Manu Raghavan, Ph.D.

Manu Raghavan is a fifth year postdoctoral researcher working with Tony Movshon at NYU. His undergraduate background is in biology and psychology with some philosophy and engineering mixed in. He completed his Ph.D. at Duke University studying how monkeys execute eye movements to track and fixate on objects in their environment. He currently studies how brain circuits from the retina through the rest of the brain analyze motion and form. At Duke, Manu builds models of visual processing to study how human subjects perceive images. He records the activity of large numbers of neurons throughout the monkey visual system in response to those same images to try to link activity in the brain to perception. Manu has mentored several undergraduates who have gone on to do everything from pursuing a Ph.D. in neuroscience, to going on to medical school, to working at companies in the area. He hopes he can help the fellow achieve their career goals.

Principal Investigator: Tony Movshon

Fellow: Vasila Abdumuminova and Sofiya Eliachova

Project
Visual search and object recognition in natural scenes
The goal of this project is to generate behavioral benchmarks that can be used to compare the performance of artificial neural networks that model core object recognition to human performance. The comparison will take place within the Simons-funded Brain-Score framework. Concretely the project will involve the design and collection of data from human subjects as they detect/search for targets ranging from simple sinusoidal gratings to objects (like faces or animals) placed at various locations in natural images. Analysis of this performance will be uploaded to Brain-Score, where performance on these tasks can be automatically compared to the performance of dozens of artificial neural networks on the same task. Analysis will be performed using the Python programming language (no prerequisite knowledge of programming required). Depending on your interest and experience we can make the project more or less computational.