use e4e latent

[loboere]

is possible load a e4e latent?

[mchong6]

If I am not wrong e4e is pretrained on 1024 resolution GAN, so it wouldn't work directly since I am using a 256 resolution GAN. However, my method should generalize to 1024 resolution GAN, so I can see you working with e4e that way.

[amadeuzou]

I used origin latent projected by face.pt with a real world image, and got a nice result:
real world image | projected image as input | generated reference image | output styled image

[mchong6]

This is very cool and works better than I imagine. Did you use the projector.py code to do GAN inversion?

[amadeuzou]

yes, the projector.py code in rosinality's project.
but i got bad results with your train.py code on disney dataset:

[sample_disney/010000.png]

[mchong6]

Did you start training with face.pt?

[amadeuzou]

yes, the early model results are normal, such as 1000iters.

[mchong6]

try using early model then. IIRC, this model I uploaded is only 2k iterations.