{"id":17816,"date":"2016-12-27T11:18:31","date_gmt":"2016-12-27T11:18:31","guid":{"rendered":"http:\/\/www.biphoo.com\/bipnews\/?p=17816"},"modified":"2016-12-27T11:20:59","modified_gmt":"2016-12-27T11:20:59","slug":"apple-artificial-intelligence-team-publishes-its-first-research-paper","status":"publish","type":"post","link":"https:\/\/www.biphoo.com\/bipnews\/technology\/mobile\/apple-artificial-intelligence-team-publishes-its-first-research-paper.html","title":{"rendered":"Apple Artificial Intelligence Team Publishes Its First Research Paper"},"content":{"rendered":"
\n

Apple Artificial Intelligence Team Publishes Its First Research Paper<\/span><\/h2>\n

Apple’s Artificial Intelligence Team Publishes Its First Research <\/strong>Paper: –\u00a0<\/span>Earlier this month, Apple announced that it would allow its artificial <\/span>intelligence, researchers, to publish research papers \u2014 a major shift in the notoriously secretive company\u2019s policy. Now, just a few weeks later, the first of these papers has been made public on the preprint server arXiv.<\/span><\/p>\n

The paper \u2014 titled “Learning from Simulated and Unsupervised Images through Adversarial Training”\u2014 deals with intelligent image recognition technology. Specifically, it describes a technique that would enable a program to recognize and decipher computer-generated images.<\/span><\/p>\n

So far, this has not been possible because, as the researchers from Apple note, \u201csynthetic data is often not realistic enough,\u201d and increasing the realism is \u201ccomputationally expensive.\u201d<\/span><\/p>\n

\u201cWith recent progress in graphics, it has become more tractable to train models on synthetic images, potentially avoiding the need for expensive annotations. However, learning from synthetic images may not achieve the desired performance due to a gap between synthetic and real image distributions,\u201d the researchers wrote. \u201cIn this paper, we propose Simulated+Unsupervised (S+U) learning, where the goal is to improve the realism of synthetic images from a simulator using unlabeled real data.\u201d<\/span><\/p>\n

In order to do so, the researchers used a technique known as adversarial learning, wherein two competing neural networks basically try to outsmart each other. In this case, the two neural networks are the generator, which, as the name suggests, generates realistic images, and the discriminator, whose function is to distinguish between generated and real images.<\/span><\/p>\n

\u201cWe develop a method for S+U learning, which we term SimGAN, that refines synthetic images from a simulator using a neural network which we call the \u2018refiner network,\u2019\u201d the researchers wrote in the study. \u201cTo add realism \u2013 the first requirement of <\/span>a S+U learning algorithm \u2013 we train our <\/span>refiner network using an adversarial loss.\u201d<\/span><\/p>\n

The paper is likely to be the first of many to be authored and published by Apple\u2019s machine learning researchers in the coming weeks and months, now that the tech giant has lifted its restrictions on sharing research. Reports suggest that this decision was taken primarily to recover the ground Apple has lost in recent years to the likes of Google, Microsoft, Facebook and the Elon Musk-backed startup OpenAI, insofar as AI-related research is concerned<\/span>.<\/span><\/p>\n

source:-http:\/\/www.ibtimes.com\/apples-artificial-intelligence-team-publishes-its-first-research-paper-2465683<\/span>.<\/span><\/div>\n

\n\t\t\t
\n\t\t\t <\/div><\/div>\n\t\t