**Advice and Guidelines for Wojciech’s students** **[Wojciech Jarosz](https://cs.dartmouth.edu/~wjarosz/)**

Dartmouth College

(This is a living document) Last update: 7 April 2021 *Credits*: Some of these tips are based on similar documents from [Frédo Durand](http://people.csail.mit.edu/fredo/student.html), [James F. O’Brien](http://obrien.berkeley.edu/Prof._James_F._OBrien/F.A.Q..html), and [Keenan Crane](https://www.cs.cmu.edu/~kmcrane/faq.html), among others. # Make and maintain an academic website As a current or prospective graduate student/researcher, you should have an academic website (no, not a LinkedIn page but a real website, preferably hosted at your academic institution). You should (at least) include a photo of yourself, your contact info, your research interests, your publications or projects, and a downloadable CV. If you don’t yet have publications, then include a section with any personal projects/coding demos you have created (that are at least tangentially related to your field of interest). This demonstrates your passion and shows you are using your spare time to reinforce your relevant research knowledge. Treat your website as an extension of your resume/portfolio. From personal experience, when hiring students or interns in the past, it was a red flag if I could not find any academic online presence. # Putting papers (pre- & post-prints) online Once your paper is accepted, you should absolutely put it online (on your website discussed above). This improves the visibility of your work and makes it easier for colleagues to find it (some researchers – e.g. in industry – don’t have digital journal subscriptions). When doing this, you should be mindful about any restrictions/requirements imposed by the publishers on pre-/post-prints. The graphics community has negotiated with most publishers the right to post personal version of their papers on academic websites. Most also require including a copyright statement on the project page, and/or in the PDF stating that the official version is available at the publisher’s digital library. Talk to me if you are unsure what the rules are or how to do this. I recommend making a separate page for each project/paper, and including as much of the following as possible with each one: - A **PDF** of the paper and any **supplemental documents** - The **video** - Include **data/scenes** and **release your code** to help the community reproduce or compare to your work. This may initially seem like it will give “the competition” a leg up in beating your approach but this is the wrong mindset. You want other researchers to build off of your work. The papers that have easily reproducible results and readily available implementations become the standard against which new approaches are compared. This builds impact. - Include any **additional/interactive comparisons** that are easier to provide on a website than in a static paper or linear video. See [this page](publications/jarosz19orthogonal.html) for an example. - Include a **BibTex entry**. See my [writing tips documents](writing.md.html) for important details on how to structure these correctly. - Add **acknowledgements** directly on the project page. - Add **slides (both PDF and source Keynote files)** once you have them available. This makes it easier for others to present (advertise) your work. - Even better, if you do a talk and record it, **post the talk to Youtube** and embed/link to it from your project page for people who couldn’t attend the conference. [Keenan Crane](https://www.cs.cmu.edu/~kmcrane) is particularly good at this. # Vacation & winter break I need to approve your vacation, so send me an email before you book. In general I approve, but I like to know, and we may need to schedule around deadlines and other constraints that you are not aware of. Roughly speaking, grad students and Post-docs funded by Dartmouth can take up to 1 month of vacation including holidays per year (¾ of that if you don’t work with me during the summer). See [Dartmouth CS’s official policy](https://web.cs.dartmouth.edu/graduate/expectations-policies-and-procedures), and the [Dartmouth HR department’s website](http://www.dartmouth.edu/~hrs/benefits/) for more details. You are responsible for your own academic success, so you need to schedule your vacations around relevant publication deadlines. For instance, if you are working on a SIGGRAPH project, it is probably not reasonable to take 10 days of vacation without working around the Winter break. Winter break and the SIGGRAPH deadline have unfortunately been a constant point of conflict for decades. In some cases working remotely is a possibility, but as above, you need to clear your vacation plans with me first. # Working from home vs. in the lab As a general rule, I expect PhD students and Post-docs to do their work in the building/lab. Among other things, this makes it easier to have ad hoc face-to-face discussions, and improves the culture of the lab. In addition to meetings with me, you have labmates which can be an invaluable resource when you need to talk things through with someone. I understand that sometimes it is more convenient to work from home. That is OK from time to time, but you should inform me by email (note: it is not reasonable to work from home e.g. 3 out of 5 work days). Likewise, if you are sick, I wish you a speedy recover, but please email me to let me know you won’t be coming in. # Making good progress on your PhD Every PhD career is different, and there is no single right way to go about it, but here are a few guidelines for what I expect of a PhD student working with me. In the Dartmouth system, the purpose of the first year is to do research with one or two faculty members with the goal of finding a permanent PhD advisor. In effect, it’s a time for both you and me to decide whether we can work effectively together. You will typically be put on a DF (Dartmouth Fellowship) during some or all of that time, which means you will additionally have TAing obligations. If you already know