We designed a uncertainty guided deep Mastering system (UGLS) to Increase the performance of existing segmentation neural networks and validated it according to the classical U-Web by segmenting the OC from shade fundus visuals as well as remaining and suitable lungs from Xray visuals. The novelty o