<?xml version="1.0"?>
<rss version="2.0" xmlns:media="http://search.yahoo.com/mrss/" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:neowin="https://www.neowin.net/">
	<channel>
		<title>Neowin News Feed for: Auto-crop</title>
		<link>https://www.neowin.net/news/tag/auto-crop/</link>
        <atom:link href="https://www.neowin.net/news/rss/auto-crop/" rel="self" type="application/rss+xml" />
		<description>Neowin News Feed for: Auto-crop</description>
		<language>en-us</language>
		<generator>Neowin Ignition News</generator>
		<managingEditor>editor@neowin.net (Managing Editor)</managingEditor>
		<webMaster>developers@neowin.net (Neowin Developers)</webMaster>
		<ttl>5</ttl>
		<image>
			<title>Neowin.net</title>
			<url>https://www.neowin.net/images/pegasus/icon.png</url>
			<link>https://www.neowin.net</link>
		</image>
		        <item>
            <title>Twitter deploys neural network to improve auto-cropping of images</title>
            <link>https://www.neowin.net/news/twitter-deploys-neural-network-to-improve-auto-cropping-of-images/</link>
            <description>&lt;div style="float:left;margin-right:10px;"&gt;&lt;img src="https://cdn.neowin.com/news/images/uploaded/2018/01/1516966827_twcrop_medium.jpg" alt="" /&gt;&lt;/div&gt;AI or human eye; Twitter has been training a neural network to impersonate the human eye, recognising faces, animals, text and other objects to tailor image previews and avoid awkward auto-cropping. &lt;a href="https://www.neowin.net/news/twitter-deploys-neural-network-to-improve-auto-cropping-of-images/"&gt;Read more...&lt;/a&gt;</description>
            <author>Michael Firth</author>
            <pubDate>Fri, 26 Jan 2018 13:28:01 +0000</pubDate>
            <guid>https://www.neowin.net/news/twitter-deploys-neural-network-to-improve-auto-cropping-of-images/</guid>
            <media:thumbnail url="https://cdn.neowin.com/news/images/uploaded/2018/01/1516966827_twcrop_story.jpg" width="760" height="428" />
                        <neowin:twitter>@techfirth</neowin:twitter>        </item>
        	</channel>
</rss>
