AI Paper Atlas

Research map / Image generation / Image restoration

MOERL: When Mixture-of-Experts Meet Reinforcement Learning for Adverse Weather Image Restoration

Tao Wang, Peiwen Xia, Bo Li, Peng-Tao Jiang, Zhe Kong, Kaihao Zhang, Tong Lu, Wenhan Luo

ICCV 2025 · Image generation › Image restoration · approach: Efficient restoration architectures · 7 citations (OpenAlex)

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 changedWhereMeasuredEffect
add MOE layer (ablation)residual-block baseline · Outdoor-RainPSNR (dB): 28.89 -> 31.48Improved
add Transformer layer (ablation)residual-block baseline · Outdoor-RainPSNR (dB): 28.89 -> 30.21Improved
add Transformer layer to MOE layer (ablation)MOERL network · Outdoor-RainPSNR (dB): 31.48 -> 32.70Improved
add channel experts to spatial experts (ablation)MOE layer · Outdoor-RainPSNR (dB): 32.20 -> 32.70Improved
add spatial experts to channel experts (ablation)MOE layer · Outdoor-RainPSNR (dB): 32.03 -> 32.70Improved
add routing prompt module (ablation)MOERL · Outdoor-RainPSNR (dB): 31.12 -> 32.79Improved
add RL training (ablation)MOERL · Outdoor-RainPSNR (dB): 32.40 -> 32.70Improved
add critic network (ablation)RL strategy · Outdoor-RainPSNR (dB): 32.54 -> 32.70Improved
add quality prompt to reward model (ablation)Q-Align reward · Outdoor-RainPSNR (dB): 32.68 -> 32.70Improved
add degradation prompt to reward model (ablation)Q-Align reward · Outdoor-RainPSNR (dB): 32.62 -> 32.70Improved
replace LLaVA with Q-Align reward (ablation)MOERL RL · average of 4 setsPSNR (dB): 33.72 -> 34.23Improved
replace InternVL with Q-Align reward (ablation)MOERL RL · average of 4 setsPSNR (dB): 33.90 -> 34.23Improved

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.

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