Development of a human visual system with the ability to detect inconsistent events

Publication Year:
2004
Usage 13
Abstract Views 10
Downloads 3
Repository URL:
http://scholarworks.rit.edu/theses/6720
Author(s):
Brazeau, Jeremiah D.
thesis / dissertation description
Current research has shown that it is possible to create a general purpose learning system that models the first few layers of the human visual system. The goal of this thesis is to build upon that idea and produce a computational model capable of learning different aspects of visual information. Simulations show that the system is capable of learning and distinguishing different types of motion. Results also give one explanation, consistent with current experiments, explaining how the human visual system learns information and as such may produce the capability to predict behavior in future experiments.