Great concept. Having a lot of fun with this one. Thank you.

😻💕👍🏻

Aye! this is fun to mess with, great work!!!

Does it get any better than this? But on a serious note, this proves that AI can generate excellent realism from a sketch, something that workflows for Flux still struggle with terribly. The utility of this idea cannot be overstated, and much more work is needed in this area. Current models have so much potential waiting to be unlocked.

all the faces seem kinda Americanized from what i see, hoping this can be changed with proper promting

That AI dude on youtube doing these animations for old TV shows is in shambles.

I wonder if it could turn realistic to anime with negative strength

worked amazingly out-of-the-box, this is dope, thank you a lot, lot of fun

POV: bro is cooking

thanks for the lora, working amazing!

is there possible to transfer immediatelly?

All my attempts result in a transform lasting only a microsecond followed by realism. How do you make the transform take longer? Do you have a sample ComfyUI workflow?

Super glad this exists, I've been trying since Wan 2.2 came out to get this same effect without loras to no avail. I'll give it a shot as-is but are you also considering giving it a 2.2 update?

I have been having SO much fun with this - I never used WAN 2.1 before but it's very amazing with your Lora ! I would be very keen for a WAN 2.2 version if possible please please ! Thank you so much for your kind donation of this Lora 😊