(This is a living document)
Last update: 10 July 2017
These are some questions that receive fairly regularly by email, with frequently given answers.
If you have taken one of my classes or done research with me, then yes I would be happy to write you a letter.
I require at least 3-4 weeks notice before the letter is due. When you ask me for a letter (by email), you should provide:
Point 1 is particularly important. Before you ask me to write a letter, you should think about what kind of letter I will be able to write for you. Creating this lists helps you think critically about whether I'm the right person for the job, and it will help me write a stronger letter. Did you actively participate in class discussions, come to office hours regularly, go above and beyond in your programming assignments, do something notable for a class project? If you cannot find something notable to mention then it is unlikely that I will be able to. A letter which simply says "this student got a B+ in my class" is not useful since it just repeats what's in your transcript — you would probably be better served by asking someone else for a letter in that case. When assembling your list keep in mind that I will only write about things that I have first-hand knowledge of/experience with.
Yes! Maybe! Since rendering research has a pretty steep learning curve, you should already be familiar with the foundations of this area. You should be a proficient programmer (most importantly in C++, but research prototyping also sometimes involves Matlab, Python, Mathematica, etc.) and feel comfortable with Calculus and Linear Algebra. If you have not already, you should sign up for my Computer Graphics, Rendering Algorithms, and possibly Computational Photography classes. Most research projects will build on the foundations established in these courses, so good performance will put you at a great advantage. You can also consider joining the weekly VCLunch reading group (talk to me or one of the VCL members to learn more).
When you know you'd like to do research with me, you should email me well before the term you want to take the thesis. Typically we will meet face-to-face to discuss possible project options. You should think about what research of mine you are most drawn to, which will help guide the discussion and help me think of appropriate projects to carve out for your research.
Probably not (unless I'm already collaborating with you or your advisor). Generally speaking, US schools don't tend to have official "internship" programs. I will occasionally host visiting scholars/students, but these arise due to an established research relationship/project need. The best way to establish a collaboration is to approach me about research project ideas (like at a conference!). If this leads to an exciting discussion with a concrete research direction, then it becomes much more interesting to host a visit.
Occasionally I may make an exception for truly exceptional students who are already doing interesting work, or who are supervised by researchers who I'm interested in collaborating with. Even in this case, first establishing a collaboration/research project is probably the right way to get the ball rolling. You may also consider having your supervisor contact me on your behalf.
Maybe! Assuming I have funding available, I am open to accepting strong, self-motivated students with a passion for graphics research. However, admission decisions are not entirely up to me but are made by a committee whose members are different each year. Instead of contacting me directly you should instead visit this webpage and submit an application. You can email me once you have applied to let me know to keep an eye out for your application, but emails simply asking "can I be your PhD student?" will either be deleted without reply or directed to this FAQ.
Convince me that you have the work ethic, background knowledge, and skills to immediately start hammering away at interesting research problems in graphics. You should have taken graphics and/or rendering classes in your undergrad/Masters institution. Getting letters of recommendations from senior researchers whom I know and trust is very helpful. Demonstrate (through class or open-source projects) that you can digest and reproduce recent graphics research (from e.g. SIGGRAPH, EGSR, etc.). Implementing one of my research papers will get you noticed.
Maybe. I enjoy collaborating with post-docs, but they are expensive, so positions are limited. It depends on a combination of your experience and my current funding.
I expect post-docs to have already established themselves as top researchers in the field. During your PhD or previous post-docs you should have already done some high-quality research and published it at top venues. I expect to see at least one or two papers at SIGGRAPH (Asia), TOG, EG, or EGSR. Ideally these would be papers that I'm already familiar with and thought were interesting. If I'm not already familiar with your research the chances of a post-doc are much lower.
If you do not already know me personally, the best way to get the ball rolling is probably to have your advisor introduce you or you could approach me directly at a conference. As with the comments on visiting scholars above, the best way to approach this is to start a discussion about research ideas. You should have a one-to-two year research agenda in mind.
Even if I want to hire you, in nearly all circumstances, I will ask you to apply for whatever fellowships or postdoc funding you might qualify for (I can help you apply). Some potential options are the NSF Postdoctoral Fellowship or the Computing Innovation Fellowship. There is also the Neukom Fellows program at Dartmouth, and many other options specific to particular countries of origin or sponsored by private organizations.