The Intern Featurette #4 — Jeya Maria Jose — National University of Singapore


Courtesy: Jeya Maria Jose

Jeya Maria Jose pursued his internship at the National University of Singapore this summer in the booming field of Artificial Intelligence. He is a tech geek who spends his free time exploring the latest technology preferments in the world. He is adept with various domains such as Robotics, Deep Learning and Circuit Design. Let’s interact with him to get his view on the process of attaining internships.

Hi, can you please let us know of your process in which you were selected for the internship?

Hello. I looked into the websites of the labs of top Universities and the research projects being done in the same. Whenever I found a project interesting or when I felt it was relevant to my interested domain , I mailed the principal investigator of that research laboratory directly. I received some positive responses and from there on I chose which would be the best possible option for me and confirmed with the Professor for the research internship. From the professor’s point of view , I think selection of a student depends solely on the availability of research openings in the lab. Only then they would even look over the cover letter and the previous works done and decide if a student would be eligible to carry out the work.

Can you share your experience of sending mails to various universities requesting for internships? How much time consuming was it?

Mailing professors and waiting for their replies , especially to those in top Universities abroad needs a lot of patience and perseverance. They would be so busy that they would just neglect any mail that has a subject “Application for Research Internship” or have filters that would directly send your mail to their spam inbox. But, most of them would be in need of interns for helping their PhD scholars or Lab Scientists on their research projects.

Also, they would definitely expect you to be good in the respective field that you are applying because it’s a waste of time for them to take you in and end up only teaching the basics of a particular domain in the internship.

So, I prepared a cover letter carefully giving details about the various projects I had done in similar fields and the notable competitions where I had secured some place. I also gave details about the frameworks and platforms I am comfortable to work on. Also , I sent the link of my GitHub repositories and attached a letter of recommendation from my previous internship and my resume.

I started mailing just after Festember. I mailed around 5 professors daily. I targeted only NUS , NTU and ETH due to personal interests. I would send a reminder mail after 2 days to the same professors if they don’t reply. I followed this routine for around 3 weeks and stopped when I received 4 worthy offers. After getting more details about the project that I would be working on from those Professors, I finalized a Professor and mailed him I would be joining his lab for the internship and to start the Visa process. This happened before the end of October.

What domain does your project belong to? Can you explain about it briefly?

I worked on Deep Learning for Medical Imaging. I used Generative Adversarial Networks (GANs) for CT image Synthesis and Segmentation of Intracerebral Haemorrhage Segmentation. I was mainly involved in building a novel Segmentation network that uses the concept of Adversarial Learning for the process of Brain Tumor and Haemorrhage Segmentation.

Also , I participated as a team with my PhD scholar and Professor for two challenges (BraTS and MoNuSeg) of MICCAI 2018, the 21st International Conference on Medical Image Computing and Computer Assisted Intervention, which is going to be held in Granada, Spain later this year. I worked on the Overall Survival Prediction of Brain Tumor Dateset of BraTS( organized by UPenn) and used the segmentation-adversarial I had worked earlier for cell nuclei segmentation for MoNuSeg.

How was the work culture? How different was it from your home university research work?

Work Culture was very different. There were no timings or any dress code. There was not even a mentor-mentee type of relationship with the Professors and scholars. They followed a collaborative culture where in we were solely responsible for taking out the best experience out of the research. You can put in all your effort and extract best results or do nothing much , have fun because the professors are chill with anything.

Even though there were no timings , I found the people in my lab working for 10 to 14 hours a day and many even turned up on Sundays. I felt their driving force was the interest they had in what they were doing which was something I found very motivating.

What’re the different sub-domains and opportunities that are available in your domain?

Deep Learning(DL) is actually a subdivision under Machine Learning in Artificial Intelligence (AI). It has been a very hot field since its revolution in 2012. In just 6 years , there has been an humongous growth in the applications of DL across various domains.

DL has its applications in Computer Vision , Robotics , Self Driving Cars, Natural Language Processing, Speech and Music Generation, Finance, Healthcare, Gaming and a lot more.

A lot of research in academia has bloomed in this sector and also most of the top companies have already heavily invested in this field. So, I believe opportunities are never-ending and don’t need any further clarification.

Did being a part of any club in college help you in attaining the internship?

I don’t think anyone gets any internship just because they are a part of some tech club.

I believe it’s the work and effort you put up to pursue research that takes you to places.

Being in a tech club can definitely give a lot of opportunities to work in interesting projects and learn new things by meeting like minded people. However one can do all this without being in a tech club too. I was a part of Spider and it did help me develop my technical skills which were helpful during the internship.

What was your previous work experience in projects on your own self, under your dept profs, internships, and how much it helped you in this internship?

I worked on some interesting project statements previously for competitions in Shaastra, Pragyan and also actively participated in TIIDC and E-Yantra. I did a couple of projects as a part of Spider. I also underwent Internship in my 2nd year in IIT Madras where I worked on Optimization algorithms for Control Systems. I tried to explore as many domains as possible and learn the basics of the same to sort of which field interested me a lot.

I guess all my previous works, projects and internships helped me do well during my internship. The coding languages and frameworks I learnt and mastered during my previous project experiences were surely of great help.

Any piece of advice to juniors who would be applying for research interns?

Plan well beforehand of what you want out of the internship and apply accordingly.

There are a lot of internship programs like CMU’s RISS , MILA’s Summer Internship apart from the classical DAAD , MITACS, and SRFP. So , google properly on all the internships available in your domain. As all these results come a bit late, better start mailing well in advance to various research groups you are interested in so that you can have a back up. In case you land on multiple offers , always bank on the ones that you think can extract the best research work from.

What’s your future plan of action? Job or highers, and why?

I am passionate about contributing to research and I look it as my career. So, I’m planning to pursue higher studies.

Feel free to contact Jeya Maria Jose for further queries!

WhatsApp — 93456 79386

Facebook — https://www.facebook.com/jeya.jose

LinkedIn — https://www.linkedin.com/in/jeya-maria-jose-357951130/

An initiative of SENSORS NIT TRICHY, the annual technical symposium of the Department of Instrumentation and Control Engineering of NIT Trichy

Source: Deep Learning on Medium