/vt/ - Virtual Youtubers

>>27701126
>>27701051
>or... could it be that robotani can't do that?
Funny you would say that because she's been getting a lot better... She can now generate the entire image from a single input (I don't need to generate a new image for every single layer) It's pretty wonky this way but it's getting better
No color on this one because mendokusai ^^; but the process began from just 4 patches of color that indicate what to generate
Right now she still relies on the shapes you draw so the image ends up looking scuffed, hence the weird face shape and ear placement.
Hopefully, soon Robotani will be able to adjust shapes and pinpoint where to add correct detail, then she'll be able to generate perfect paws from nothing but rough squiggles then we will never need real paws again :]