<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Perceptual Losses for Real-Time Style Transfer and Super-Resolution · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/1603.08155/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/1603.08155/feed.xml" rel="self" type="application/rss+xml" /><item><title>Perceptual Losses for Real-Time Style Transfer and Super-Resolution — mentioned by reiinakano</title><link>https://news.ycombinator.com/item?id=15384856</link><guid isPermaLink="false">urn:hackerpapers:mention:1603.08155:15384856</guid><pubDate>Mon, 02 Oct 2017 15:59:35 GMT</pubDate><description>&lt;p&gt;Ahh I see, these are some resources that should help you. The original Neural Style Transfer paper should help you understand the loss functions involved: https://arxiv.org/abs/1508.06576 The paper introducing Real-Time Style Transfer is basically the algorithm used here: https://arxiv.org/abs/1603.08155&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1603.08155/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item><item><title>Perceptual Losses for Real-Time Style Transfer and Super-Resolution — mentioned by nilkn</title><link>https://news.ycombinator.com/item?id=14712911</link><guid isPermaLink="false">urn:hackerpapers:mention:1603.08155:14712911</guid><pubDate>Thu, 06 Jul 2017 18:56:30 GMT</pubDate><description>&lt;p&gt;Slightly off-topic, but there&amp;#x27;s some pretty fascinating work that&amp;#x27;s been done recently on using neural networks for style transfer: https://arxiv.org/pdf/1603.08155v1.pdf http://www.anishathalye.com/2015/12/19/an-ai-that-can-mimic-...&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1603.08155/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item><item><title>Perceptual Losses for Real-Time Style Transfer and Super-Resolution — mentioned by chas</title><link>https://news.ycombinator.com/item?id=13836845</link><guid isPermaLink="false">urn:hackerpapers:mention:1603.08155:13836845</guid><pubDate>Fri, 10 Mar 2017 06:46:05 GMT</pubDate><description>&lt;p&gt;This approach surprised me. Why are they doing feature extraction and then feeding that into a DNN? It seems much more straightforward to have the input of the network be noisy samples and the output be clean samples a la super resolution[0] in images. They probably wouldn&amp;#x27;t want to use fully-connected layers in that instance, but I don&amp;#x27;t see any fundamental barriers if they have enough computational power to run a neural network already. Am I missing something? [0] https://arxiv.org/pdf/1603.08155.pdf&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1603.08155/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item><item><title>Perceptual Losses for Real-Time Style Transfer and Super-Resolution — mentioned by j2kun</title><link>https://news.ycombinator.com/item?id=12064811</link><guid isPermaLink="false">urn:hackerpapers:mention:1603.08155:12064811</guid><pubDate>Sun, 10 Jul 2016 07:07:24 GMT</pubDate><description>&lt;p&gt;IIRC you only get to choose from a fixed set of styles to process your image, right? Then they&amp;#x27;re probably using this algorithm to precompute the style model and seriously speed up rendering: http://arxiv.org/abs/1603.08155&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1603.08155/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item><item><title>Perceptual Losses for Real-Time Style Transfer and Super-Resolution — mentioned by dharma1</title><link>https://news.ycombinator.com/item?id=11898125</link><guid isPermaLink="false">urn:hackerpapers:mention:1603.08155:11898125</guid><pubDate>Mon, 13 Jun 2016 22:07:01 GMT</pubDate><description>&lt;p&gt;Try this, should be near realtime https://arxiv.org/abs/1603.08155&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/1603.08155/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.CV</category></item></channel></rss>
