Computer Vision News - August 2018

Anand Joshi is a principal analyst with a focus on AI and computer vision at Tractica. He is also an AI strategy consultant, advising many startups and Fortune 500 companies, particularly in reference to hardware, computer vision, AI strategy, video analytics, 3D and related technologies. You are in the ideal position to see current trends in the computer vision arena. Can you tell us what impressed you the most in the first six months of 2018? Back in 2014, I described computer vision as analogous to rookie quarterback who played well in college, and he was just about to enter the National Football League. Everybody had high expectations. I think, if you look at 2018, I can say that those expectations are coming through, and we are starting to see computer vision taking more and more center stage and going into more and more applications. In particular, automotive is a very sweet spot. A lot of companies are actively working on research projects including Level 4 and Level 5 autonomy, which is somewhat still far. With 1, 2, and 3, we are getting there. A lot of companies are shipping their products. There is a good market traction. Robotics is another area where there have been announcements and products introduced. About drones, there is a decent volume in the sense that a lot of DJI drones actually use some sort of navigational capabilities enabled by computer vision. There is a surveillance market, which has always been there, but it is now being enhanced by computer vision. I would say that 2018 is the year that computer vision has started to go from a rookie quarterback to a professional quarterback, in the sense that there are a lot of dollars going in. We will see more and more applications and dollars coming into the picture. One of your favorite questions: who are the winners and who are the losers? The de facto winner right now is NVIDIA. There is no question. NVIDIA has made significant investments in the chipsets as well as the whole infrastructure that you need to create the computer vision applications. Not far from NVIDIA is also Intel who have been acquiring a lot of companies and investing quite a bit to enable computer vision applications. There is, of course, companies like AMD, Qualcomm and the like. They are investing, although they are lagging somewhat behind Intel and NVIDIA in terms of their product features. A lot of different companies and startups appear all across the board who are essentially creating a point solution, if you will, in a particular niche. Over time, we are going to start seeing them grow to other areas and start seeing some dominant players. 28 Anand Joshi - Tractica Guest Computer Vision News Guest