Computational Perception and Scene Analysis

15-485/785 and 85-485/785

Spring semester, 2004

Instructor:

Mike Lewicki
Computer Science Department and
Center for the Neural Basis of Cognition
Carnegie Mellon University

Units: 9 (undergraduate) or 12 (graduate)

Days/Times: Tuesdays and Thursdays, 3:00 - 4:20

Location: 200 OSC (Old Student Center, 4902 Forbes Ave.)

Course Description

This course teaches advanced aspects of perception, scene analysis, and recognition in both the visual and auditory modalities, concentrating on the essential computational processes that allow us and animals to behave in natural, complex environments. The goal of this course is to teach how to reason scientifically about problems and issues in perception and scene analysis, how to extract the essential computational properties of those abstract ideas, and finally how to convert these into explicit mathematical models and computational algorithms.  Throughout the course we examine important findings from perceptual science that have provided key insights into the function and organization of human and animal perceptual systems.

Specific topics include sensory coding, perceptual invariance, spatial vision and sound localization, visual and auditory scene segmentation, many aspects of attention, and approaches to recognition in natural visual and auditory scenes.

Mathematical topics covered include Bayesian inference, information theory, linear systems analysis, neural networks, independent component analysis, and selected algorithms in computational vision and audition.

Prerequisites: 15-385, 85-370, or permission of the instructor.

Contact:

Mike Lewicki
lewicki@cs.cmu.edu
http://www.cs.cmu.edu/~lewicki

Office: Mellon Institute 115K, x8-3921

Other links: