CVPR Daily - Tuesday

DAILY T u e s d a y Trevor Darrell 10 Trevor, it is an honor to have you with us. We have featured your research in our magazine before, but this is our first interview with you. Can you tell our readers about your work? I’ve been working in computer vision for as long as I can remember and feel fortunate to be part of a field that is constantly evolving. Research and computer vision touch on so many fundamental things that relate to understanding intelligence, the brain, thinking and perception, which are all these grand, fantastic goals, but at the same time, it involves concrete and practical things too. Can I recognize this piece of fruit? Can the robot pick up this block and decide whether it will balance on this stack? That’s what has kept me interested and excited all this time. I can remember the first CVPR I went to back in 1988. The funny thing is, you could take some of the titles from that year’s conference and mix them in with some of the titles from this year’s conference and people would be fooled. Some of the ideas are still applicable. Of course, if you looked at the paper, there would be no experiments on GPUs, so that would be a dead giveaway! Like many other events around the globe, this year’s virtual CVPR will be a different experience to the physical conference we are all used to. Do you have any thoughts on the situation we are all currently facing? I would say, let’s all get through this together and figure out away to continue to do great research. The research process hasn’t significantly slowed down. We’re all still productive in a virtual mode, and a lot of what we do to learn is about reading on arXiv and in blogs. If anything, I think this is a moment to take a break from all the professional stress we’ve been under and take some time to relax and focus on things outside of work. If we don’t have to travel and go Trevor Darrell "One crazy suggestion I have is to remove the absolute requirement for novelty!"