Computer Vision News - December 2016

Computer Vision News lists some of the great stories that we have just found somewhere else. We share them with you, adding a short comment. Enjoy! Now Machines Recognize Something After Seeing It Once We are used to think that algorithms need thousands of examples to learn something. Which can be quite annoying when time and data are scarce. It seems that researchers at Google DeepMind found a way around that. Read about it , get the paper , watch the video Computer Vision News Spotlight News 38 How Machine Vision Is Changing Sports Spectating Those who follow sports events are well aware of the increasing use of technology to boost viewer’s experience, improve athletes’ training and make judging more accurate. Think at Hawk-Eye in tennis and goal-line technology in football. This nice article tells you more about machine vision ’s impact in this area, including Intel’s spectacular 360-degree replay of MLB action. Read… AI Curated a Magazine Using Image Recognition Among the many things that a machine can do with a picture, here is a new one: decide whether it is aesthetic or not. Not like an unskilled viewer would do, but like the 18 million expert photographers members of the EyeEm community, whose magazine was curated by an AI trained to understand moods, feelings and beauty. Read... Automated Inference on Criminality using Face Images Two Shanghai Jiao Tong University researchers claim that their study is the first to offer automated inference on criminality based solely on still face images. They do it exactly as you would expect: machine learning , classifiers, a dataset of 1856 real persons controlled for race, gender, age and facial expressions. Half of them were convicted criminals. It seems that variation among criminal faces is much greater than that of the non-criminal faces. Read… Trainspotting via the Caltrain Rider How many of our readers have taken a ride on a Caltrain , between SanFran and the Silicon Valley? Probably most of them. And how many of them think that delay estimates provided by Caltrain are reliable? Probably many less… Engineers at a Silicon Valley firm used image processing and recognition to improve arrival predictions. Read… The Motley Fool tells us what is cooking at Apple. 3 revelations found looking into Apple Job Listings! Read…