          |
GRASP Seminar Series: Fall 2005
November 11, 11:00 AM, 307 Levine Hall
Pietro Perona
California Institute of Technology
"Towards Visual Serendipity"
Abstract: A field scientist visiting a new environment will carefully take note of the presence, frequency and location of plant and animal species, landscape features, human physiognomies, behaviors and customs. We are beginning to understand how to replicated this ability: given a training set of images collected by an expert, our machines can now learn `visual' models of categories. They can subsequently detect and catalogue these categories in new environments. Good scientists, however, will also be able to spot anomalies and identify things that they were not trained to recognize. These are often the most valuable pieces of information. Can we hope to build machines that can do the same? I will discuss a statistical approach to visual recognition, describe the state of the art and speculate on realistic goals for the next decade.
Biography: Pietro Perona studies the computational aspects
of vision; his current focus is visual recognition. He has published on
applications of PDEs to image segmentation, human texture perception and
segmentation, dynamic vision, grouping, perception of human motion, learning
and recognition of object categories, categorization of scenes in human
vision, human perception of 3D shape, interaction of attention and recognition.
Perona is Professor of Electrical Engineering and of Computation and Neural
Systems at the California Institute of Technology (Caltech). He is the
Director of the National Science Foundation Engineering Research Center
in Neuromorphic Systems Engineering at Caltech.
GRASP
Laboratory
Site maintained by graspadm@grasp.cis.upenn.edu
Last update:
23 August, 2005