CMSC 35500
Computer Vision (STAT 37900)Prerequisites: Consent of instructor.
Catalog Description: This course covers deformable models for detecting objects in images. Topics include one-dimensional models to identify object contours and boundaries; two-dimensional models for image matching; and sparse models for efficient detection of objects in complex scenes. Mathematical tools needed to define the models and associated algorithms are developed. Applications include detecting contours in medical images, matching brains, and detecting faces in images. Neural network implementations of some off the algorithms are presented, and connections to the functions of the biological visual system are discussed.
Instructors: Yali AmitQuarter offered: Winter
Last Verified by Sharon Salveter on 8 April, 2003.