Dehaze-RL: Physics-Guided Reinforcement Learning (ECCV 2026) by Liu ChangDehaze-RL: Physics-Guided Reinforcement Learning (ECCV 2026) by Liu Chang

Dehaze-RL: Physics-Guided Reinforcement Learning (ECCV 2026)

Portfolio project
Liu Chang
Motivation & Challenge: Unpaired dehazing needs reliable restoration without costly diffusion sampling. Dehaze-RL learns physical parameters, enhancement factors, and spatial gating, guided by perceptual rewards. Results & Venues: Improved restoration quality with a physics-grounded hybrid action space and multi-granularity rewards. Accepted at ECCV 2026. Link: https://media.eventhosts.cc/Conferences/ECCV2026/pdfs/3070.pdf

1 like · 0 views · Sep 15, 2026

Motivation & Challenge: Unpaired dehazing needs reliable restoration without costly diffusion sampling.

Tags