Figure 1. Overparameterizing the StyleGAN latent space during training improves inversion quality. We obtain near-perfect reconstructions ("Inversion A" and "Inversion B") of corresponding arbitrary real images ("Target A" and "Target B"), without having to sacrifice editability, as exemplified by interpolating between reconstructions (middle columns).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.