Computer Vision News - December 2018

Almost everyone gets it immediately that Tabasco is hot. If you think about it, really, really carefully, you could say, “ Well, fire extinguishers are really for putting out heat so doesn't this say that Tabasco is not hot? ” No, it doesn't mean that. So it could miss the point completely! This is true for all metaphors. You have to decide what property is being transferred. The textbook example of a literary metaphor is “ the classroom was a zoo ”. So what property of a zoo is being transferred? It means that the kids were out of control and wild like animals, but it could equally well mean that you had to pay $12.50 to get in or that the children were in cages being abused. All sorts of things that could apply, don’t. It is unusual with the Tabasco case that the fire extinguisher is the opposite of being hot, but still, it is sort of a symbol of being hot. Most symbols are highly ambiguous. They need a little bit of context and a little bit of background knowledge to ground them at all. This is an amazing part of the human language’s ability to communicate. Whereas, nothing we say is perfectly 100% clear and a lot of the time we can fill in the gaps with our background knowledge. As long as 50% of people get it, and there's no other meaning that is too threatening to it, it’s pretty good. If you look into any metaphor or anything really too hard, you’ll find the flaws and the point where it cracks. No one really has time to think through all of this. It’s about those first impressions. You have spent the biggest part of this century in academia: Washington, Stanford, MIT and Columbia. What is that keeps you in academia? I have done internships at Microsoft and at Google, sort of on the research side. I like to pick problems that will take five to ten years to do well. It’s hard to see the immediate business value of it. Understanding visual metaphors and creating them is not something that I can just go to a company and say, “ Hey, I want to spend my time doing this. It’s going to require careful study, and it’s not going to be useful for quite a number of years. It may not work at all. ” I’m attracted to those kinds of problems. I like things that I don’t know how to do, 34 Women in Computer Vision Computer Vision News Women in Science “Your brain is like: ‘Huh, something is wrong there’!”