Recognition of Occluded Objects as Statistical Inference: A Neurophysiological Study in Awake, Behaving Monkeys

In the real world, visual objects commonly block, or occlude, each other from view. Very little is known about how the brain makes up for the missing information about an occluded object and recognizes it nonetheless. This project will help elucidate the mechanisms by which the brain recognizes occluded objects using neurophysiological methods and will determine what individual cells in the relevant brain regions do when the brain recognizes an occluded object. The study will also help characterize which tell-tale features in the visible portions of an occluded object the brain uses for this purpose, and how it uses them. For example, someone's eyes may serve as a dead give-away for them, thus the brain may use this information about the eyes to recognize that person, even when the rest of the face is occluded. Given the ubiquity of occluded objects in the real world, understanding how one perceives occluded objects is critical to understanding, in a larger sense, how the brain works. This research will also facilitate the development of better machine-vision tools, so as to help computers better recognize objects of interest in the real world. This project also has a significant educational component. For example, classroom courses and hands-on research experience opportunities will not only impart scientific knowledge, but also help students develop an active interest in pursuing careers in science and technology. Undergraduates and graduate students from the Investigator's institution as well as from many other local colleges and universities will participate in this project, providing additional opportunities for students. The results, resources, and tools generated by this project will be made publically available on the Investigator's website at http://www.georgiahealth.edu/medicine/discovery/bbdi/hegde.