Research map / Image generation / Image restoration
Restormer: Efficient Transformer for High-Resolution Image Restoration
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 |
|---|---|---|---|
| Use Restormer instead of SPAIR | Image deraining · Five-dataset average | PSNR: 1.05 dB | Helped |
| Use Restormer instead of MPRNet | Single-image motion deblurring · GoPro, HIDE, RealBlur-R, RealBlur-J | PSNR: 0.26 dB | Helped |
| Use Restormer instead of IFAN | Defocus deblurring · DPDD combined scenes | PSNR: ∼0.6 dB | Helped |
| Replace the UNet Resblocks baseline with the full MDTA + GDFN Transformer block (ablation) | Restormer · Urban100 σ=50 | PSNR: 29.11 -> 29.62 | Helped |
| Replace MTA with MDTA, adding channel-wise transposed attention (ablation) | Restormer attention block · Urban100 σ=50 | PSNR: 29.28 -> 29.43 | Helped |
| Add gating to the feed-forward network (ablation) | Restormer feed-forward network · Urban100 σ=50 | PSNR: +0.12 dB | Helped |
| Add local depth-wise convolution to the feed-forward network (ablation) | Restormer feed-forward network · Urban100 σ=50 | PSNR: 29.42 -> 29.54 | Helped |
| Add decoder level-1 concatenation (ablation) | Restormer · Urban100 σ=50 | PSNR: 29.62 -> 29.66 | Helped |
| Add the refinement stage (ablation) | Restormer · Urban100 σ=50 | PSNR: 29.66 -> 29.71 | Helped |
| Train on progressively larger patches rather than fixed 128² patches (ablation) | Restormer · Urban100 σ=50 | PSNR: 29.71 -> 29.78 | Helped |
| Increase model dimension from 48 to 64 (ablation) | Restormer · Urban100 σ=50 | PSNR: 29.71 -> 29.63 | Hurt |
Closest papers by meaning
- Sharing Key Semantics in Transformer Makes Efficient Image Restoration (NeurIPS 2024 · similarity 0.87)
- DreamClear: High-Capacity Real-World Image Restoration with Privacy-Safe Dataset Curation (NeurIPS 2024 · similarity 0.87)
- Adapt or Perish: Adaptive Sparse Transformer with Attentive Feature Refinement for Image Restoration (CVPR 2024 · similarity 0.87)
- Omni-Kernel Network for Image Restoration (AAAI 2024 · similarity 0.87)
- Intra and Inter Parser-Prompted Transformers for Effective Image Restoration (AAAI 2025 · similarity 0.86)
- Efficient Visual State Space Model for Image Deblurring (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.