AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

TransWeather: Transformer-Based Restoration of Images Degraded by Adverse Weather Conditions

Jeya Maria Jose Valanarasu, Rajeev Yasarla, Vishal M. Patel

CVPR 2022 · Architectures and efficiency › Attention and transformers · approach: Block-sparse attention kernels · 420 citations (OpenAlex)

Check 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 changedWhereMeasuredEffect
Use TransWeather instead of All-in-OneAll-in-One adverse-weather restoration · Test1 (rain+fog)PSNR: 24.71 -> 27.96Improved
Use TransWeather instead of All-in-OneAll-in-One adverse-weather restoration · Snow100K-LSSIM: 0.8820 -> 0.9308Improved
Use TransWeather instead of All-in-OneAll-in-One adverse-weather restoration · RainDrop testPSNR: 31.12 -> 28.84Hurt
Use TransWeather instead of All-in-OneAll-in-One adverse-weather restoration · RainDrop testSSIM: 0.9268 -> 0.9460Improved
Add a hierarchical encoder (ablation)Transformer Base · RainDrop testPSNR: 30.12 -> 31.62Improved
Add a hierarchical encoder (ablation)Transformer Base · RainDrop testSSIM: 0.8512 -> 0.8671Improved
Add Intra-PT blocks to the hierarchical encoder (ablation)Hierarchical encoder · RainDrop testPSNR: 31.62 -> 32.37Improved
Add Intra-PT blocks to the hierarchical encoder (ablation)Hierarchical encoder · RainDrop testSSIM: 0.8671 -> 0.9463Improved
Add learnable weather queries and a transformer decoder (ablation)Hierarchical encoder with Intra-PT · RainDrop testPSNR: 32.37 -> 34.55Improved
Add learnable weather queries and a transformer decoder (ablation)Hierarchical encoder with Intra-PT · RainDrop testSSIM: 0.9463 -> 0.9502Improved
Use TransWeather instead of DDMSNETSnow removal · Snow100K-LPSNR: 28.85 -> 28.48Hurt
Use TransWeather instead of Quan et al.Raindrop removal · RainDrop testPSNR: 31.44 -> 28.84Hurt

Later papers that compared with this method

1 later accepted paper reports TransWeather as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy