Research map / Image generation / Image restoration
A Comparative Study of Image Restoration Networks for General Backbone Network Design
Xiangyu Chen, Zheyuan Li, Yuandong Pu, Yihao Liu, Jiantao Zhou, Yu Qiao, Chao Dong
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 |
|---|---|---|---|
| replace half of the transposed self-attention blocks with overlapping cross-attention spatial blocks | Restormer · x4 SR Manga109 | PSNR (dB): 31.96 -> 32.38 | Improved |
| add spatial self-attention blocks to Restormer | Restormer · Denoising sigma=50 Urban100 | PSNR (dB): 30.02 -> 30.24 | Improved |
| add spatial self-attention blocks to Restormer | Restormer · Deblurring HIDE | PSNR (dB): 31.22 -> 31.76 | Improved |
| add spatial self-attention blocks to Restormer | Restormer · Deraining Rain100H | PSNR (dB): 31.48 -> 32.09 | Improved |
| add spatial self-attention blocks to Restormer | Restormer · Deraining Test1200 | PSNR (dB): 33.22 -> 32.31 | Hurt |
| add spatial self-attention blocks to Restormer | Restormer · Dehazing SOTS Indoor | PSNR (dB): 41.97 -> 42.90 | Improved |
| add spatial self-attention blocks to Restormer | Restormer · All-in-one dehazing | PSNR (dB): 36.95 -> 38.24 | Improved |
| use X-Restormer (U-shape) versus plain-architecture HAT on SR | X-Restormer · x4 SR | PSNR (dB) | Hurt |
| use SwinIR (plain architecture, local attention) for global-context tasks | SwinIR · Dehazing SOTS Indoor | PSNR (dB): 41.87 -> 29.14 | Hurt |
| train one all-in-one model across five tasks with a more general backbone | X-Restormer · All-in-one restoration | PSNR (dB) | 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 9 of 10 shared results
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
- Dual Residual Networks Leveraging the Potential of Paired Operations for Image Restoration (CVPR 2019 · similarity 0.90)
- Deep Generalized Unfolding Networks for Image Restoration (CVPR 2022 · similarity 0.89)
- Revitalizing Convolutional Network for Image Restoration (TPAMI 2024 · similarity 0.89)
- Training Your Image Restoration Network Better with Random Weight Network as Optimization Function (NeurIPS 2023 · similarity 0.88)
- A Restoration Network as an Implicit Prior (ICLR 2024 · similarity 0.86)
- Debiased All-in-one Image Restoration with Task Uncertainty Regularization (AAAI 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.
