Research map / Image generation / Image restoration
Defocus Deblurring Using Dual-Pixel Data
Abdullah Abuolaim, Michael S. Brown
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 dual-pixel L/R views as input instead of single blurred image (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 24.34 -> 25.13 | Improved |
| add dual-pixel L/R views as input instead of single blurred image (ablation) | DPDNet · DP defocus deblurring test (combined) | LPIPS: 0.277 -> 0.223 | Improved |
| add dual-pixel L/R views as input instead of single blurred image (ablation) | DPDNet · DP defocus deblurring test (outdoor) | PSNR: 22.25 -> 22.90 | Improved |
| add blurred combined image I_B as extra input (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 25.13 -> 25.07 | No clear effect |
| use full network instead of removing E-Block 1 and D-Block 4 (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 24.89 -> 25.13 | Improved |
| use full network instead of removing E-Block 1 and D-Block 4 (ablation) | DPDNet · DP defocus deblurring test (combined) | LPIPS: 0.268 -> 0.223 | Improved |
| train with 512x512 patches instead of 256x256 (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 25.01 -> 25.13 | No clear effect |
| filter 30% homogeneous patches instead of 0% (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 24.98 -> 25.13 | Improved |
| filter 30% homogeneous patches instead of 15% (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 25.01 -> 25.13 | Improved |
| filter 30% homogeneous patches instead of 45% (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 25.00 -> 25.13 | Improved |
| train with 16-bit data instead of 8-bit (ablation) | DPDNet · DP defocus deblurring test (combined) | LPIPS: 0.228 -> 0.223 | Improved |
| train with 16-bit data instead of 8-bit (ablation) | DPDNet · DP defocus deblurring test (combined) | PSNR: 25.18 -> 25.13 | Mixed |
Closest papers by meaning
- Learning to Deblur Using Light Field Generated and Real Defocus Images (CVPR 2022 · similarity 0.94)
- Learning To Reduce Defocus Blur by Realistically Modeling Dual-Pixel Data (ICCV 2021 · similarity 0.93)
- Quad-Pixel Image Defocus Deblurring: A New Benchmark and Model (CVPR 2025 · similarity 0.91)
- Residual Diffusion Deblurring Model for Single Image Defocus Deblurring (AAAI 2025 · similarity 0.90)
- LDP: Language-driven Dual-Pixel Image Defocus Deblurring Network (CVPR 2024 · similarity 0.90)
- Multi-Focus Image Fusion via Explicit Defocus Blur Modelling (AAAI 2025 · similarity 0.89)
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.
