Issues in CogSci: Oliver Layton, "Using Dynamical Systems to Model Human Heading Perception"


Issues in CogSci: Oliver Layton, "Using Dynamical Systems to Model Human Heading Perception"

Sage 4101

March 26, 2014 12:00 PM - 1:30 PM

Using Dynamical Systems to Model Human Heading Perception

How do humans autonomously navigate about the ever changing world? During navigation, the brain must continuously adapt to varying environmental conditions. For example, an independently moving object may require the human observer to adjust the intended path of travel to prevent collision. A successful model that describes how humans perceive where they are heading should then fundamentally address how the temporal unfolding of new visual information impacts perception. Despite the abundance of approaches to mathematically model the primate brain, few simultaneously incorporate temporal dynamics, make mechanistic predictions about how populations of neurons interact to give rise to behavior, link the activity of neurons to our perceptual experience, and quantitatively simulate cell recordings from neurophysiology. I will discuss a dynamical systems approach to neural modeling that considers navigation, and other behavioral and perceptual phenomena, as emergent outcomes of dynamical interactions between populations of neurons in different areas of the primate brain. I will first descibe how dynamical systems can provide mechanistic insight on a well-known visual illusion, and then show how I have used this approach to model human heading perception in the presence of independently moving objects. The model proposes the neural mechanisms that underlie why humans systematically misperceive their heading direction when a moving object crosses their path.


Access the paper here. 

Add to calendar