Prof. David McLaughlin, Provost Emeritus, NYU; Silver Professor of Mathematics and Neuroscience; Courant Institute, Tandon School of Engineering, NYU Shanghai; Chief Science Mentor, NYU Shanghai; Associate Investigator, Neuroscience Institute at NYU Langone Medical Center
lecture hall, 2rd floor of Tsung-Dao Lee Institute (east of Pao Yue-Kong Library)
In this lecture, I will use our work in visual neural science to illustrate the potential that large-scale computational modeling presents to neural science today. Neural science is primarily an experimental science, with major advances following closely upon advances in experimental technology. Similarly, advances in computational technology over the past two decades have positioned computational scientists to contribute to the theoretical understanding of neuronal systems. For some time now, our group at NYU has been developing a large-scale computational representation of an input layer of the primary visual cortex (V1) of Macaque monkey – the “front end” of the monkey’s visual system. Neurons in V1 are “edge detectors” – detecting the orientation of edges within the visual scene. While monkey V1 is tiled by an ordered map of orientation preference, mouse V1 is tiled by a disordered “salt and pepper” map. In recent work, we have adapted our monkey V1 model to mouse V1, and have studied the differences in neuronal response in the presence, and the absence, of an ordered map of orientation preference. Our mouse model reproduces laboratory observations, and allows us to analyze the mechanisms by which the model achieves the observed responses – with its disordered map of orientation preference.