Why study Graphics, Vision and VR?
My answers to an interview done by a cmput101 student.
Hi,
I am a CMPT 101 student at the U of A. I am working on a presentation
project for this course. My topic is "If I was a Computer Scientist,
I'd work in Graphics and Virtual Reality." I was wondering if you
might have time to answer a couple questions to further my
understanding of this topic?
If you could answer some of these questions I'd really appreciate it,
thanks.
Corinne
I'd be happy to answer your questions. There's more information
on my web site. If you like to know more feel free to ask or
come to visit.
Cheers,
Martin
Here are my questions:
What exactly do you do in this field?
Mainly research into how to use 2D images (from regular cameras)
to measure 3D properties or reconstruct a whole 3D geometry
from these images alone. See my web pages for more information.
This research used to be thought of as computer vision but is
now becoming more and more important to graphics and visualization.
What attracted you to this line of work?
This research is mathematically interesting (and challenging
as are most "inverse" problems -- Just rendering images is
a forward problem and much easier). Yet (unlike some mathematics)
there are very direct applications in creating
What do you enjoy about it?
Certainly to some extent the visual results, but also that there
is a particularly strong foundations in math and the physics of
imaging, yet there are new results to be discovered.
What sort of background or qualifications are required?
This depends on if one is interested in fundamental research or
applications. For applications some mathematical background combined
with talent in visual arts and presentations.
For fundamental research a strong background in linear algebra,
calculus and numerical analysis is needed. Furthermore additional
courses in geometry, optimization and differential equations
are helpful.
Are there any commercial benefits or purely academic challenges with
this job?
Microsoft has been hiring some of the very best researchers over
the past several years with very good salaries, and have large
groups in computer vision both in Redmond and Cambridge. Likewise
there are several companies working with computer vision for
special effects (e.g. match moving) in movies.
However, as the knowledge needed is still somewhat esoteric and
quite advanced, most people have graduate degrees whether they
work in industry or academia.
Martin