Stable Diffusion - Video Input Strengths

preview_player
Показать описание
Input video made in Blender using a Mixamo model and animation, ran through Deforum Stable Diffusion at strengths of 1 (source), 0.75, 0.5, and 0.25.
Рекомендации по теме
Комментарии
Автор

Strength(0.25) looks like me when I'm looking for free ai generators on the web

ClintochX
Автор

Strength (0.5): It can cut you like a knife, if the gift becomes the fire,  On a wire between will and what will be

Strength (0.25): She's a maniac, maniac on the floor
And she's dancing like she's never danced before

blackryan
Автор

Hey great content! Can i ask you a question? Is that a init strength?

dimaswps
Автор

Stable Diffusion doesn't have "temporal cohesion". Maybe you can take a look at embeddings and hypernetwork training (it's user friendly)

Lucas_Simoni
Автор

Is it img2img with controlnet. Or anything else u used? Thanks for in advance.

vfxartist
Автор

Nice, do you have a tutorial for doing that? For people that do not know how to code

moisanchez
Автор

But why do her leg moves still look weird? Even in super speed... 🤔

nillyk
Автор

Name of the extension or link to stable diffusion?

Manuel-qnuv