Computer Vision News - March 2022

19 Computer Vision: Algorithms and Applications students work so well together. When I moved to Seattle in 1995 to work at Microsoft Research, they had a connection with the university and encouraged us to work with people there and co-supervise students. I met people who became lifelong collaborators, like David Salesin, Steve Seitz, and Brian Curless. For me, the biggest joy in doing computer research has been working with smart people and up-and-coming students. Just that sheer joy of coming out of a meeting with ideas and excitement of things I hadn’t thought about before. That’s what makes the University of Washington very special. What’s your message for the community? I’m looking forward to seeing all my friends andtheyoungerpeopledoinggreat research in person once the conferences restart. I’ve retired now, so I’m in a position where I can look back over the last 40 years, and I’ve had a wonderful career. The field has exploded. We can do things now that we couldn’t back then. The conferences were about 300 people when I started, and now they’re over 5,000. Computer vision has become useful in practice. I love it because it combines visual elements with deep mathematics. I’m so excited to share what I know and love about the field with others! can do is sign up for a class at the university or talk to the professor who does computer vision and see if you can do a research internship. Most of us who teach the class believe you can easily shape a complete degree around just computer vision. There are so many different topics. If you want to understand the whole thing, you have to take many classes. There’s no way to learn computer vision without putting in a lot of effort. You can take introductory classes, online courses, and do some of the great tutorials in deep learning associated with PyTorch and TensorFlow. Dive Into Deep Learning by Zhang and co-authors at Amazon has inline code snippets. That’s an excellent way to teach yourself. Looking back over your career to date, can you pick one thing that has particularly impressed you? Boy, that’s hard because I would go to conferences most years and come away with pure delight at something that happened! My PhD was in computer vision, but I’ve always loved computer graphics. The 3D graphics that Pixar produces now came out of people like Ed Catmull, whowas publishing papers in the ‘80s on computer graphics. I was still working on my master’s degree at the University of British Columbia, and we drove down to Seattle in 1980, where the SIGGRAPH conference was being held. They showed this short video called Vol Libre, flying around a fractal terrain. It was just amazing because we’d never seen anything like that! You have spent so much of your career at the University of Washington. What has kept you there so long? The computer science department and the Graphics and Imaging Lab (GRAIL) are extremely collegial. The professors and