Computer Vision News - February 2020

2 Summary Artifici l Intelligence 2 AI Year 2019 in Review by IBM Research We liked very much this very nice review of key topics and notable papers in 2019 by IBM Research . It's an excellent analysis of what was notable during the year: a growing focus on making trustworthy AI work in practice, creating new AI engineering paradigms to scale AI for a broader use , and continuing to advance core AI capabilities in language, speech, vision, knowledge & reasoning, human- centered AI, and more. You should read the article: it is written by John Smith , an IBM Fellow and an old friend of our magazine. Read... An algorithm that learns through rewards m ay show how our brain does too Like animals learn through punishments and rewards (remember Pavlov?), it seems that computer couldcontinuously learn through similar reinforcement to solve a virtual maze. As reinforcement learning continued to mature, it in turn helped the field of neuroscience discover those mechanisms, feeding into a virtuous cycle of advancement between the two fields. In a paper published on Nature, DeepMind has once again used lessons from reinforcement learning to propose a new theory about the reward mechanisms within our brains. Read... Neural Data Server: an AI-Powered Search Engine for AI Neural Data Server (NDS) is a large-scale search engine for finding the most useful transfer learning data to the target domain. Introduced by awesome Sanja Fidler (with Xi Yan and David Acuna), Neural Data Server consists of a dataserver which indexes several large popular image datasets, and aims to recommend data to a client, an end- user with a target application with its own small labeled dataset. These fine UofToronto folks show the effectiveness of NDS and demonstrate state-of-the-art performance on several target datasets and tasks such as image classification, object detection and instance segmentation . Read... Computer Vision News has found great new stories, written somewhere else by somebody else. We share them with you, adding a short comment. Enjoy!