Computer Vision News - August 2020

access. At some point, patients have to be willing to consent to give data. There have to be people in hospitals who actually make the effort to create these datasets. It sounds so easy when you explain it. Science needs it! Progress needs it! But would you advise your parents or grandparents to share their data? Um, it depends. For some data, I think there is no risk at all. The reconstruction of the ID or some hidden information is very low, but for some data I wouldn’t. It really depends on the dataset. Right now, if we share our own data, if we just give it away, we don’t have any information about how it is used. What would be super nice, and I’m continuing to do the research for this, is if the data actually explains the impact. We are tracking how long we sleep, how long we run, who we are related to. We could create a network where our data donors will actually see what we achieve all together by giving this data. Our readers have read interviews with other people at NVIDIA. I interviewed Jan Kautz a couple of years ago and Chip Huyen, who was at NVIDIA when I interviewed her. Also, Sanja Fidler , before she joined the company. Tell us something about NVIDIA that we don’t know. I don’t think it’s a coincidence that you interview several women from NVIDIA. NVIDIA is very supportive. What you probably don’t know is that our team is roughly 50/50, 50% women and 50% men. This actually creates a really nice community. Did they create the proportion on purpose or did it happen by chance? NVIDIA encourages women in science, STEM, mathematics, and technology, but there is no quota. There are a lot of people within NVIDIA that are very open minded and very supportive within this community. You changed your focus over the years. Women in Computer Vision 42 Best of MIDL 2020 “When I talk to researchers, I want to learn from them.” 0