- Research Interests: Audio Signal Processing, Music Information Retrieval
- GPA: 9.29/10 (at time of application)
- GRE: 321, 166 Q + 155 V
- TOEFL: 114, 28 Reading, 30 Listening, 29 Speaking, 27 Writing
- University: IIT Bombay B.Tech + M.Tech (BS + MS) in EE specializing in Signal Processing
- Experiences:
- With a Professor at my University: Have worked with her since the beginning of Junior (3rd) year, she is also my Master’s thesis advisor. My initial work with her was on Music Information Retrieval, and for my Master’s Thesis, I worked on generative audio synthesis.
- Research internship at Kyoto University in statistical signal processing for neuroscience
- Research internship with the Music Tech Group at UPF Barcelona in Music Information Retrieval
- Publications:
1 domestic conference paper (NCC,accepted), one poster at an international conference (ISMIR,extended abstract), 1 Journal (Neural Networks, submitted), 1 International Conference (ICASSP, submitted) - Presentations:
Presented my accepted paper at NCC. Presented my internship research work at Honda Research Institute Japan. - Letter of Recommendation:
3, one from each professor I worked with. Pretty sure I got solid recommendations from all 3 as I had worked with them for a good amount of time and also built a good rapport with them.
I want to do my Ph.D. in audio signal processing/music information retrieval. In the US, I applied to a mixture of Music Technology Ph.D. programs (which comes under the departments of music), and EE programs where I saw faculty working in signal processing/audio. I was ready to apply to Music programs there were faculty working at the intersection of signal processing/MIR and a variety of other interesting topics. However, one thing to be aware of is in some departments, Music Ph.D.’s are not considered STEM, so international students will not be applicable for the STEM F1-OPT extension (3 years) to work after studying, so contact your department and verify this! Unlike the US, UK Ph.D.’s don’t guarantee funding with their admissions, so you either have to find funding separately (fellowships, scholarships, etc. )or pay. Also, the UK Ph.D.’s funding applications expect you to have a detailed idea of what you would be working on (research proposal + discussion with advisors + rough timeline of your Ph.D.), so pace yourself accordingly. In the UK, I applied directly to programs that had funding guaranteed, either through fellowship or a scholarship.
- US:
- Stanford Music (CCRMA): PhD in Computer Based Music Theory and Acoustics → Rejected on 3 March
- NYU Steinhardt Music Technology: PhD in Music Technology → Rejected on 5/ March
- Georgia Tech Music: PhD in Music Technology → Accepted with partial funding on 2 March
- UIUC: PhD in Electrical and Computer Engineering → Accepted with fully funded RA offer on 5 Feb
- Johns Hopkins: PhD in Electrical and Computer Engineering → Accepted with full funding (fellowship for first year) on 24 Jan
- USC: PhD in Electrical Engineering → Accepted with full funding (fellowship for first year) on 14 Feb
- UK:
- Center for Digital Media, Queen Mary University London: Artificial Intelligence and Music Program → Accepted with funding on 22 March
- Center for Vision, Speech and Signal Processing, University of Surrey: PhD in Vision, Speech and Signal Processing → Rejected on 24 March
First steps are to give GRE and TOEFL, and get them out of the way. I gave GRE and TOEFL in September, within a week of each other. Once done with that, start searching for Professors who you would be interested in working with. To do this, just go to the university department websites, and check out the professors and their research groups. They also often post additional information on their websites (do they need new students, what are they working on currently etc.)
I also found it really advantageous to mail the Professor before applying and asking them directly if they will be taking new students etc. This serves 2 purposes: You know that prof is available in case you want to work with him and also helps you in selecting schools. If he is not taking students, then you might be better off not applying to that school (if no one elses research interests you). I found the following links 1
2 to be helpful while drafting mails. This is
the template I used while mailing:
Hello Professor XYZ,
I’m Krishna, a (current degree) student and I am writing this email to you as a prospective Ph.D. applicant.
I’m currently (describe current work in 1 or max 2 lines). I have attached my CV below, which highlights some of my research in (xyz). I’ve also linked my homepage, where you can find more information on my research.
I find your work on (xyz) fascinating, and (a sentence indicating that you have read his work or worked on something similar) I also like your work on (something else that interests you) and would love to hear more about graduate opportunities with you.
Thank you for your time and, I look forward to hearing from you!
Sincerely,
Krishna Subramani.
Homepage: (url to homepage)
With the following mail template, I heard back from almost everyone I mailed (7 out of 8 places I applied to). Another thing that I felt that helped was building a small webpage, where you can describe your research in more detail
(publications/project reports/code etc). I used Jon Barron’s website template.
You can use github pages to host your website, or a server if your university provides one (IITB provides one through bighome)
The next important component is the SOP. This is not the most important part of the application (your research experience, it’s relevance with your prospective profs current research and your recommendations are very important). What the SOP does however is highlights your research strengths, and allows the department to match you to prospective profs (whom you can name in your SOP). Here are two blogs 1 2 which I found really helpful while drafting my own SOPs.
A few more miscellaneous takeaways from my experience:
- If possible, meet-up faculty if they come to local conferences/university symposiums or talks - I was lucky in this, the professor I wanted to work with at UIUC was in a workshop in India, so I could attend the workshop and talk to him.
- Talk to graduate students who work with the advisor you want to work with. They can give you a good insight (advising style etc.). Talk to multiple grad students if possible to get a less noisy estimate of the advisor.
- One potential reason for my rejection from music programs might be my lack of formal music training, which a few schools desire. Also, the Music Technology Ph.D. programs are more competitive than the ECE Ph.D. programs in that they only take 2-3 students every year. So, if you are applying to Music Technology Ph.D. programs, keep all these in mind.
- For the UK programs, contact the potential advisor beforehand, as they can help you in drafting the research proposal, and they can also help with your application in general.