I am releasing my next blender addon here: Syntexmex.
Which is “Synthesis of PBR Textures on UV-Meshes based on Example Images” (== Syntexmex )
Right now, there are pre-built packages available for Linux and WindowsI don’t have a Mac so we’ll see about that ;). Depending on the number of requests for this I will either do it or not … . I would also help anyone with a mac to build package for maxOS.
The addon does pretty much what its name says: it takes an example image and
can then synthesize a new texture of almost arbitrary size on a UV map. Additionally
it makes the seams between UV-islands & cuts seamless. It will automatically
create a new material from the synthesized images and it also works
on PBR textures!
A list of some examples which I hope I can extend pretty soon once poeple start using it ;):
Please post all your questions here in this thread! This method is pretty new for blender users and I have a little bit of a biased view on it as I am so deep in the topic. I will support all questions in this thread and also change the documentation accordingly if things are unclear.
Don’t be surprised about the size of the addon (~95MB). That is because the addon makes use of quiet a few 3rd party libraries.
The error usually gets shown due to these things: 1. Not chosen a material 2. There needs to be an image texture in the material to choose from. 3. Library size value is too high (the algorithm needs more memory than is available on you system in that case, turning down library size will help in this case.). And if course there might still be some bugs. As this add-on is really new. So I actually need the community to test it a little bit ;).
@atekdigital: I am afraid there isn’t one yet. I hope though that the documentation is sort of complete? If you have any question don’t hesitate to ask and I will include them in the documentation.
The Size of the addon btw is due to the number of 3rd-party libraries that are used to enable the functionality. It makes use of some AI-libraries that are quiet large in size (sklearn).