COMPUTATIONAL MODELING OF VISUAL MOTION PERCEPTION

Carnegie-Mellon University, Pittsburgh PA

Investigators

Linked publications & trials

Abstract

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The goal of our research is to understand how the biological mechanisms involved in vision work. Currently, we have been modeling visual motion sensing mechanisms that are believed to exist in the primary visual cortex. Although modeling neural networks is not a new concept, we have been building biologically realistic models based on neural physiology that may answer questions that simpler models cannot. We believe that secondary effects in biological neural networks are important to their function. Conventional neural network models like the integrate-and-fire model do not model these effects. However, with our models, based on Hodgkin-Huxley equations and simulated with the General Neural Simulation System (GENESIS), we have made considerable progress in understanding these effects. For example, with GENESIS we are able to simulate the voltage pulses that neurons use to communicate with other neurons. However, we can also observe cell membrane voltages. These voltages might be one example of a secondary effect that will enable us to understand how motion is perceived in the brain. In order to test our theories, we need to build neural network models that are large enough to exhibit behaviors that can be confirmed with psychophysical experiments. We plan to use Parallel GENESIS (PGENESIS) for this stage of our project. Since our current simulations consume most of our resources now, we know that the more complex simulations must be performed by a much more powerful machine. One of the reasons we are interested in simulating these models at the Pittsburgh Supercomputing Center is because it is where PGENESIS was developed and is well supported.