The aim of this project is to shed light on whether the partial blindness of
infants is helpful to the development of visual capabilities.
Melissa Dominguez (Ph. D. Thesis Work)
Robert Jacobs (Thesis Advisor)
Melissa Dominguez
The noise present in the top picture makes matching stereo images
difficult. However, the blurring in the bottom figure makes the
problem much easier because only the features important for matching
survive the blurring process.
One of the most intriguing aspects of the human brain is that its
abilities develop tremendously over the course of the life of its
owner. A human infant is with extremely limited perceptual abilities,
yet within a few years he has developed into a person with complex
abilities far beyond those of the most advanced artificial
intelligence program.
We propose that studying the developmental process of the human visual
system will provide useful insights for developing better learned
computer vision systems. We present a machine learning system which
applies theories from developmental psychology to the computer vision
problem of stereopsis, and discuss further applications where human
developmental data would be helpful in solving perceptual artificial
intelligence problems.
http://www.cs.rochester.edu/~melissad/Research.html
Developmental Vision, Binocular Disparity, Motion Velocity
Sensitivity, Computer Vision, Melissa Dominguez, Robert Jacobs,
University of Rochester Computer Science Department