**Graphics & Imaging Lab -- Recruitment**
Derek Nowrouzezahrai
![](http://gfx.lab.mcgill.ca/img/DigitalLightAndMatterLabLogo.png width="62.5%" style="background-color:#000;")
This page provides guidelines and a Q&A for anyone interested in joining my lab.
!!! Warn
Please read the **whole page** before contacting me.
How To Join The Lab
====================
I get a lot of e-mails from candidates interested in joining my lab, whether for undergraduate internships, graduate studies or post-doctorate roles.
If you're interested in joining my lab, please:
1. read this entire document *before e-mailing me*,
2. review my [home page](www.cim.mcgill.ca/~derek/) to ensure our interests align, _and_
3. follow the [procedure](#Procedure) to properly format your initial e-mail to me.
!!! Error: Before we start -- are you about to send me SPAM?
Please don't contact me unless you're interest in **computer graphics** _and/or_ **applied machine learning**. I will not answer generic e-mails.
Procedure
-----------
If you decide to e-mail me, please format your e-mail as follows:
* Format your **Subject** according to `[Type,NumTopTier,NumPubsAndSubs]` where:
- `Type` is the _type_ of position you are interested in. Possible values are:
- `Mp` for Master's project (both M.Sc. and M.Eng.),
- `Mt` for Master's thesis (both M.Sc. and M.Eng.),
- `P` for Ph.D. student,
- `PD` for a Post-doctoral position, _or_
- `O` for any other role (e.g., intern, McGill B.Eng thesis/capstone/[SURE](https://www.mcgill.ca/engineering/students/undergraduate/research))
- `NumTopTier` is the number of technical papers you've published at the following venues (do not count papers _in submission_): **SIGGRAPH**, **SIGGRAPH Asia**, **TOG**, **NeurIPS**, **ICML**, **ICLR** and **CVPR**.
- `NumPubsAndSubs` is the _total_ number of papers you've published and/or have currently in submission to _any_ venue. This count should include the papers you considered in `NumTopTier`; and so `NumTopTier` $\leq$ `NumPubsAndSubs`
!!! Warn
I will **ignore** e-mails whose Subject lines don't adhere to this format.
* Keep your e-mail concise, listing as much of the following information:
- your current institution, position and current GPA,
- your research areas of interest: be _as precise_ as possible (e.g., "I like computer graphics and machine learning" might be suitable for a Bachelor student, but it's way too broad for a Ph.D. applicant)
- links to your **github** profile, **website** and/or **CV**.
For example, if you're interested in a Ph.D., you've published one SIGGRAPH paper and have an ICML paper in submission, your e-mail should look something like:
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ none noscroll
From: you@your-domain.com
Subject: [P,1,2]
My name is Fizz and I'm a Master's student at Buzz studying applied procrastination. I'm
interested in generative adversarial networks for smoke rendering, and I'd like to extend
your work on anti-aliased emojis to handle real-world data and dual-space representations.
My webpage is www.bestmodesty.com and it links to my CV and github.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Admission Requirements
-----------------------
You have to apply to McGill University through either the [Electrical and Computer Engineering](https://www.mcgill.ca/ece/graduate/graduate) or [Computer Science](https://www.cs.mcgill.ca/academic/graduate/admission/) Departments.
In addition to institutional & departmental requirements, I have some guidelines:
1. The best was to pursue a Ph.D. in my lab is to have already conducted good research with me, i.e., typically during a Bachelor's or Master's project.
2. McGill students should take [ECSE 446/546](ecse446.html). Doing well in the course, as well as in [COMP 557](https://www.cs.mcgill.ca/~kry/comp557F18/) (cross-listed as ECSE 532) and/or [COMP 559](https://www.cs.mcgill.ca/~kry/comp559W19/), doesn't hurt.
3. Ph.D. applicants **who have not yet worked with me** should have at least one publication, and ideally at least one at a top-tier venue. If you don't, a strong letter of reference from someone I know/trust is **required**.
- So, if your Subject is [P,0,0] and I don't already know you (or your supervisor), I won't consider your application nor reply to your e-mail.
4. Post-doc applicants **must** have at least two **top-tier** publications in my areas of interest. For example, I won't consider an applicant with Subject [PD,1,1].
Here's a visual plot of how various profiles would rank, both relative to each other _and_ on an absolute scale of "likelihood I'll seriously consider your application":
********************************************************************************
* [PD,0,0] [P,0,0] [Mt,0,0] [Mp,0,0] [Mt,1,0] [P,2,1]
*
* spam |-----------+ 0% +--------------------+ %50 +--------------------| 95%
*
* Likelihood of Admission
********************************************************************************
General Questions
====================
- Q: What kind of projects do we work on in the [Graphics & Imaging Lab](http://gfx.lab.mcgill.ca/)?
