Maneesh Agrawala: 30+ Years of Computer Graphics
by Anika Rai '23 for Silver Quest
“I didn't realize just how special it was,” Maneesh Agrawala, ‘90, says recounting his experiences in the Blair Magnet Program. “The ability to have depth in subjects that I really was interested in, computer science and science, those things were incredibly helpful to me in undergrad, grad school, and continue to be in my work today.”
Today, Agrawala is the Forest Baskett Professor of Computer Science and Director of the Brown Institute for Media Innovation at Stanford University. The work he does in his lab cuts across “three main areas, which are computer graphics, human computer interaction, and visualization,” Agrawala explains. “A lot of the tools that we've developed have been aimed at various kinds of visual content, which directly relates to my PhD in computer graphics. But, more and more, we're doing things outside of visual content.” This includes making tools to create audio or physical objects.
“[With] most video editors, even today, what happens is that you have to spend a lot of time scrubbing the timeline, right? You get all your raw video, you capture it, and then you have to find the parts of it that you want to edit into the final. And that's, that's just a ton of work.” Instead, Agrawala and his lab started developing tools where to obtain high quality transcripts of audio, and time align it to speech in a video. “That already is useful,” he says, “because with that alignment, you can read the transcript, reading is way faster than scrubbing, and watching. And, you know, you can click on a word to jump to that part of the transcripts, and navigation becomes a lot easier.”
Going even further, the lab worked on applying cut, copy and paste to the text and editing the underlying video. They have applied this to audio as well. “When people think about that, they often think about the raw video or audio in terms of the speech, not in terms of the timecode or the specific point in time you are in the video, right? And so it just fits better the way they think about the underlying material.”
Taking a trip down memory lane, Agrawala explores the path he took to get to where he is today. “I was always kind of interested in math and computer science. My dad, being a computer scientist, I think helped with that. And the Magnet program was fantastic for letting me kind of pursue those directions in great depth,” he reiterated.
After graduating from Blair, he pursued an undergraduate degree in mathematics at Stanford University. “I was still looking around for what to do and what I might be interested in… so I started taking some computer science classes at Stanford, and really just loved computer graphics. So I took the first course on that, and basically decided that I needed to pursue that further,” Agrawala acknowledges. “So I finished undergrad, I got the degree in math, but really, for a lot of the last part of that I was basically taking computer science classes and more specifically computer graphics.” He went on to complete his doctorate in computer science under Pat Hanrahan and after graduating, worked at Microsoft Research for three years, before joining the faculty at the University of California, Berkeley.
The flexibility of being a PhD student is what led to Agrawala’s interest in research even in the computer science industry. “ I liked the ability to synthesize ideas from, you know, various places or develop new ideas that might address various kinds of problems,” he says. Today, “we’re very interested in what people are doing in terms of making various kinds of media … There's a lot of video making and sharing going on, on social platforms so we’re interested in understanding that process and trying to see if we can make it better and how we might intervene to make it easier.”
Back to the program, Agrawala emphasizes how formative it was to him that the Magnet was not only other students in the cohort, but also that they were embedded in a “regular” high school. “The ability to interact with students that were outside of the magnet was really influential for me.”
During his time at Blair, Agrawala competed in SuperQuest, a program intended to stimulate the integration of computational science and its tools into high school science. “The company was asking students to work in a team to propose various projects,” he explains. “There were four people on a team, and each one proposed their own project. And the finalist teams all went to Minnesota for the summer to work to implement their own project on the supercomputer.” It was truly a formative experience.
“The education that this public school could provide was just unbelievable,” he says referring to the quality of teachers and classes at Blair. He advises students to take advantage of what the program can provide. Agrawala recommends focusing and delving deep into a few topics rather than spreading oneself too thin. “It is important to try and identify the things that you really care about.” He leaves students with some notable advice: “One of the challenges is that it can be easy to be interested in too many things. Try your best to figure out what you're most interested in and then try even harder to pursue it.”