[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/vt/ - Virtual Youtubers

Search:


View post   

>> No.50555534 [View]
File: 402 KB, 512x512, 00197-3702154406.png [View same] [iqdb] [saucenao] [google]
50555534

What I've noticed about AI art so far is that the gestalt is as if an art critic could do art. Doing a generation and walking through the gallery and just going "myeeeeehh, I'd like this if it were more like this or like that..." and then just being able to have your way like that, for as many iterations as you want if you know the techniques and tricks.

Anyways, I'm sick of trying to fix this one, inpainted faces just come out worse than this generated one or leave edges. Doing a full img2img fucks up other details that are fine. I really don't like bad AI art so I'm pretty hesitant in posting anything even if I put in a lot of time into doing and learning. So tell me if it's a bother and if you'd rather me not post cruddy stuff.

Navigation
View posts[+24][+48][+96]