A new framework called P2GCL has been developed to jointly improve the estimation of atmospheric turbulence strength and the restoration of infrared video. This framework uses a cooperative learning approach where one model estimates turbulence strength and provides this information as a physical prior to a second model that restores the infrared images. The restored images are then fed back to the first model to refine its turbulence strength measurements, creating a cyclic collaboration. Experiments show that this method significantly improves both turbulence strength estimation and image restoration quality. AI
IMPACT This research introduces a novel cooperative learning framework that could improve the quality of infrared imaging in challenging atmospheric conditions.
RANK_REASON This is a research paper describing a new framework and experimental results. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
