17
$\begingroup$

Are there any diffusion-based (text + image) -> image generation models available in Mathematica, Wolfram Repositories, or from third parties?

In particular, I'm looking for any stable-diffusion based models, not the original latent-diffusion. These neural nets (which are vastly superior to GANs) for image generation are now very mainstream: DALLE-2, Imagen, Midjourney, etc.

Here's a recently open-sourced diffusion model: https://stability.ai/blog/stable-diffusion-public-release

If it's possible to train or fine-tune that would be nice, but just inference with these models as NetGraph objects would be enough to poke at and play with them.

References:

$\endgroup$
1
  • 2
    $\begingroup$ Please find it here for the diffusion part:mathematica.stackexchange.com/questions/269181/…. The difference between what is posted and stable diffusion, is that stable diffusion applies a VQ-GAN for data preprocessing, which is currently not available. $\endgroup$ Commented Jan 2, 2023 at 11:45

1 Answer 1

4
$\begingroup$

Stable Diffusion-v1 is in the WNNR(https://resources.wolframcloud.com/NeuralNetRepository/resources/Stable-Diffusion-V1/).

$\endgroup$
1

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.