Parth, Suren, Tejas (ABS Team), Vaibhav, and Kushagra (Titane Team)
Summer Internship at INRIA, France
What inspired you to take up a research internship?
Kushagra - I have always been interested in geometry and math since high school. Research has also interested me a lot, and I had always wanted to give it a try.
Vaibhav - I had completed a Machine Learning course through Coursera, which piqued my interest further, and I wanted to explore it in depth.
Did you consider applying for SURA?
Since SURA was announced rather late, all of us had already secured internships by then, so we did not consider applying. INRIA's research environment is not significantly different from SURA's, but INRIA offers more flexibility and a greater scope for social development.
Were you drawn to the internship because of the topic or the research involved?
Suren - I was not particularly inclined toward any specific topic. I wanted to try out research and have always preferred the theoretical side of Computer Science over the practical side.
Vaibhav - I was asked to read the professor's research papers, which helped me become familiar with the topic at hand.
Kushagra - I would advise not to be hesitant and avoid any area just because it is highly competitive.
Parth - One must try internships that require courses that have already been covered. This doesn't require too much background; otherwise, it may take a long time to become familiar with the topic at hand.
Is there a luck factor involved?
Everyone agreed that luck plays a role.
Kushagra - Targeted emailing and knowing the experiences of previous interns help reduce the luck factor.
Parth - I emailed five or six professors after reading the abstracts of their research papers, which helped me receive a few positive responses.
Tejas - There was a message shared on WhatsApp groups about a possible internship under a professor. Luckily, I paid attention to it and then told Suren about it. The ABS team had received many competitive applications. After Suren and I applied, only two positions were left. Word spreads quickly, so it's important to stay in touch with your seniors.
Did you specifically prepare for the internship?
Vaibhav - I completed a course on Coursera. Since my work in the internship was a continuation of previous research, I also reviewed the earlier work.
Suren - The professor sent me some reading material about the recent developments in my field of research.
Kushagra - I watched the professor's YouTube videos and simulations to understand my topic better.
Parth: My professor sent me short videos on bioinformatics algorithms.
Would you all now introduce your research problems please?
Kushagra: My work involved computational geometry. I was tasked with improving the CGAL algorithm related to alpha shapes and generating a specific type of 3-D mesh wrap for a given model, which is used in various geometry processing and graphics applications. I improved the complexity and performance by adding another parameter tolerance to stop subsequent refinement based on suitable approximation.
Vaibhav - I worked on a sub-project as part of the larger BIM2TWIN project. The aim of my project was to classify different objects in LiDAR point clouds using Deep Learning models. This has potential applications in automated progress monitoring of construction sites.
Parth - My work was on hierarchical clustering. I was given a set of RNA databases extracted from different parts of the world. My task was to develop a tree to identify the parent and child COVID samples. I managed to develop new heuristic algorithms that greatly improved time complexity.
Tejas - I worked on the sampling of proteins with multiple loops. My aim was to generate conformations of biological proteins efficiently, and I implemented an algorithm in C++ to achieve this.
Suren - My work was part of an ongoing project. The previous intern was developing a clustering algorithm. My task was to fix a loophole in the algorithm regarding an optimization method, which was not guaranteed to work, even if it did in practice.
What was the size of the team that you worked in, and to what extent did you experience collaboration?
The general consensus was that the team typically consisted of a head professor, senior researchers, post-doctorates, and interns. These individuals were very knowledgeable and helpful. The supervisors, who were senior researchers and post-doctorates, provided substantial support during the internship.
As the internship progressed, did you notice any difference between reality and your expectations beforehand?
Parth - I was initially nervous and determined to meet the professor's expectations. However, the professor and supervisors turned out to be very friendly and did not hold unrealistic expectations.
Vaibhav - I expected a smoother experience, but at times, small issues escalated into bigger problems during my internship.
Suren - My experience was better than expected. The professors did not expect second-year undergraduates to know everything. I had the opportunity to interact with people from multiple teams, which allowed me to learn about many active fields of research in Mathematics and Computer Science.
Kushagra - I attended multiple seminars and workshops, which was a great learning experience for me.
If professors do not expect a lot from undergraduate students, what do they expect?
Professors believe that intern students are sharp at effectively implementing algorithms. There is also an element of demand and supply at play. They also look for interns who could potentially become PhD candidates.
How were the professors? Did they encourage you to pursue higher studies?
Professors often seek PhD students. They helped us prioritise and provided key insights about career decisions. Independent thinking was also encouraged. The supervisors frequently shared their past research experiences and what they would do differently if given the choice to start over. Overall, it was a healthy environment and a welcome change from the IIT Delhi campus.
What skills did you acquire during the internship?
Suren - Besides the academic experience, I realised that I lacked in collaborative efforts. I also learned to write cleaner code, as the professor pointed out.
Tejas - I improved my presentation skills and significantly enhanced my LaTeX and C++ coding skills.
Kushagra - I learned new things in the field of topology and algorithms and gained respect for other research fields as well.
Vaibhav - I improved my communication skills by engaging with a variety of people. In my branch, there are no collaborative assignments, so it was a refreshing change and a great experience.
How does an offline internship compare with an online one?
If research experience were only about academics, there wouldn't be much difference between an online and offline internship. An offline internship offers great potential for social development, teaching you how to maintain professional relationships and adapt to work culture. You are not always bombarded with work, and there is greater appreciation for your efforts. An offline internship is also a good way to convince your parents for a vacation. We visited neighbouring countries like Monaco, Switzerland, Italy, and Vatican City for sightseeing, an experience not possible in an online internship.
Were the submissions and deadlines quite challenging?
Kushagra and Vaibhav - We worked from 9 A.M. to 6 P.M. with two hours for lunch. We had only two strict deadlines and occasional presentations. The internship offered significant flexibility. In our free time, we enjoyed playing in the music room and having fun.
Tejas - Professors did not expect us to work after 5 P.M.
How was the internship experience in general?
Vaibhav - The provided accommodation was quite good, with a clean microwave, weekly room service, and subsidised food. The academic experience was significantly different from what we were used to at IIT Delhi.
Kushagra - Here, we often rush to complete courses and assignments. At INRIA, the pace was slower, allowing us to apply the topics better and with less stress.
What was the most memorable thing about the internship for you?
Kushagra - The tourist experience was great. Despite being a vegetarian, I managed to find good food. We also interacted with students from IIT Bombay and IIT Kanpur and learned about their campus culture. We made many new friends as well.
Vaibhav - The evening coffee breaks were memorable. We discussed our research problems and cultures. There was a Dutch post-doctorate who had great knowledge about Indian history, leading to insightful discussions on philosophy and ethics.
Finally, what advice would you give to intern hopefuls?
Suren - There is no right time to start contacting professors. It's better to begin as soon as possible. Use websites like csrankings.org to identify professors based on your interests. You are not expected to have an in-depth knowledge of their research papers, so don't be discouraged from applying to a subject. Don't hesitate to contact professors, and don't expect any prerequisites.
Tejas and Kushagra - Specifically for dual-degree students, if you plan to pursue a third-year research internship, it's advisable to take your second-year research internship seriously and aim for a publication or a letter of recommendation. For BTech students, there's little time between completing a second-year internship and the third-year internship tests. Allocate about 1.5 - 2 hours every day during the summer to prepare. During the application process, consider taking relevant courses to stand out. Focus on your institute courses as well, as there's limited time for additional courses.