views:

133

answers:

2

What are the best known labs for research in CG especially raytracing/rendering? I want to pursue Masters/PhD in this field and though I have been working on projects such as raytracers, it is all self-driven. What would be a good place to start, perhaps as a research intern? Also, what kind of background is expected for admissions in these programs, in terms of programming skills, formal education etc.?

+2  A: 

I'd say that getting a job at Pixar or some other computer graphics factory might be a better place to start than academia.

But if you must go get a degree, UC Davis has some active stuff going on.

Looking at up to date conference proceedings like SIGGRAPH 2010 might tell you something about what's being done and who's doing it.

You have to understand the mathematics of computer graphics: linear algebra, trig, wavelets, Fourier transforms, etc.

duffymo
"Seems done" sounds a lot like "Everything that can be invented, has been invented".Raytracing seemed done until Radiosity arrived. Then it was done, until somebody invented Photon tracing. Then it was done until somebody at PDI tried to render milk. Then it was done again, until GPUs came along. See a pattern? There are *plenty* of problems with raytracing begging for clever solutions. Eg. tracing through non-homogenous translucent media via subsurface scattering is still an active research field. Somebody who came up with some analytic accelerations would be pretty popular.
kibibu
Thank you for the correction, kibibu. I'll amend my answer to reflect your comment.
duffymo
Thanks kibibu- my interests lie in these seemingly super-specialized areas of raytracing. I belong to the school of thought who are ever hopeful raytracing will trump rasterization soon.Unfortunately such specific taste makes it difficult to find universities which would be interested in me. As for companies like Pixar, you are expected to have several years of experience or a MS/PhD in the field.
Shrey
Larry Gritz
@Larry - thank you, I'll correct it.
duffymo
+5  A: 

In the spirit of "teach you to fish" rather than "give you a fish", you really need to find out who is doing work in your area. The gold standard for publishing cutting edge graphics research is the annual SIGGRAPH conference -- there are two now, one in North America, one in Asia, both are published as editions of the ACM's journal Transactions on Graphics (TOG). There are also some other specialized conferences such as High Performance Graphics (HPG), which is a combination of the older Graphics Hardware (GH) and Interactive Ray Tracing (IRT) conferences.

The best thing to do is to look at the list of papers last couple years of these journals and conferences, and see the affiliations of the authors. There's really no better way to figure out which schools are doing work in which areas. It's infinitely more useful and accurate than listening to people here (often who don't work in the field) list specific schools).

This web site is a fantastic list of the best conferences in the field, and conveniently lists the papers, authors, and affiliations. Quickly scanning through the last couple years' of the best conferences, you will see which schools tend to be doing a lot of work in each area.

For the lazy, top schools for computer graphics research tend to be Stanford, University of Utah, University of North Carolina, Cornell, U. Washington, UC Berkeley, UIUC, Brown. But there are many others, as well, and different schools specialize in different parts of the field, so you should be trying to identify your particular interests, then go to the schools that are doing active work in that area.

Once you have identified a school doing work that interests you, these days all the info you need is online -- their admission requirements, course catalogs, syllabi, and prerequisites expected for the courses you may be interested in. Also, you may also have noticed that the schools I mentioned as being active in computer graphics tend to be among the top-rated CS academic departments generally, so you need to have a fairly beefy math and CS background and excellent undergraduate performance in order to get accepted in the very competitive admissions.

Larry Gritz
That's a great link, thank you!I really like the research at UNC-Chapel Hill, Stanford and Utah. I majored in Electrical Engineering though.
Shrey
An EE background should not be a hinderance, provided you have distinguished yourself therein. Lots of people in graphics started out in other fields (math, EE, physics, etc.). In fact, hardware graphics is very important nowadays, and obviously it's an amalgam of EE (design of the hardware itself) and CS (the graphics algorithms it embodies).
Larry Gritz