Home Department: Computer Science
Mentors: Keren Haroush (Neurobiology) and Shaul Druckmann (Neurobiology and Psychiatry & Behavioral Sciences)

“Applying Deep Learning to Social Gaze Patterns in Competitive and Cooperative Decision-Making”

Among rhesus macaques, the gold standard of neurophysiology research, eye contact is interpreted as an aggressive cue and is unlikely to be utilized in prosocial interactions. However, other primates such as humans and marmosets can use direct gaze prosocially. In this project, Melinda is designing various tasks involving collaborative and competitive behaviors among marmosets, which will be recorded and used for the refinement of machine vision algorithms that can track marmosets’ gazes. By perturbing activity in areas of the social brain and other network nodes, these algorithms will lead to understanding how neuronal changes in primates affect social gaze patterns.