you want to work with me upon being admitted then you are ahead of the game. Ideally, you already have some domain knowledge to dive immediately into interesting research projects. By the end of the first year, you should aim to make sufficient progress on a research project that it passes my threshold for submission to an international publication venue (whether it gets accepted is out of my control). If your research hasn’t reached that point, you should at the very least be able to convincingly describe a viable path towards a solution to an interesting research problem and a publication strategy. Beyond the first year, you should aim to submit your research to the best venues in our field (e.g. SIGGRAPH (Asia), TOG, EGSR, EG). Getting a paper accepted to one of these venues every 1-1.5 years is good progress (along with potentially some publications on projects where you are not the primary student). Since a typical PhD in CS at Dartmouth is 5 years, this means your dissertation may be the culmination of about 3-5 published papers. One final comment about being a researcher or getting a PhD: It can be one of the most rewarding careers, but it will not be enjoyable (and you will not be successful/competitive) if you are not passionate about it. It is important for you to carefully maintain your work-life balance to remain satisfied. From my experience, the most successful researchers accomplish this by allowing the passion for their work to permeate into their hobbies, and so the distinction between work and play becomes blurred. If you are working in graphics/rendering, this could mean one of your hobbies is working on your own rendering system “on-the-side”, or writing graphics demos, or tinkering with optical systems or 3D printers. These may not be directly related to your chosen research topic, but it is an incredible asset if the time you spend on a hobby reinforces your depth and breadth of knowledge as a researcher. Your hobby may even lead to a new research idea. Consider someone who does this in their spare time vs. someone who treats a PhD as a 9-5 job. Now ask yourself who will likely be more productive at research. That being said, it is important to strike a balance and not allow your hobbies to drain time away from making actual progress on your PhD. # Classes vs. research I recommend getting started on research as soon as possible. Given the expectations of the first year, you need to show me within a relatively short amount of time that you have the potential to be a serious and productive researcher. If you are also on a DF and TAing, then you may want to postpone taking classes until after your first term. The exception to this is if you don’t already have the necessary domain knowledge to immediately start on research. In this case you may need to ramp up by taking classes, or learning the material on your own starting in the first term or before you arrive. # Attending conferences Conferences are a great opportunity to meet colleagues, get inspired for your next research project, and start new research collaborations. Talk to people. Ask questions during the presentations, or approach the authors after the presentation if you prefer a quieter setting. Reading upcoming research papers in our reading group can allow you to articulate more meaningful or relevant questions. If you were puzzled about something during our reading group discussions and the conference presentation doesn’t clarify it, ask the authors during the conference. This is perfectly natural. Try to meet and hang out with people from other labs, and not just your own lab. If this is intimidating, you may find it useful to leverage your own labmates to approach other students/researchers as a group. You could also “tag along” with a more senior student or researcher (or me!) and have them introduce you. Assuming there is available funding, I like to give PhD students the opportunity to attend a conference (typically SIGGRAPH) in their first-year even if they did not submit a paper. It is very helpful to see what the top researchers in the field are doing early on in your career, to get inspired, and to calibrate against the expectations of the research community. After the first year, I will provide financial support for you to attend conferences when you have an accepted paper or talk. If you do not have a submission but want to attend the conference, I encourage you to apply to be a volunteer at the conference (which sometimes includes partial financial support). Talk to me about it first, and I might be able to provide additional financial support if you already show initiative and secure a volunteer position. If funding is plentiful, you may have the opportunity to attend additional conferences. # Internships Internships can be both helpful and detrimental to a student’s career. They are a great way to get a different perspective and to see what problems are most relevant “in the real world”. Internships are also a great way to get industrial researchers interested in your work which can lead to employment opportunities after graduation. You should, however, keep your PhD timeline and goals in mind – e.g. doing an internship every summer will probably impede your PhD progress unless the internship is your PhD research. Particularly with Dartmouth’s PhD system, I recommend not doing an internship in your first year, but waiting until further in your PhD when you have more experience and can land internships more relevant to your research agenda. Prefer internships that could lead to a publication, or can turn your research into a visible tech transfer. If you are considering doing an internship over the summer, you should come to me first (yes, before you apply) so we can discuss how it fits with your PhD research timeline. I may also be aware of specific opportunities or have suggestions for where to apply. I am happy to serve as your reference for internships, but it is awkward if I’m contacted for a reference by a company and I don’t already know your plans.