(Note:The weight of the example image from left to right is: -1,0,1)
Adjust the age of the characters while maintaining feature consistency, while minimizing the impact on the composition and style of the original image.
After testing, it has been found to be effective in male, female, furry, anime, and realistic settings.
POTENT SUGGESTION WEIGHT:[-1,1][0,1.5],
WEAKEN SUGGESTION WEIGHT:[-2,2][0,2.5],
MATURE SUGGESTION WEIGHT:[0,1.5],
Be cautious when using negative weights!
Use positive weights to increase character age and negative weights to decrease character age
Description
This version has more extreme age control, but there are greater changes to the composition
FAQ
Comments (9)
Is there any way you could post these LoRAs on TensorArt?
I have never used Tensorart before. If you want to move it, please indicate the original website and original author.
@lse14 I looked here and saw that someone had already posted your LORA there!
@NamelessKing2 I have already read it, and they have marked the original Lora website and author for the move, so I am too lazy to spend time on it
it only work for 1 person? what if theres multiple characters? how the lora detect which character should be positive or negative age
Lora applies to all characters that appear, with positive weights becoming older and negative weights becoming younger
@lse14 so technically i can't do one girl to be aged up and the other to be aged down, i see. .
@monicalucci Perhaps you can use Lora during inpainting
Hi! I always appreciate your work! I have tried both the weak and potent versions, and I prefer the potent one at a -0.7 to -1 weight for younger subjects. Above that, the output gets washed out and makes the subject look more like a chibi. The weak one, true to its name, has a weaker effect than the potent version. It’s the perfect choice for anyone who wants a younger face while maintaining the subject's original body shape. Overall, they both serve a good purpose. Great decision not putting all of that into a single LoRA!