Computer Vision News - October 2018

Now, with 786 papers, accept it is more difficult! Exactly. Now, it is more difficult. I’m also older and don’t have as much energy. I have a second trick, which is I find some of the junior students in my lab and I ask them: “ What are the posters that you think I would like? ” They are the explorers for you! Exactly! They already have a good model of what kinds of things I like, so they tell me where I should go and which posters I should see. Can you tell us about one paper at this conference that particularly impressed you? We’re still at the beginning, so I’m not willing to point to a particular paper. I could say already that I like that there is finally a lot of work starting to happen that is moving away from direct supervision. People are interested in doing things that are self- supervised, unsupervised, or somehow moving away from, “ Let’s just label those things and go for it .” I’m very excited about this trend. I think there is a lot of work to be done in this direction. We are definitely not there yet, but I’m happy that a lot of people are starting to work like that. Would you agree that it’s a very positive thing that there is an increase in industry involvement at this year’s conference? I think it’s wonderful that now vision is actually useful to industry in the real world. When I started, nobody cared about vision because it didn't work. Now, this is a testament to the fact that it’s starting to work. This is great. It’s also great that a lot of companies are starting research labs and starting to be interested in, not just development, but also doing research and publishing papers. It does make for a more crowded conference. This explosive growth is kind of the downside of it, but it’s the price we have to pay. Industry is also taking some of the best minds out of academia. There is a competition for talented engineers and scientists which could cause an issue for academia in the medium to long-term. Do you think that is another downside? I think, in the short term, this is great. It’s fantastic that some of the best and brightest people go off to industry and make a difference in the real world. I do worry about the medium to long- term. We could be in the position where we run out of teachers and advisors to bring up the next generation. Then these same companies will be in trouble because they will not be able to hire new software engineers. We’re already seeing a situation now where a lot of the young junior researchers don’t have a solid foundation in classical work. A lot of people have never heard of David Marr. A lot of people have never heard of Jan J. Koenderink. A lot of people haven’t read the classics. Daily Wednesday Alyosha Efros 13 “… these same companies will be in trouble because they will not be able to hire new software engineers!” Research