voula@utcsrgv.UUCP (Voula Vanneli) (02/11/85)
UNIVERSITY OF TORONTO
DEPARTMENT OF COMPUTER SCIENCE
(SF = Sandford Fleming Building, 10 King's College Road)
(MC = McLennan Physical Laboratories, 60 St. George Street)
(RS = Rosebrugh Building, Taddlecreek Road)
GRAPHICS SEMINAR - Friday, February 22, 11 a.m., RS 211
Dr. Robert P. Futrelle
Department of Genetics and Development
University of Illinois at Urbana-Champaign
"Vision, Knowledge, and Natural Language in Graphics" *
Computer graphics has made tremendous strides as a
means of creating artifical worlds and as a convivial system
for design. But these uses of graphics are generative and
depend strongly on user input. There is a huge and impor-
tant use for graphics which is little worked on today -- the
use of graphics to integrate knowledge. In this mode graph-
ics is integrated with world knowledge and natural language.
Specifically, graphics would use image analysis (computer
vision) to build its own knowledge base. Natural language
would be tied tightly to the graphics knowledge to build a
coordinated whole. Natural language would also be a part of
the interface to the human user. It is our view that the
knowledge and reasoning structures underlying vision and
natural language are the same -- they involve the same
models of how the world functions. We will illustrate some
of these points by discussing a long-term project for
machine understanding of scientific text and graphics.