Computer Vision News - May 2018

Laura Waller is an associate professor at UC Berkeley in electrical engineering and computer science. Laura, tell us about your work. I work on computational imaging, which is the joint design of optical systems and computational algorithms to design systems that use hardware and software together efficiently. You don’t just design an optical system, take the data, then get somebody else to do the image processing. Nor do you write up the map, figure out the best imaging system, then try to build it. You go back and forth between what’s practical and cheap to build and what is convenient and efficient to do computationally. Replace some of the optical functions with computation, but only where it makes sense. Among all the fields that a young scientist could choose, why did you choose this one? In my undergraduate, I was an electrical engineering and computer science major and I was focused on the signal processing track. Then I randomly worked in an optics lab one summer, because I wanted to go home and it was in the town. [Laughs] It just seemed really beautiful to have all this physics that matches the theory that I had learned. In my graduate career I learned more optics and started doing this combination of the two, and it was the perfect fit for my interests and skills. And you do not regret it… Yeah, that’s right! [ Laugh ] Do you think it’s a field where one individual scientist can have a big impact? Computational imaging is relatively new – it’s only the past 10 years or so that it’s been becoming popular – so it’s a good time to be in that field. It’s a very creative and open field, where it’s very much about working on design, so you have a lot of free parameters. Figuring out what problem to solve is more of the research than actually solving the problem. It’s a very open-ended field and you have a wildly broad set of different applications, from biological microscopy to imaging black holes to medical imaging like MRI and such. Some students will be attending the conference for the very first time and may not have chosen their specialist field yet. Do you have any advice for them? Field-wise, I think people find a lot of Laura Waller 32 Saturday ISBI DAILY Keynote BEST OF