The Stable Diffusion Neural Network can be further trained to generate images in a certain style.

Versions with two pupils are quite common.

The Stable Diffusion Neural Network can be further trained to generate images in a certain style, as in the Waifu Diffusion. And the editor of PiyarSquare took and trained a neural network on footage from The Simpsons.

The result is a model that can turn any human (or animal) into a Simpsons character. This is what it looks like in the test images provided by PiyarSquare itself.

We got interested, and we immediately downloaded the two-gigabyte model and started experimenting. It seems like the neural network was worth spending a little more time training: it’s got some good choices, but that’s the exception, not the rule.

We generated Simpsons versions of several famous people or popular characters without filtering out the bad footage. Here’s what it looks like; the most common problem is multiple pupils, but there are other unusual disadvantages.

Hideo Kojima is a.

♪ I’m the one who’s gonna make you ♪.

The Cratos.

Daryl from The Walking Dead.

The Herald .

I ‘m not going to lie .

Triss .

The circus.

I ‘m not sure .

No, we don’t know why he looks more like Geralt.

Eloy , please .

Sauron is.

I ‘m not going to let you do this .

Elrond .

You can download the model and experiment with it yourself here.

Similar Posts