ON THIS PAGE:

Course Objectives

Prerequisites

Approach

Course Outline

The topic for COMP 785 varies from year to year. This document is a sample syllabus, from the Spring 1995 offering. The offerings have been: Neural Networks and Vision (Spring 1991, 1993, 1995); Neural Networks and Control of Action (Spring 1992); and Neural Networks and Pattern Recognition (Spring 1994). Please contact the instructor for details about the next offering.

Course Objectives
This introductory graduate seminar will investigate the neural mechanisms and circuitry underlying visual perception. The seminar will have a special focus on how visual systems represent 2D and 3D space, on the mechanisms of visual attention, and on visual plasticity and learning.

Prerequisites
The seminar is at an introductory level. Prior exposure to material on neural networks or vision is not a prerequisite, but would likely be helpful. The seminar is open both to students who have not studied neural networks or vision and to students who have. Graduate students in other departments, as well as those in computer science, are encouraged to take the seminar. Faculty members and postdocs are also welcome. The seminar is also open to students who have previously taken my Neural Networks and Vision course (1991/1993), since the material covered will be about 75% new.

Approach
Because self-organizing neural networks are intended to correspond with how real brains process information, the seminar will be of interest to cognitive psychologists and neurobiologists, as well as computer scientists and engineers.

Introductory information will be provided on neural networks and on vision. Introductory topics include:

• basic structure of animal visual systems;
• basic introduction to neural networks;
• how to read papers in neurophysiology, psychophysics, and neural networks;
• computational aspects of vision;
• computational behavior of neural networks;
• edge detection, orientation detection;
• perception of visual motion, depth, transparency, and occlusion;
• visual segmentation, boundaries, and grouping;
• learning, adaptation, and self-organization.

Later topics will include:

• attentional focusing on selected features or spatial regions;
• converting visual data from self-centered to world-centered maps;
• behavioral aspects of attention in human and animal vision;
• attention and space in the hippocampus and visual cortex;
• representing invariant relationships of objects to one another;
• grouping operations (establishing linkages among perceptually related data);
• scission operations (breaking linkages between unrelated data);
• "filling-in" or diffusion of color, motion, depth, and surface data across image regions.

Neural network modeling is concerned with understanding both how neurons in animal brains communicate and perform tasks for perception, cognition, and motor control, and how networks of artificial neuron-like processing elements solve computational problems. Neural network research is highly interdisciplinary: computational theory helps explain and predict findings in neurobiology and psychophysics, while neurobiological and psychological results suggest new methods for solving computational problems. The visual systems of vertebrates are a microcosm of the full brain--nearly all the complexity and function is present.

Workload
The seminar will cover selected papers from the recent literature. Workload will consist of readings, class presentation(s), class participation, writing short critiques, and either a research project/paper or a take-home final exam.

The class meetings will consist mainly of focused discussions on the readings and on related issues. This will not be a lecture course, although the instructor will give occasional mini-lectures. There may be occasional in-class assignments.

The course emphasizes critical and creative thinking. That means that you are expected to search for conceptual flaws in the reading material and in discussions and to propose your own ideas for improvements.

Course Outline
Numbers in parentheses indicate approximate number of weeks

• Intro to neurophysiology & psychophysics (1.5)
• Receptive field plasticity (1.0)
• Illusory contours (1.0)
• Surface perception and filling-in (2.0)
• Stereopsis (1.0)
• Apparent motion (1.5)
• Smooth motion (1.0)
• Image stabilization and binding (0.5)
• Visual attention (2.5)
• The hippocampus (1.5)
• Object recognition (0.5)