DeepArt.io introduces style transfer for videos

It bases on:<p>- video: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=Khuj4ASldmU" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=Khuj4ASldmU</a><p>- paper: Manuel Ruder, Alexey Dosovitskiy, Thomas Brox, &quot;Artistic style transfer for videos&quot; <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1604.08610" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;1604.08610</a><p>- code: <a href="https:&#x2F;&#x2F;github.com&#x2F;manuelruder&#x2F;artistic-videos" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;manuelruder&#x2F;artistic-videos</a>

At 1k EUR for a 5 minute video, I can&#x27;t imagine they&#x27;re getting much in the way of sales.

Today I learned that I could be making a lot more money on the side doing this for half the cost with a better method.

I wonder if there are ways to reduce the GPU RAM requirements on this. Even the commercial offering doesn&#x27;t go beyond 720p - I guess that fills up 12GB on a Titan X?<p>On stills style transfer you can process the image in multiple tiles, but I don&#x27;t expect that would work with video with the temporal consistency

<p><pre><code> University of Tübingen has a pending patent application for the Neural Art technology. </code></pre> I wonder what that is (can&#x27;t find on Google)?

<a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=11873806" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=11873806</a>

How big companies scale dnn as Google Photos or Siri? Any approach? you need to train big networks for different users?

Watching those videos is how I always imagined what being on LSD would be like.