search

LynnHo / EigenGAN-Tensorflow

EigenGAN: Layer-Wise Eigen-Learning for GANs (ICCV 2021)
MIT License
344 stars 40 forks source link

Projecting real images into the latent space #2

Closed gabemarx closed 2 years ago

gabemarx commented 3 years ago

Hello!

I was wondering if there is a planned functionality to get the latent variables (U,L,z etc) of a real image that is similar to the dataset the GAN was trained on, essentially a way to recreate the metrics in Table 1 of the publication...

LynnHo commented 3 years ago

@gabemarx There is no plan now. I think the current architecture is not powerful enough, so it's hard to learn a good inversion. I think cooperating with StyleGAN architecture should be better, we will try in the future.