Skip to the content.

This Food Does Not Exist 🍪🍰🍣🍹

We have trained four StyleGAN2 models to generate food pictures. The images below are purely synthetic!

The code optimized for TPU training as well as the pretrained models are openly available.

cookie-256.pkl

cookies

cheesecake-256.pkl

cheesecake

cocktail-256.pkl

cocktail

sushi-256.pkl

sushi

Why not DALL·E/diffusion models? 🤔

Recent methods like diffusion and auto-regressive models are all the rage these days: DALL·E 2, Craiyon (formerly DALL·E mini), ruDALL-E… Why not go in this direction?

Realism vs control

StyleGAN models shine in terms of photorealism, as can be some by some of our food results. For another example, the website ThisPersonDoesNotExist.com produces very believable face images. While GANs are still better at this, diffusion models are catching up and this may change soon.

Diffusion models offer better control and flexibility, thanks in large part to text guidance. This comes at the cost of larger models and slower generation times.

Training resources

We were able to train the provided models in less than 10h each using a single TPU v4-8:

Training plots

FID (Fréchet inception distance) is a metric used to assess the quality of images created by a generative model.

In comparison, Craiyon is being training on a v3-256 TPU pod which means 32x the resources (albeit using the previous TPU generation) and the training has been going on for over a month.

Result comparison

No cherry-picking!

Ours

bdc76775-2c9f-4110-a2f1-fcbc07a588e7

Craiyon (“a pile of cookies on a plate”)

a-pile-of-cookies-on-a-plate

DALL·E 2 (“a pile of cookies on a plate”)

Screenshot 2022-07-20 at 15 31 55

Acknowledgements 🙏