Computational Basis of Visual Motion Detection in Insects

TIME: 3:30 PM

LOCATION: GMCS 314

SPEAKER: Patrick Shoemaker, Computational Science Research Center, San Diego State University

ABSTRACT:

I describe computational models for how insects detect various classes of motion observed in their visual fields, based on temporally-varying signals from their photoreceptors. Early work was aimed at modeling a fundamental operation involving signals from neighboring ‘pixels’, which gives a form of motion sensitivity, and subsequent work describes how large neurons in the deeper brain might integrate signals from such elementary detectors to achieve sensitivity to various patterns of motion across their visual fields. I then summarize the area of my own research: the visual detection of small moving targets by animals that chase prey or mates on the wing. A model for an elementary detector sensitive to small targets is described, as well as a form of predictive facilitation in target detection that is a subject of current research.

HOST: Jose Castillo

VIDEO: