Research map / Image generation / Image restoration
Spatially-Adaptive Image Restoration Using Distortion-Guided Networks
Kuldeep Purohit, Maitreya Suin, A. N. Rajagopalan, Vishnu Naresh Boddeti
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 spatial feature modulator (SFM) guided by Net_L (ablation) | dense encoder-decoder · Raindrop AGAN | PSNR (dB): 30.72 -> 31.39 | Improved |
| add spatial feature modulator (SFM) guided by Net_L (ablation) | dense encoder-decoder · GoPro deblurring | PSNR (dB): 30.85 -> 31.44 | Improved |
| replace dense conv blocks in decoder with mask-guided sparse convolution (SC) (ablation) | Net2 · Raindrop AGAN | PSNR (dB): 31.39 -> 31.80 | Improved |
| replace dense conv blocks in decoder with mask-guided sparse convolution (SC) (ablation) | Net2 · GoPro deblurring | PSNR (dB): 31.44 -> 31.62 | Improved |
| add generic non-local layer (NL) in decoder (ablation) | Net3 · Raindrop AGAN | PSNR (dB): 31.80 -> 32.17 | Improved |
| use distortion-guided sparse non-local (SNL) instead of generic non-local (NL) (ablation) | Net4 · Raindrop AGAN | PSNR (dB): 32.17 -> 32.73 | Improved |
| use distortion-guided sparse non-local (SNL) instead of generic non-local (NL) (ablation) | Net4 · GoPro deblurring | PSNR (dB): 31.79 -> 32.06 | Improved |
| use SPAIR instead of MSPFN | deraining · average of 5 test sets | PSNR (dB): 30.75 -> 32.91 | Improved |
| use SPAIR instead of RCDNet on heavy rain | deraining · Rain100H (trained on Rain100H) | PSNR (dB): 31.28 -> 31.69 | Improved |
| use SPAIR instead of RCDNet on light real rain | deraining · SPANet | PSNR (dB): 41.47 -> 44.10 | Improved |
| use SPAIR instead of best prior shadow method | shadow removal · ISTD | RMSE: 6.37 -> 5.88 | Improved |
| use SPAIR instead of Suin et al., trained only on GoPro | deblurring · HIDE | PSNR (dB): 29.98 -> 30.29 | Improved |
Closest papers by meaning
- Adapt or Perish: Adaptive Sparse Transformer with Attentive Feature Refinement for Image Restoration (CVPR 2024 · similarity 0.89)
- StarIR: Convolutional Image Restoration With Spatial-Frequency Fusion (TPAMI 2026 · similarity 0.89)
- GIQE: Generic Image Quality Enhancement via Nth Order Iterative Degradation (CVPR 2022 · similarity 0.88)
- Image Restoration via Multi-Domain Learning (TPAMI 2026 · similarity 0.88)
- AdaIR: Adaptive All-in-One Image Restoration via Frequency Mining and Modulation (ICLR 2025 · similarity 0.87)
- Continuous Expert Assembly: Instance-Conditioned Low-Rank Residuals for All-in-One Image Restoration (NeurIPS 2026 · similarity 0.87)
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.
