Computer Vision News - September 2021

real problems. This is truly a real problem that we need to address. It has an easy valuation. I don’t have robots that can feed people today, so I can see that it doesn’t exist. It’s really exciting to push in this area. How do theoretical researchers put in so much effort but never really see the fruits of their labor? Actually, my husband is in theory. He works in areas that are exactly like that. [ laughs ] They sometimes just prove the existence of things. They don’t even have a constructive way of finding it. But it’s an interesting puzzle. It’s interesting to push on that puzzle even though you might not see the effects of that in an application today. There’s an excitement over long term effects. If I solve this puzzle, maybe in 20 years, 40 years, it would actually be used in practice. That’s true in a lot of algorithms that you see In which kinds of situations would anyone need that? In general, the core application for that is in an assisted setting for people with disabilities who cannot feed themselves. They often use robot arms on a wheelchair. This robot arm can help them on various types of tasks, like feeding themselves or putting clothes on. All of these are pretty challenging tasks, and the person still has some control over the robot. But it’s really difficult to use a joystick or to use the devices that patients with disabilities have, like sip and puff devices to control these robot arms. You need to have good algorithms that are intuitive to control and easy to operate to help the person. It seems like you really want to leave your mark on what you are doing. As I said, I personally get motivated by 32 Women in Science