Prisma was a land grab because the original authors didn't make an app. Hell I started working on my own version at the time since it was such an obvious app.

The original authors finally got around to making their DeepArt app, so we have at least two players in this space atm.

I think either of these two players will incorporate this research, but there's no clear incentive for anyone else to try and get in on this market now. It's not clear that there's a whole lot of money to be made here, but who knows maybe usage will explode. I mostly expect an acquisition.

Looks like Zuckerberg is working on a version too - https://www.facebook.com/zuck/videos/10103204449698911/

Plug for my mostly ignored version, that runs on Mac & Windows. It processes high res images unlike Prisma, and runs entirely locally. And can process videos of arbitrary length & resolution. http://macdaddy.io/Style/

I will try it out on Windows10 in the next days! Seems interesting for processing photos instead of having to use Prisma on Android.

I imagine we will see many clones when Google open sources their implementation too (as mentioned in the article, along with others mentioned in this thread as well). I wasn't aware of DeepArt. I'll check it out.

There are already quite a few open source implementations. Some of them are even pretty decent.

Prisma never had "ownership" of this effect; the papers of Neural Style transfer have been out for months, with this Google method being another iteration.

Prisma was just the first polished app released using style transfer. (However, since the technique is public, it will make an acquisition of Prisma less worthwhile)

Sure, but I was talking about market ownership, i.e., the most market share for users looking to achieve this type of effect, not IP ownership.

Of what use is market ownership when no one is going to buy you. Facebook is already working on its own style transfer algos.

Yeah, see my original comment at the top of the thread.