Computer Vision News - March 2022

18 Computer Vision Book reference book that most computer vision engineers and researchers would have on the side. It’s just there if they want to know a bit more about a particular topic or which papers to read. It’s a perfect starting point for people already doing computer vision who want to get into a subject they don’t quite know yet because it’s peppered with references and citations. Do you have any other books in you? Some students, like undergraduate students, find all the references distracting, so a shorter book with fewer references would make the ideal textbook for them. But I have never had the energy to write it! I’m hoping someday a professor will step up and say, “ Okay, let’s do that! ” Computer vision is such a broad field. Do you have any tips for people new to the community who may be finding it hard to navigate? If you’rea student, oneof thebest things you Have you ever received a suggestion that gave you a real eureka moment? The help I get is at a variety of levels. I have long-time collaborators like Sing Bing Kang and Daniel Scharstein, who read chapters or co-author things. I was fortunate that in the class I was teaching at the University of Washington, an undergraduate called Matt Deitke would ask excellent questions. It became apparent there were parts of deep learning, such as LSTMs and transformers, that he understood a lot better than I did. He gave me pages worth of comments, and at some point I said, “ Would you be willing to write a section of the book? ” He agreed, and there’s a whole chapter on the more advanced topics of deep learning that he wrote. That was my biggest aha moment! Do you knowwho the people downloading the book in their thousands are? It’s used as a textbook for classes, so lots are undergraduate and graduate students. It’s a