Computer Vision News - November 2020

34 Shiry Ginosar Jitendra is also very intuitive. He's very good at saying, “This is how it works.” He understands the high level idea of how something very technical works. He's also an amazing encyclopedia. He knows all the papers, all the topics, and who wrote what, like 40 years ago. He knows how to connect things. We were recently talking about transformers. He walks in, and he's like, “Yes, this is exactly like anisotropic diffusion. It’s exactly like this, or it’s exactly like that. This is the whole historical current that led to this cutting-edge thing that is happening today.” He's also very, very technical. He will continue to be an inspiration for me forever, especially as the deep learning revolution happened. Before that, he would say, “Deep learning mumbo jumbo. Whatever, it's never going to work!” Then he saw that it works. Then he took several months. He sat down and read every single paper there was to read. He brought himself to the cutting- edge forefront by decision. They went and wrote R-CNN, and from then on, he has been leading the field. His strengths are that he's very agile, pragmatic, and technical. His memory is just amazing! I think that’s part of why he's such a great researcher as he actually knowsto connect all of these different things together. He remembers everything that happens and how they're all connected. What does it mean for you to have your work displayed to the public, such as the exhibition at the Israeli Design Museum? What I push for, obviously, is to have my work displayed, scientifically, technically in conferences and in person, in workshops, and so on. That's the work I do. I think that is the most important part of what we do as scientists is to kind of open up our processes, and our code, and our data, and everything to everybody else in the community so it can be replicated and learned from. A secondary way that different parts of my works have been exposed was through different kinds of media articles. For example, we were published in the New Yorker. This gives people that are outside of the community a peek into what we're doing. The kind of work that we do is trying to show that we're not done. AI is not solved. It's not a good idea to invest in the AI bond market yet. We have many things that are not solved. I think that's interesting to the general public, and I think it's important for the general public to know. That’s inspiring! Yes! Also, we had five different people who emailed us who couldn’t move for different reasons, such as being paralyzed. They asked, “Can you please make me move even on the screen?” You can connect with so many more interesting ideas and inspire people in the way that you wouldn't necessarily just from a poster. It’s a whole different world!