Tip, advice, guidelines, policies, rules

for Frédo Durand's students

Putting papers online


- It is extremely strongly very "beaucoup" recommended that you post your papers and theses online. It greatly improves their impact and makes it easier for colleagues in the field to access the material

- I advise to dedicate a separate web page for each paper. It makes it easier for me to link to them, you can put images and additional material. And when we are good, we can add hindsights and corrections. (Hugues Hoppe is very good with this).

- Systematically copy-paste the acknowledgment section from the paper at the end of the web page. It makes sponsors happy (NSF is asking about it for yearly reports).

- post the slides as well. Both powerpoint and pdf. It makes it easier for people to talk about and advertise your work.

- of course, include videos.

- if there is data that can help people reproduce or compare to your work, make it available (e.g. input and output images). It will make your life easier later and will avoid that you waste time when random people ask about data that you cannot localize anymore.

- If you are better than me, it's a good idea to put the bibtex

- putting code online is greatly encouraged.

Vacation

Please send me an email before you book vacations. I usually approve, but I want to know. Yes, this means before you buy the plane ticket. Yes, even if you told me orally, still send me an email. Yes, even if it's right after the Siggraph deadline and it's "natural" to take a vacation, you should still let me know.

Similarly, if one day you do not come to the lab because you are sick, or you work from home, or any other reason, please let me know by email. If you work from home, I'd appreciate email responsiveness. If you are sick, all my wishes of prompt recovery.

If you are involved in a siggraph project, it is probably not reasonable to take ten days without working around the Winter break. You're welcome to work remotely though.

I consider four to five weeks of vacation per year a reasonable amount (3/4 of this if you don't work with me during the summer). This can be modulated by effort and productivity (e.g. if you work 90 hours a week, I probably won't frown at your sixth week of vacation).

Writing and communication

I have a separate page for advice and sites about writing/communication.

In particular, look at my talk on how to write a bad paper and my notes on writing.

Going to conferences

Conferences are important opportunities to meet colleagues, get fresh ideas, and "network." Your goal should be to meet new researchers, get to know the top people in your area and have stimulating discussions. It can be intimidating to go and talk to a famous researcher, but it's well worth it. Introduce yourself, ask them about their past or current research. Ask their opinion about competing approaches, ask them about what they view as the big unsolved problems, etc. If you know somebody who knows the researcher, you can ask them to introduce you, it's perfectly normal.

As Anne Hunter would say "Pretend you are Harvard students, and for every person in the room, go and find out how they can be useful to you."

Learn to take compliments. If you have a paper at the conference, it's likely some people will positively comment on it. Do not react in a way that would indicate embarassment or unhappiness about the work, be polite, thank the person, say something like it was a fun project, or that it was frustrating at time but well worth it, etc. Showing excitment about your work will stimulate good conversation and will reflect positively on you.

If you see a talk you like, don't hesitate to talk to the speaker after their talk. People always appreciate it when you show interest. Ask questions, suggest follow-ups, tell them why you think the work is deep and innovative. Ask their opinion about the future of that area.

Avoid excessive "clustering" where you only interact with students from yoru own lab. You see them all year long, conferences should be an opportunity to meet new people.

Summer internship

They're good, just try to find an internship that might lead to a paper or that will teach you new things. In particular, it can be a great way to meet other researchers in the field and get them to appretiate you.

Another option is to teach (e.g. at MIT, Columbia, etc.)

Feedack, meeting-meeting, pre-submissions

There are bad misconceptions about feedback and the ineterst of presenting your work to the rest of the lab. I believe that this is because people look for the wrong kind of feedback at the wrong stage of a project's life.

At the beginning of a project, before it's fully scoped you can get useful help about technical solutions and scope. Is this a useful problem to solve? Should it be re-scoped to be easier, harder, more relevant, more general, etc. Does it remind someone of a similar problem, are there proven solutions that can be adapted? Is the technical approach you are planning to use well known to be flawed? Are there easier better alternatives?
Such early stages are the ideal moment to give a meeting-meeting talk.
In terms of writing/presubmission, it can be useful to write an introduction+previous work as well as a short overview of the proposed approach.

When the project is farther along, you are less likely to get as much technical feedback, because anyway, you have already solved your technical problems. In addition, things gets very specialized and few people understand the bells and whistles of what you do. But this is great time to get feedback about the exposition and demonstration of your contributions. This is the best time to write a presubmission and see if people understand what you are trying to solve, if they are convinced that it's useful, if your paper is readable by someone else than you, if you fail to separate your contributions from well-know methods. Do your demos really showcase your contributions?

Goals of grad school

Get tho know a research area in depth and breadth

find a research problem

solve it

expose the solution, in writing and orally

After grad school

Slides about life after the PhD.

Advising statement

My group work on image generation and acquisition from a variety of perspectives: computational photography, 3D lighting simulation, performance engineering and compilers, and deep learning. We work on both theoretical and applied aspects, and some of our work gets transferred to real-world products within a couple of years, while other research is more exploratory with a much farther horizon.

Working with students on exciting research is what motivates me every day. I like to have about 6-8 graduate and a couple of post-docs. I often co-advise people with other researchers, including people outside the department. I meet with each student once a week and have a group meeting where two students each week informally discuss their ongoing research. When research accelerates or around deadline, I am happy to add impromptu meetings. I often help junior students find a research topic but I expect senior students to define their research agenda. I expect students to conduct research, write papers and give presentations at conferences. I send students to conferences. I encourage students to work with and potentially supervise undergraduates (urops). I expect students to be in the lab “often enough around normal hours” but do not have rigid schedules. I think everyone is entitled to a healthy amount of vacation (4-5 weeks a year). I enjoy collaborations between students but I like to make sure that ownership and expectations get discussed. My group also enjoys collaborations with a diversity of faculty members in the department ( Bill Freeman, Wojciech Matusik, Phillip Isola, Saman Amarasinghe, John Guttag, Greg Wornell), outside the department (John Ochsendorf in architecture), and outside MIT both in industry (e.g. Sylvain Paris and Andrew Adams at Adobe, Miki Rubinstein at Google), and other Universities (e.g. Jaakko Lehtinen at Aalto, George Drettakis and Adrien Bousseau at INRIA).

I expect a typical PhD thesis to include 2-3 projects of publishable quality. While some amount of coherence is appreciated, I put more weight on critical mass and research quality. I am happy to let students graduate as soon as critical mass is achieved, but I also discuss their career goals and how to best achieve them, which leads some students to stay longer to strengthen their resume, especially if they want a faculty position.