4446.1746 In reply to 4446.1745
<<Is it only possible to render images selected
randomely by the AI
Yes because you have the total Prompt! ;)
and the seed, format etc...
so you must have the same image with your own variations if you want!


- or is it also possible to feed the AI wit a specific image pool
it to mix/melt them together ?
Yes you can also make that! ;)
Fields of possibilities are some large!
example : Steampunk cityscape of new york city, by tim hildebrandt, wayne barlowe, syd mead, john harris, bruce pennington, donato giancola, larry elmore, oil on canvas, artstation, pixiv, cinematic composition, dramatic, beautiful lighting

Model
Stable Diffusion 1.5
Seed
929179597
Guidance scale
7
Dimensions
512 × 512

so for Wombo take juste something like this and put capitale letters to Names
Careful just 200 letters!!!
steampunk cityscape of New York city oil on canvas, cinematic composition dramatic beautiful lighting Tim Hildebrandt
Original Stable

Inside Wombo

Hi there,
Please be aware that any AI image generator relying on Laion5B is built upon massive art theft and privacy infringement, as the images used for training were not opt-in. None of the artists whose work was used gave their consent to it.

As a matter of fact here are two crowdfunding campaigns currently ongoing to rectify that, one in Europe and one in the US :

https://www.gofundme.com/f/help-protect-our-art-and-data-from-ai-companies
https://www.gofundme.com/f/protecting-artists-from-ai-technologies

Also important :

https://www.youtube.com/watch?v=5Viy3Cu3DLk

AI-driven image generation is undoubtedly fascinating, but in its current state is highly unethical. It gets even worse when people start prompting it to emulate the style of specific artists living or dead like Moebius, Kim Jung Gi, Miyazaki, Syd Mead, and so on. Feeding their work into ML is IMHO the polar opposite of their legacy built on their dedication to their craft.

@ Pilou : Thank you for your great tips. very appreciated !!
.
@ Pior : It is incredibly fascinating what the latest AI is capable of doing
but of course copyright must be respected and this is surely a very
hot topic - steal or no steal ?!

Like Chees, Go Game, etc... Ai are now more strong than humans!
I have believed that I don't see during my life a computer win a Go Game play against a Pro Player : in 2 years that was finished!
4 -1 for Alpha Go... some time after the new Alpha Go (who learn alone) 100 - 0 again previous Alpha Go!!! Human has no chance except he changes all that he has learned over the millennia ! :)

That will be the same for graphic, 2D, 3D, Design, Expert System etc...
Explore this site... https://www.altair.com

Well, as far as I am concerned there is no debate, and all I can do is bring up awareness to the issue and link to the crowdfunding efforts in the hope that people donate. And hopefully, besides legislation one can hope that the cultural backlash against this stuff will be strong enough. I am not interested in arguing "for or against" here.

Whether or not someone decides to trample over an artists legacy just because it amuses them and fulfills their need for instant gratification using an unethically developed tool is up to them.

@FP : These AI images generators are not just cleverly designed algorithms. They only exist and function as well as they do because they are using scraped artwork as an input in a way than none of the artists ever agreed to. You are free to do as you please of course since you are the one making your own moral choices. All I can say is please look into the issue further. IMHO every single one of these images is an insult to the artists who got unknowingly fed into the dataset.

@Prior
it's the beginning...
Yes it's not fair to use the artist's work without permission or without some retribution when something is sold.
But how solve this ? Here artists are dead from centuries...
Seems it's not an insult but more an homage!
Of course that is for publicity but this is an another story! :)

As said, up to you to consider it an "hommage". As far as I am aware I have not been dead for centuries, yet my work (and the work of millions of others of course) is part of the DB of scraped artwork. SamDoesArt is very much alive as well, yet this is happening :
https://youtu.be/5Viy3Cu3DLk?t=513

Also, the future is not set in stone - there has already been precedents where companies using unethically gathered data for ML have been obligated to destroy their model. So there's hope on that front.

So again, I won't be arguing here because firstly this is not the place to do so, and secondly I believe that you probably aren't aware of what's actually going on, and there is likely some stuff being lost in translation too. So please look into it. A good place to start would be to look at the GoFundMe pages linked earlier.

Prompt "mickey Disney" (we can't write anymore prompt in free version :(
images not for sale only for pleasure of the eyes!
The problem is for the company if they don't retribute Disney not for for me! ;)

In 3D...

HI there,
You're obviously not understanding (or pretending not to understand) my earlier posts about the source of the current 2D ML models, so I'll leave it at that.

100 variations https://imgbox.com/g/4hJLVf6w7W

@Pilou :-) Really impressive how well the new addition fit to the original !
"laitiere" image !

Yes this "Outpainting" function is incredible!



And this other thing absolutely terrific in 2D painting!!!
https://twitter.com/i/status/1604186054923210752

A free one for increase image https://github.com/upscayl/upscayl

Almost scary what theAI is capable of but its is like with any invention -
the advantage and disadvantage depends on the ethics of the users ...

Maybe more the ethic of companies! ;)

Prompt "ethic" - style Surreal

Very cool Pilou !
.
I did some research and found out that what I am looking for is
called "Algorithmic Paredolia" performed by digital neural net-
works. The first algorithm of this kind was created by Google and
is called "Deepdream"
.
I wonder if it is the same / modified or a new algorithm used
for the "remix" function of Midjourney which generates pretty
amazing results.
.
Now I am searching for such an offline open-source tool to
merge imagery and expect to find some sophisticated stuff
which is capable to "learn" what I expect it to do by iterative
processes.

Ah yes the Pareidolia effect! :)
It was the beginning of the Ai period!
https://surrealismtoday.com/beginners-guide-artificial-intelligence-art/ (bottom page)

But alas Deep Dream is not more free!
https://deepdreamgenerator.com/subscription/plans

I imagine a "Pareidola" software with a node based system where I can link
a pool of images, patterns and textures together and to be able to teach the
neural ntework based system in which way I wont those nodes applied to
my base image. And to be able to create quick iterations by changing the
respective parameters.

Not sure that teach a neural network is an easy thing!

For a still image you can create a big prompt very elaborated who will render that you want! ;)

choose your image(s) !
https://lexica.art/?q=pareidolia