prompt: DISCO ZOMBIES, 80s movie poster, text
I mainly wanted to see how SDXL can handle text. Here's a few other examples with different prompts. No ControlNet/LORA models were used.
realistic photo of a witch holding up a sign that says "happy halloween!"
robot holding up a sign that says "AI RIGHTS MOVEMENT", realistic photo, new york city street
a haunted house with a sign posted on the ground that says "abandon hope all ye who enter", realistic photo
all images were generated using RealVisXL: https://civitai.com/models/139562/realvisxl-v20
11 Comments
KingGordon@lemmy.world · 5 pts · 2y
Green Shirt zombie is totally Kevin Bacon. No?
Skrapz101@lemmy.world · 3 pts · 2y
I was also thinking that
myfaceistupid@lemmy.world · 2 pts · 2y
Kevin Bacon does kind of look like a skull
7112@lemmy.world · 4 pts · 2y
These are all interesting. I haven't thought about how AI handled text.
ScoobyDuBois@lemmy.world · 4 pts · 2y
I would watch a "Disco Zombies" movie in a heartbeat. (pun intended)
vine@lemmy.world · 3 pts · 2y
This needs to be a real movie.
Justas@sh.itjust.works · 2 pts · 2y
Their faces kinda remind me of the aliens from "They live"
Misterbonger@lemmynsfw.com · 2 pts · 2y
The clown image is disturbing -- it's like a cross between Ronald McDonald and Phoenix's Joker. Now THAT would be a movie to watch.
intensely_human@lemm.ee · 1 pts · 2y
That robot’s protest sign was definitely made in the CNC shop
yata@sh.itjust.works · 1 pts · 2y
Kinda like how text isn't actually text when you are dreaming because your brain doesn't want to spend resources recreating actual words and meaning when you are sleeping.
intensely_human@lemm.ee · 1 pts · 2y
Pretty soon (aside from copyright nerfing of course) we’ll be able to get full length feature films and 10-season series from prompts like this.