Research map / Image generation / Image restoration
Efficient Dynamic Scene Deblurring Using Spatially Variant Deconvolution Network With Optical Flow Guided Training
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 a dilated backbone instead of the baseline backbone (ablation) | Deblurring network · GOPRO testing dataset | PSNR: 28.34 -> 29.01 | Helped |
| Add a dilated backbone instead of the baseline backbone (ablation) | Deblurring network · GOPRO testing dataset | SSIM: 0.9124 -> 0.9237 | Helped |
| Add the feature deconvolution module with modulated deformable convolutions (ablation) | Dilated-backbone deblurring network · GOPRO testing dataset | PSNR: 29.01 -> 29.57 | Helped |
| Add the feature deconvolution module with modulated deformable convolutions (ablation) | Dilated-backbone deblurring network · GOPRO testing dataset | SSIM: 0.9237 -> 0.9338 | Helped |
| Add optical flow guided training (ablation) | Feature-deconvolution deblurring network · GOPRO testing dataset | PSNR: 29.57 -> 29.81 | Helped |
| Add optical flow guided training (ablation) | Feature-deconvolution deblurring network · GOPRO testing dataset | SSIM: 0.9338 -> 0.9368 | Helped |
| Use Ours (B) rather than SRN | GOPRO deblurring · GOPRO testing dataset | PSNR: 30.10 -> 29.81 | Hurt |
| Use Ours (B) rather than DeblurGAN-v2 | GOPRO deblurring · GOPRO testing dataset | SSIM: 0.9340 -> 0.9368 | Helped |
| Use Ours (B) rather than SRN | GOPRO deblurring · GOPRO testing dataset, 1280×720 images | Runtime: 0.4 s -> 0.01 s | Helped |
| Use Ours (B) rather than DeblurGAN-v2 | GOPRO deblurring · GOPRO testing dataset, 1280×720 images | Runtime: 0.35 s -> 0.01 s | Helped |
| Use Ours (B) rather than DeepDeblur | GOPRO deblurring · GOPRO testing dataset | Model size: 303.6 MB -> 3.1 MB | Helped |
Closest papers by meaning
- Dynamic Scene Deblurring Using Spatially Variant Recurrent Neural Networks (CVPR 2018 · similarity 0.93)
- Deep Multi-Scale Convolutional Neural Network for Dynamic Scene Deblurring (CVPR 2017 · similarity 0.93)
- Exploiting Deblurring Networks for Radiance Fields (CVPR 2025 · similarity 0.89)
- Efficient Multi-scale Network with Learnable Discrete Wavelet Transform for Blind Motion Deblurring (CVPR 2024 · similarity 0.89)
- Parameterized Blur Kernel Prior Learning for Local Motion Deblurring (CVPR 2025 · similarity 0.88)
- DeblurDiff: Real-Word Image Deblurring with Generative Diffusion Models (NeurIPS 2025 · similarity 0.88)
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.