Computer Vision News - August 2018

An interview with Francois Chollet on PyImageSearch: Computer Vision News publishes many interviews. We think that we know how to do that and we enjoy very much doing that. But we recognize when others do a great job too. Read the interview we would have liked to do… Facebook AI Research ExpandswithNewAcademicCollaborations: Another major news from Facebook : Yann LeCun announced that Jitendra Malik joined from UC Berkeley to lead FAIR out of Menlo Park. Jessica Hodgins will lead a new FAIR lab in Pittsburgh. Both retain dual affiliation with their university. Read the other additions! Lethal Autonomous Weapons Pledge! Let’s start with a positive message, about the role that AI might play in the future with regards to military systems : lethal autonomous weapons selecting and hitting targets without human intervention. Many of us have read leaked memos involving key scientists and corporations. Hundreds of companies and thousands of individuals say NO: among them, friends of our magazine: Yoshua Bengio , Jürgen Schmidhuber , M. Musa , Eyeris and many more! Other stories you shouldn’t miss: AutoGraph converts Python into TensorFlow graphs Google Glass is back - Nowwith AI! What ML means for software development Video: China - Artificial intelligence better than human doctors Intro to Keras Layers Why Businesses Fail at Machine Learning: Yes, some do and for many reasons. Find them out in this nice article by Google’s Cassie Kozyrkov on Hackernoon . Our comment is that when things are becoming too complicate, don't bother with algorithms and R&D : outsource your project and RSIP Vision will do all you need in the best way. Read Now! Towards a Monopolization of Research in Artificial Intelligence? And since we decided to talk about good causes, here is another subject which worries Yoshua Bengio and even UNESCO : the concentration of AI power in few hands might endanger both democracy and the economy . Read More! GoogleAIblog:UpdatetoOpen Images-NowwithBounding-Boxes Last year Google introduced Open Images , a collaborative release of about 9M annotated images over 6K categories , as a useful dataset for machine learning research. Google’s Vittorio Ferrari just posted to announce an update to Open Images with millions additional images . Read more… Spotlight News 11 Computer Vision News