Automated Design of Salient Object Detection Algorithms with Brain Programming

Despite recent improvements in computer vision, artificial visual systems' design is still daunting since an explanation of computing algorithms remains elusive. Salient object detection one problem that open due to the difficulty understanding brain's inner workings. Progress on this research area follows traditional path hand-made designs using neuroscience knowledge. In years two different approaches based genetic programming appear enhance their technique. One idea combining previous methods through and fuzzy logic. The other approach consists improving computational structures basic models evolution. This work proposes expanding dorsal stream a proposal solve salient problems. uses benefits main aspects area: fixation prediction objects. We decided apply fusion saliency image segmentation as template. proposed methodology discovers several critical template present results benchmark designed by experts with outstanding comparison state-of-the-art.