I tried a couple of agents and stuck with open art consistent character. gonna try higgsfield as well. the hardest part is her body. i generate a lot of junk until i get something that looks like her. for consistency in video, i ask veo 3 to change the angle and screenshot for a new frame to video prompt, so it looks like a doc with two cameras.
Here's a podcast from "the real" 2065 analysing the project https://youtu.be/e67OwpvhUfk
Sorry, I'm not up to date with image generation - but how do you achieve consistency (same face and body) across multiple pictures and a video?
I tried a couple of agents and stuck with open art consistent character. gonna try higgsfield as well. the hardest part is her body. i generate a lot of junk until i get something that looks like her. for consistency in video, i ask veo 3 to change the angle and screenshot for a new frame to video prompt, so it looks like a doc with two cameras.
https://www.reddit.com/r/StableDiffusion/comments/196ikpo/wh...
From what I've seen LoRAs tend to focus on styles, characters, hand fixes, face fixes, clothing. They're quite adaptable.
lora has been on radar. i tried to run locally via confyui, but i don't have a very good GPU available. it takes AGES.
https://about.ideogram.ai/character Ideogram released consistent character from a single picture 2 days ago.
seems nice!! gonna check it out
Music sounds like the 20's? 40 years ago it was very different, so I assume 40 years in the future it will change a little as well.
yep! she's retro. :)
Can't wait, but the acturial tables tell me I will dead by then.
sounds like you literally can't wait :)
i'm definitely not able to wait.
lol "my grandma listened to Billie elish, so that's where this retro sound comes from"
billie will be old one day lol
https://m.youtube.com/watch?v=-M6cnv8MXfI
We're almost there, friends.
ok