSNR: 26.52dB | Improved |
Closest papers by meaning
- Weakly-Supervised Depth Estimation and Image Deblurring via Dual-Pixel Sensors (TPAMI 2024 · similarity 0.94)
- LDP: Language-driven Dual-Pixel Image Defocus Deblurring Network (CVPR 2024 · similarity 0.91)
- K3DN: Disparity-Aware Kernel Estimation for Dual-Pixel Defocus Deblurring (CVPR 2023 · similarity 0.90)
- Quad-Pixel Image Defocus Deblurring: A New Benchmark and Model (CVPR 2025 · similarity 0.88)
- Passive Snapshot Coded Aperture Dual-Pixel RGB-D Imaging (CVPR 2024 · similarity 0.86)
- Blurry-Edges: Photon-Limited Depth Estimation from Defocused Boundaries (CVPR 2025 · similarity 0.86)
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.
