Bates College

Professor Greene's primary research goal is to understand the mechanisms that enable rapid, intelligent perception of our environment. Towards this end, she uses computational methods from machine learning and computer vision to model the information that may be used by the human brain for visual understanding. This information can come from the scene’s objects, from global layout, from knowledge of the actions afforded by an environment, or from other prior visual experience. Professor Greene then compares these models to human performance, either in the laboratory using well-established psychophysical measures, in neural activity or eye-movement patterns, or by using crowdsourcing to collect human-generated data at scale. This enhanced toolkit has allowed her work to make an impact beyond psychology, generating follow-up work in computer vision, as well as in applications ranging from fingerprint identification to pedagogy. She examines visual perception from two directions: a bottom-up approach of measuring and manipulating the contents in the scene, and a top-down approach of exploring the internal representations and knowledge that influence how a scene is understood.