- A: We work on animation and rendering algorithms for video games and VR, physics-based models for visual effects in films, and applied machine learning.
- Q: I like math/theory more than code (or vice-versa); are there projects for me in your lab?
- A: Maybe! We work on projects that run the spectrum from purely theoretical to more systems-oriented, so chances are there's a fit for your skills and interests. More important than your existing skills, however, is your willingness to learn (many) new tricks.
- Q: What kind of job can I get after working in the Graphics & Imaging Lab?
- A: Graduates from our lab move on to successful careers in graphics, video games, VFX, and industrial research. Many of our projects include collaborations with industrial partners (see [below](#collaborators)), and so students often undertake industrial internships during their studies.
- Q: Do you work with other groups or industrial partners?
- A: Yes. Students in the Graphics & Imaging Lab often work on projects with collaborators from around the world. We also have very strong ties with several academic and industrial partners:
](images/Logo_Ubi_Stacked_HOR_Montreal_BLACK.png width="20%")
We work closely with Ubisoft Montreal's [**La Forge** R&D lab](https://montreal.ubisoft.com/en/our-engagements/research-and-development/), through [NSERC's Industrial Research Chair](http://www.nserc-crsng.gc.ca/Chairholders-TitulairesDeChaire/Chairholder-Titulaire_eng.asp?pid=990) program, to explore how modern rendering and simulation methods can be adapted to the needs of high-performance interactive graphics experiences. Members of the Graphics & Imaging Lab can work on collaborative projects with Ubisoft, which includes the opportunity to work on-site with R&D specialists at their studios in Montreal.
 and [FAIR's Montreal Lab](https://research.fb.com/blog/category/facebook-ai-research/)](images/FRL_Facebook_Reality_Labs.png width="20%")
Our lab maintains strong ties with [Facebook's Reality Labs](https://research.fb.com/blog/category/augmented-reality-virtual-reality/) (née Oculus Research), in collaboration with [Facebook's Artificial Intelligence Research](https://research.fb.com/blog/category/facebook-ai-research/) groups. Together, we work on new hardware and graphics software to enable the next-generation of virtual and augmented reality displays.
](images/autodesk_research.png width="20%")
We collaborate with Autodesk's Montreal office and Solid Angle's R&D group to develop new image synthesis techniques for realistic 3D content creation. These tools will push the boundaries of offline rendering techniques.
](images/elementai_logo.png width="20%")
We're working with Element AI to understand how our understanding of the physics behind image formation can help teach computers to understand the world around us. We're designing approaches that allow computers to understand and express 3D structure implicit in 2D images.
](images/mila.jpg width="20%")
We collaborate closely with the [MILA](http://mila.quebec) to apply modern machine learning to problems in graphics and imaging, as well as exploring how graphics models and techniques can help machine learning models better understand the world. Students working on applied machine learning in the Graphics & Imaging lab are eligible for MILA membership.
Current McGill Students
=======================
I've had the pleasure of supervising fantastic undergraduates in the past, many now pursuing careers in the games, digital media, graphics and feature films industries. Undergrads work on real problems tailored to their interests and the lab's needs. In the past seven years, roughly 60% of undergrads who've worked with me choose to continue with graduate studies in the lab.
- Q: Can you supervise my Honours Thesis project? Do I have to come up with my own project, or do you have a list of existing thesis projects?
- A: I'm happy to chat about HT projects. I have a list of topics, but I'm open to ideas you may have that are related to our lab's interests.
- Q: Do you offer paid research internships for undergraduates?
- A: Sometimes! I'm always looking for students excited about computer graphics, numerical simulation, computational physics and applied machine learning. If you're serious about tackling challenging and fun projects in any of these areas, then I'm happy to arrange a meeting to chat about it.
!!! Note
If you're a McGill undergraduate interested in working with me, please follow the [procedure](#Procedure) detailed above and contact me.
Applying for an [NSERC Undergraduate Student Research Assistantship](http://www.nserc-crsng.gc.ca/Students-Etudiants/UG-PC/USRA-BRPC_eng.asp) is a good start, and this is one of the few outlets I will support for paid undergraduate internships. If you're an engineering student then the [SURE](http://www.mcgill.ca/engineering/current-students/undergraduate/research) program is a great way to supplement an NSERC USRA, and the Graphics & Imaging Lab has [several projects](https://www.mcgill.ca/engineering/current-students/undergraduate/research/electrical-computer) listed on the SURE website.
In exceptional cases, I may coordinate joint internships with industrial partners, like Ubisoft, Facebook Reality Labs, Autodesk and ElementAI.