Computer Vision News - February 2017

that we want work with. The second direction is building novel imaging systems for underwater research that don’t exist yet. The main example is underwater microscopes [ see cover ]. I concentrate on in-situ imaging so algorithms will work on images taken in-situ, as opposed to lab experiments and so on. We aim to validate everything through experiments in diving, collecting data by ourselves. We have underwater cameras. We design the experiments and collect the data. We also design the hardware and the imaging setup for the specific algorithms. I mentioned that we also build imaging systems. If we cannot use it from off the shelf, we also build the systems ourselves for underwater use. The experiments are done either by diving or deploying the systems from boats or underwater robots. Do you dive yourself? Well, now that I am faculty I don’t have so much time! When I was a PhD student and a post-doc I went diving a lot. Now I dive less, and the students do most of the diving. You should have never made your PhD then! You should have remained a PhD student… [ laughs ] One of them said that. It’s not worth to be faculty, if you don’t have time to go diving… That is probably true. How did you find yourself as faculty? When I graduated my PhD, I went for a Postdoc in San Diego. I was there for four years. Half of the time, I was in the Department of Computer Science. The other half, I worked in the Scripps Institution of Oceanography. I started looking for faculty positions in Israel. It wasn’t easy, as the marine technology field was not (and still isn’t) very developed in Israel. When people heard what I was doing, it sounded a bit esoteric [ laughs ]. I was in touch with a few engineering departments. I also visited the school of marine sciences. Back then the Department for Marine Technologies didn’t exist. Then one day they emailed or called me in San Diego and said that they had received a large donation to establish a marine technology department. It was perfect timing! It was exactly when I needed it. Can you tell us more about your work involving observation of the sea and ecology? The computer vision algorithms that I’m developing are applicable for many applications. They improve visibility. You can use them for defense, surveys, and so on. There are some things that are more specific for ecology. For example, I work a lot with corals because it’s a fun topic. They are beautiful, and I love to go diving there, that’s also important! 14 Computer Vision News Women in Computer Vision Women in Science “ I work a lot with corals because it’s a fun topic ”