2 Dynamics in Perception

Overview

In this chapter we will explore dynamic models using some examples from perception. We’ll start with introducing Hebbian Learning, then introduce a Hopfield Network (Hopfield, 1982) that takes noisy input and settles on a stable interpretation, then we’ll explore some simple examples of dynamic networks evolving over time, and finally consider bistable percepts that can be interpreted multiple ways. Along the way we’ll cover numerous examples and concepts that recur and are useful throughout this course.

Content

2.1 Hebbian Learning

2.2 Hopfield Networks

2.3 Dynamic Systems and Bistable Perception

Slides