Research map / Image generation / Image restoration
CogSENet: Blind Image Deblurring with Blur-Conditioned Semantic Routing and Explicit Frequency Fusion
Pan Wang, Yihao Hu, Xiujin Liu
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 |
|---|---|---|---|
| use CogSENet (8.9M) over FFTformer (16.6M) | CogSENet · GoPro | PSNR: 34.21 -> 34.72 | Improved |
| use FullBFFB at bottleneck (Ours+) instead of LightBFFB only | CogSENet · GoPro | PSNR: 34.72 -> 34.91 | Improved |
| add semantic-driven state space module (SDSSM) (ablation) | CogSENet · GoPro | PSNR: 33.95 -> 34.31 | Improved |
| add semantic prompt in SDSSM (ablation) | SDSSM · GoPro | PSNR: 34.15 -> 34.31 | Improved |
| add semantic token regrouping in SDSSM (ablation) | SDSSM · GoPro | PSNR: 34.21 -> 34.31 | Improved |
| add implicit frequency branch in BFFB (ablation) | BFFB · GoPro | PSNR: 34.15 -> 34.31 | Improved |
| joint CBF and CLIP modulation instead of CBF only (ablation) | CogSENet · GoPro | PSNR: 33.80 -> 34.31 | Improved |
| joint CBF and CLIP modulation instead of CLIP only (ablation) | CogSENet · HIDE | PSNR: 31.29 -> 31.68 | Improved |
| apply CogSENet to Rain100H | CogSENet · Rain100H | PSNR: 32.17 (IDT) vs 32.15, +0.69 over Restormer 31.46 | Mixed |
| apply CogSENet to BSD68 denoising across noise levels | CogSENet · BSD68 | PSNR: sigma=15: 34.02 vs AirNet 34.14; sigma=50: 28.29 vs 28.23 | Mixed |
| apply CogSENet to SOTS dehazing | CogSENet · RESIDE SOTS | PSNR: 28.21 -> 28.72 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Restormer — Restormer: Efficient Transformer for High-Resolution Image Restoration (CVPR 2022) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Surface-Aware Blind Image Deblurring (TPAMI 2019 · similarity 0.87)
- Reblurring-Guided Single Image Defocus Deblurring: A Learning Framework with Misaligned Training Pairs (IJCV 2025 · similarity 0.86)
- Blur2Blur: Blur Conversion for Unsupervised Image Deblurring on Unknown Domains (CVPR 2024 · similarity 0.86)
- GS-Blur: A 3D Scene-Based Dataset for Realistic Image Deblurring (NeurIPS 2024 · similarity 0.85)
- BlurDM: A Blur Diffusion Model for Image Deblurring (NeurIPS 2025 · similarity 0.85)
- Unsupervised Blind Image Deblurring Based on Self-Enhancement (CVPR 2024 · similarity 0.85)
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.
