2013
Nicholas Baker
I am a rising senior at Johns Hopkins University, studying cognitive science. I am interested in vision, and how studying computers can inform us about how people compute meaning from sensory input. I am also interested in how algorithms from computer vision can be inspired from what heuristics we find to be effective from studying human vision. My current project is concerned with visual saliency, which is thought about in two distinct components. Bottom-up, data driven visual saliency seeks to identify those parts of a scene that stand out due to the distinctness of their features compared to the surrounding scene. The top-down component of visual saliency is task driven. In top-down processes, exogenous information informs the way the viewer analyzes the scene, shifting the weight of what is most salient based on what the viewer knows is important. My project aims to use eye tracking data from people watching videos to study what determines salience in human viewers to improve the way computers evaluate saliency. I also hope to use this saliency data to improve object detection in visual scenes.
Weekly Presentations
Poster Presentation
Final Report
Back to 2013 REU Participants
