AnnouncementsMatrixEventsFunnyVideosMusicAncapsTechnologyEconomicsPrivacyGIFSCringeAnarchyFilmPicsThemesIdeas4MatrixAskMatrixHelpTop Subs
3
Add topics
Comment preview
[-]JasonCarswell
1(+1|0)1 month ago

You'd think they'd be able to give extra attention to the faces and hands.

[-]HighQulaityDickPics1(+1|0)1 month ago

There are nodes that do such things. Face swap for example, isn't just to swap faces. Its a face detection system, and a good method to insure continuity across works. So if you have a comic, and you need your characters face as the subject for each panel, then you would use something like that.

[-]JasonCarswell
1(+1|0)1 month ago

OOOOOOOOOH! Me likey!

I could draft up model sheets and let the A.I. do the rest. I'd authorize what passes or fails to build up a better library (I don't know if that's the correct terminology), and go in and correct manually occasionally.

I mentioned before, I'd like to roughly sketch, let the A.I. improve it and add details, then go finish up inking or whatever to get "my look" with penmanship, style, flourishes, etc. Once the full black & white storyboard/graphic novel is done I'd then develop colour treatments. And/or maybe animation. I'm telling you my Bittersweet Seeds story (started in 2001) starts slowly but should blow everyone away - especially truth-seekers, A.I.-phobes, and hard sci-fi fans. Maybe I should focus on drafting a treatment folks can read, but so much more is in the carefully specific details.

Among other projects, I have been developing another animated series for almost a decade. Formerly "Trutherism 101", now "Pintle Pygmalion & Galatea Gudgeon", features fantasy characters that you might appreciate. They're on a curious island where they just talk about things in enlightening, educating, and entertaining ways. The primary goals: quality over quantity + free-thinking.

I'm almost tempted to develop a Santa Libertas cartoon too. I hate being in front of the camera, but I'm all I have and all I can 100% count on. Essentially the content would be more or less the same (live or animated, Santa or Pintle & Gudgeon) with a focus on quality over quantity. The idea is to prove that I can make great content to raise funds and attract high quality talent to collaborate, both on camera and off.

[-]HighQulaityDickPics0(+0|0)4 weeks ago

I really want to say "Well, the first step in all of that is downloading comfyUI!" But thats a bad approach because its just going to leave you confused and feeling frustrated because thats what happens to me on a daily basis. So maybe thats the way, to tell you that is horribly frustrating- as is all technology.


There are 2 kinds of ?platforms? that uses this stuff. AAAA1 Automatic and comphyUI. AAAA1 is easier to get into but much more limited in what it allows you to do. Comphy, on the other hand, has a much higher learning curve -BUT it lets you do crazy cool things. Like generating an image and automatically masking out the subject and then using that to replace who is in your generated image on the fly. So thats why i would toss my hat in that ring. Fortunately, there is a lot of good info out there (for now), and a lot of free workflows (for now) that will show/help you do the thing you want to do.




If you want to talk further about this sort of stuff, i enjoy doing it so would enjoy talking about it. Building houses with walls of text is my specialty!