AI Paper Atlas

Research map / Label-efficient and robust learning / Test-time adaptation

Towards Multi-Domain Single Image Dehazing via Test-Time Training

Huan Liu, Zijun Wu, Liangyan Li, Sadaf Salehkalaibar, Jun Chen, Keyan Wang

CVPR 2022 · Label-efficient and robust learning › Test-time adaptation · approach: Test-time adaptation of vision-language models · 56 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
Apply proposed test-time training to multi-domain GDNGDN · RESIDE indoor SOTSPSNR: 26.67 -> 26.83Improved
Apply proposed test-time training to multi-domain GDNGDN · RESIDE outdoor SOTSPSNR: 27.18 -> 27.26Improved
Apply proposed test-time training to multi-domain GDNGDN · O-HazePSNR: 23.13 -> 23.21Improved
Apply proposed test-time training to multi-domain GDNGDN · NH-HazePSNR: 18.93 -> 19.04Improved
Use helper-network test-time supervision without meta-learning (ablation)GDN · RESIDE indoor SOTSPSNR: 26.67 -> 26.69Improved
Use helper-network test-time supervision without meta-learning (ablation)GDN · RESIDE outdoor SOTSPSNR: 27.18 -> 27.16Hurt
Use helper-network test-time supervision without meta-learning (ablation)GDN · O-HazePSNR: 23.13 -> 22.89Hurt
Use helper-network test-time supervision without meta-learning (ablation)GDN · NH-HazePSNR: 18.93 -> 18.65Hurt
Jointly train GDN with dehazing and haze-reconstruction losses (ablation)GDN + Helper · RESIDE indoor SOTSPSNR: 26.67 -> 26.71Improved
Jointly train GDN with dehazing and haze-reconstruction losses (ablation)GDN + Helper · RESIDE outdoor SOTSPSNR: 27.18 -> 27.19Improved
Jointly train GDN with dehazing and haze-reconstruction losses (ablation)GDN + Helper · O-HazePSNR: 23.13 -> 22.94Hurt
Jointly train GDN with dehazing and haze-reconstruction losses (ablation)GDN + Helper · NH-HazePSNR: 18.93 -> 18.71Hurt

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