Researchers have developed a novel semi-supervised framework for removing nighttime lens flares from images, addressing the challenge of limited paired data. The proposed method, Semi-LAR, enhances learning from unlabeled images by refining pseudo-labels through quality assessment and filtering, and by employing a flare-aware contrastive loss. This approach encourages representations that are discriminative against flare patterns while remaining consistent with reliable targets, demonstrating improved performance and robustness across various benchmarks. AI
Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →
IMPACT Introduces a new technique for image artifact removal, potentially improving visual quality in AI-generated or captured imagery.
RANK_REASON The cluster contains an academic paper detailing a new method for image processing. [lever_c_demoted from research: ic=1 ai=1.0]