Vale ’18 Named Regeneron Science Talent Search Finalist

Regeneron Science Talent Search named senior Vinjai Vale a finalist this January. The Science Talent Search, with several Nobel Prize-winning alumni, is considered one of America’s most prestigious math and science competition for high school seniors.

"Vinjai is remarkably good both at doing terrifically creative, interesting research on his own and at working with other people to accomplish common goals such as running the Exeter Puzzle Hunt so that PEA student teams can have fun together. He is always thinking and a lot of that thinking is about other people.”

Over 1,000 students compete each year. Students are chosen based on a personal research project of their choice. In early January, the search named 300 applicants as Scholars, then reduced the pool to 40 finalists, including Vale. These finalists are invited to the final round of judging in Washington, D.C., where they will compete for a total of $1.8 million in awards; there, they will have the chance to display their work to the public and meet with other notable scientists.

Vale focused his research on artificial intelligence, specializing in computer vision and how programs process images. He worked to improve the capacity of current methods so that computers can analyze complex scenes closer to human accuracy. “The current state of the art paradigm for doing this kind of computer vision is using something called a neural network,” Vale said. “They’re really good at [classification], but they often struggle to represent the spatial composition of these objects, which comes naturally to people.” To tackle this problem, Vale borrowed insights from psychology and neuroscience and applied these techniques to computer science.

This problem represents a long history of interest in computer science for Vale. He met his mentor at MIT PRIMES, a high school research program for math and computer science. The two began working on a project similar to what Vale eventually ended up researching for the Talent Search. However, he faced setbacks. “I spent a few weeks to just take a step back and think about where what I was doing was going wrong and reading some papers,” Vale said. “That’s how I formulated this new problem, and that’s what I worked on from August to now.”

Vale described his process of research as “going into uncharted waters.” In the fall, he woke up at 5:30 a.m. to work for two hours every day before his morning commute to Exeter. “You have to be up to date with the latest research, so you have to be reading papers very often,” Vale explained.

Computer Science Instructor Sean Campbell praised Vale for his passion and dedication. “I met with him roughly twice a week last spring term, but this was more often him showing me how his work was progressing and evolving,” he said. “Vinjai has a strong work ethic combined with the ability to find novel approaches to problems.”

A big part of Vale’s project was simply thinking—something easier said than done. Vale would break down his main research problem into smaller sub-problems so he could attack it from different angles. “[It’s] a very immersive process. There would always be some problem that I was thinking about in the back of my head,” he said. “I think that’s how you make progress, just thinking about it a lot and trying different things because most of them don’t work.”

Biology Instructor Townley Chisholm commended Vale’s research. “Exeter has a long history of astonishingly gifted science and math students; Vinjai is one of the very best,” he said. “Vinjai is remarkably good both at doing terrifically creative, interesting research on his own and at working with other people to accomplish common goals such as running the Exeter Puzzle Hunt so that PEA student teams can have fun together. He is always thinking and a lot of that thinking is about other people.”

“He’s such a hardworking student,” upper Jenny Yang, who has worked with him in the Computer Science Club, said. “He helps younger students with difficult algorithms and pursues his own research outside of school and the club. He has been a great role model for the past few years, and I have learned a lot from him. Not only in terms of science knowledge, but also from his time management and efficiency.”

Campbell applauded Vale’s achievements, saying, “Apart from the direct opportunities afforded by his selection as a Regeneron finalist, [this experience] also sets him apart from most students his age. He is working on graduate-level research as a teenager under the direction of an expert in the field.”

The process, though arduous, confirmed Vale’s passion for research and computer programming. “Doing research is like a rollercoaster. You have lows that last a really long time and you just feel like you’re going nowhere. But then something clicks and you come up with an idea and it opens up a whole new field of questions,” he said. “And that’s something I want to continue doing probably for the rest of my life.”