Scaling up GANs for Text-to-Image Synthesis · HackerPapers discussions

Scaling up GANs for Text-to-Image Synthesis — mentioned by godelski

Thu, 14 Dec 2023 23:17:04 GMT

Sorry, I left too much knowledge for inference. I presumed it was safe to let users infer that a 2018 (StyleGAN) or 2019 (StyleGAN2) model was not going to compete with performance of a 2023 model, regardless of architecture. There has in fact been improvements in GANs in the last 5 years. Text conditioning is not unique to diffusion and is rather a subnetwork for conditioning (actually StyleGAN's whole innovation was creating a subnetwork for conditioning synthesis) There are definitely modern GANs that are T2I and computationally cheaper than a latent diffusion model for a comparable…

Paper and discussion history. A mention is not an endorsement.

Scaling up GANs for Text-to-Image Synthesis — mentioned by gwern

Wed, 23 Aug 2023 16:17:00 GMT

If you're wondering how it's so fast and cheap and they can generate variants so easily, it's because they're using GANs (see the footer). GANs are way faster than diffusion models because they generate the image in a single forward pass and their true latent space encoding makes editing a breeze. (And if you're wondering how it can look so good when 'everyone knows GANs don't work because they're too unstable', a widespread myth, repeated by many DL researchers who ought to know better, GANs can scale to high-quality realistic images on billion-image scale datasets, and become more, not…

Paper and discussion history. A mention is not an endorsement.

Scaling up GANs for Text-to-Image Synthesis — mentioned by GaggiX

Fri, 10 Mar 2023 02:29:54 GMT

Link to the paper: https://arxiv.org/abs/2303.05511

Paper and discussion history. A mention is not an endorsement.