AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

TIR-Agent: Training an Explorative and Efficient Agent for Image Restoration

Yisheng Zhang, Guoli Jia, Haote Hu, Shanxu Zhao, Kaikai Zhao, Long Sun, Xinwei Long, Kai Tian, Che Jiang, Zhaoxiang Liu, Kai Wang, Shiguo Lian and 2 more

ECCV 2026 · Reinforcement learning › Visual and representation RL · approach: RL with tree search for discovery · 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 SFT stage before RL (ablation)TIR-Agent · MiO100 Group APSNR: 19.49 -> 22.07Improved
add RL stage after SFT (ablation)TIR-Agent · MiO100 Group AMUSIQ: 60.17 -> 63.69Improved
add exploration-driven perturbation (EDP) to SFT data (ablation)TIR-Agent · MiO100 Group APSNR: 19.97 -> 22.07Improved
add task-order perturbation alpha^t (ablation)TIR-Agent · MiO100 Group APSNR: 20.61 -> 22.07Improved
add tool-level perturbation alpha^m (ablation)TIR-Agent · MiO100 Group ALPIPS: 0.3256 -> 0.2874Improved
add multi-dimensional adaptive reward (MAR) (ablation)TIR-Agent · MiO100 Group APSNR: 20.85 -> 22.07Improved
add MAR (ablation)TIR-Agent · MiO100 Group ALPIPS: 0.2830 -> 0.2874Hurt
add decoupled per-metric advantage normalization (ablation)TIR-Agent · MiO100 Group APSNR: 21.86 -> 22.07Improved
add adaptive metric weights omega^theta (ablation)TIR-Agent · MiO100 Group APSNR: 20.51 -> 22.07Improved
increase task-order perturbation probability alpha^t above 0.3 (ablation)TIR-Agent · MiO100 Group APSNRHurt
increase tool perturbation alpha^m (ablation)TIR-Agent · MiO100 Group APSNRNo clear effect
train agent with SFT+RL instead of training-free tool traversalTIR-Agent vs AgenticIR · MiO100 Group AInference time (s): 184.57 -> 61.58Improved

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