Hi, I'm Vicente and I'm a faculty at the University of Virginia. If you are reading this webpage chances are that you have interest in becoming part of our new UVA Computer Vision group. I'm happy you are considering us. This is a first step and you will probably achieve what you are looking for if you keep persisting (here or elsewhere).
Sending me emails: I receive many emails from prospective applicants, especially close to the graduate applications deadline. I read all of them but I only read an email from beginning to end if it is a short email (i.e. up to three very concise paragraphs), and probably only the first paragraph or less if they are too long. On the other hand, while I read all of them (to some degree) I can not reply to all these emails. The problem with replying is a mix of prioritizing time and my weak memory. For instance it might happen that I think you are indeed a stellar applicant but by the time I remember I received your email it is already buried among other emails from other prospective applicants and other emails I have to reply more urgently. Please include in your email [Your Name][UVA Vision and Language Research - Prospective Applicant] in the subject of your email. I routinely take time to search my inbox with these keywords and reply to emails.
Chances to obtain an admission: The short answer is that I can not tell what are your chances for admission. I can not, for various reasons. One reason is that admissions are competitive and judgments can only be made when looking at the entire pool of applicants in any given year. Another reason is that eventhough individual professor recommendations on applicants count a lot, the final admission decisions are made by a committee. So if I tell you that you have good chances then you might resent me if you end up not getting an admission, and I don't want that. I could however tell you that you have a strong profile without implying any judgment on your chances for admission. Here are a few examples of things that would impress me (and did impress me in my first round of participating in the admission process) in a PhD applicant:
You published paper in a reputable venue in any field of AI: ICCV, CVPR, ECCV, NIPS, ICML, AAAI, ACL, NAACL, EMNLP, TACL, TPAMI, IJCV, KDD, IJCAI, SIGGRAPH, SIGGRAPH Asia, ACM MM. I would also be impressed by a paper in a next tier vision conference that I am also familiar with like BMVC, ACCV, ICIP, or WACV, especially if the research is related to the projects in our group. I am less impressed however by a long list of publications in regional or other minor venues (I might even be worried if this list is really long), in that case I would also look for other things listed here.
You worked with another Vision/NLP/AI/ML professor at your school and you get a recommendation letter from said professor. If your school does not have an active professor in any of these fields and you are in the US, there are opportunities like NSF REU or CRA DREU to obtain such experience in another institution. If you qualify and are interested in CRA-DREU contact me, I will try my best to support you. If you are an international applicant, an equally good thing would be any type of international research experience, for instance if you spent a summer or a semester abroad participating in a research lab. This is especially helpful if I already know your supervisor at this research lab by reputation or you produced something I could actually judge (a paper on arXiv, a repository on github, a publicly available dataset, etc).
You have a track record for your engineering skills. For instance if you had an internship at a reputable technological company, and/or if you obtained a recommendation letter from somebody experienced in Vision/NLP/AI/ML from said internship. You can also impress me if you participated in a program like the Google Summer of Code in a cool project. You can also build points if you had success in programming competitions like the Google Code Jam, Top Coder Opens or have a reputable history of participating in their regular Round Matches, ACM Collegiate Programming Contests, Informatics or Math Olympiads, etc. I would also be impressed by open source projects that you started/significantly participated and is public and reasonably popular (e.g. on Github).
Is your group recruiting new PhD students next Fall? Currently, I am open to recruiting students who are interested in Computer Vision and Machine Learning, so no need to ask this.
Long answer: I get this question a lot, and understandably so, admissions cost some amount of money (and a considerable amount depending on your country) and you want to maximize the number of applications in places that are actually hiring in the areas of your interest. However, this question is a bit complicated. First, in the US, most applicants are admitted to a program and not directly to a research group of a professor, as is customary in some other countries. PhD students are guaranteed funding on their first year regardless of matching to a specific professor. Second, professors often do not know sufficiently in advance what is going to be their funding situation a year from now, therefore are very reluctant about making definitive assertions about how many students they will be recruiting a year from now. However, since you are funded by the school during your first year, professors will actually recruit you probably not next Fall but next next Fall. This makes the question more complicated because a professor now needs to predict his funding situation two years from now. In summary, the relevant question is not whether a professor will be recruiting students next Fall, but will he be recruiting students next next Fall. This is why in most cases it is not very productive to ask a professor whether he will be recruiting students next Fall. My advice is to always apply to schools where there is more than one professor that you could potentially work with. If you are interested more broadly in images and machine learning, please keep reading because I am not the only faculty working in these areas at the University of Virginia.
Why the University of Virginia?: Again, glad you are considering a graduate education in the Department of Computer Science at the University of Virginia. Our school is among the most reputed public universities in the United States according to the most widely used (but very often maligned) US & News rankings, as of this year trailing behind Berkeley and tied with UCLA at 2nd place and just above Michigan and UNC. More relevant to our program is that we are placed within the top #30 departments in Computer Science by the same rankings. Another interesting ranking based on which schools produce more PhDs that become professors at other schools places us #34 [Clauset et. al. ranking]. Another notable fact is that UVA is a powerhouse in Business and Law placing #8 in the rankings, surpassing many other world renowned institutions. This last bit is not super relevant to us but a lot more people will be familiar with your school because of this.
Another interesting fact is that the UVA
campus Grounds are designated as UNESCO World Heritage Site, which is impressive considering that the other man-made structures in the US with this designation are the Statue of Liberty in New York and the Independence Hall in Philadelphia. Charlottesville itself is a nice small city in Virginia with many restaurants, bars, and shops. I live walking distance from the Downtown Mall which is nice but I end up spending a lot more time at school anyways. I like that the airport is really close and has flights to a reasonable number of places. I like the place but to be fair I spend time in New York for the holidays/breaks, you really can not compete with that. Finally, we are a two-hour train ride away from Washington, DC, home of many great museums.
If you are interested in Vision/AI/Machine Learning , this is also a great school. I'm starting a new vision lab focusing on Vision and Language, my collegue Kai-Wei Chang is widely known for structured prediction in Natural Language Processing, Honging Wang in Information Retrieval and Data Mining, Yanjun Qi in Machine Learning and Bioinformatics, Connelly Barnes in Computer Graphics and Vision, Quanquang Gu in Statistical Machine Learning, and it is expected that there will be more faculty in Data Science and Artificial Intelligence joining us in the future. Closely related departments at UVA also have strong faculty working in related areas like the faculty at the Predictive Technology Lab in the Systems Engineering Department, and the Video&Image Analysis Lab in the ECE/BME Department. Also check the amazing research by other faculty working in other image domains by looking at the Image Processing Seminar series website.
Last Modified: December 2016.