junyanz / iGAN

Interactive Image Generation via Generative Adversarial Networks
MIT License
3.97k stars 587 forks source link

projecting an image onto the manifold? #2

Closed dribnet closed 7 years ago

dribnet commented 7 years ago

Is the model for projecting an image back onto the manifold also packaged with the pre-trained models listed in the README? The code is organized very well and I've been able to use the Model class in `dcgan_theano.py to sample from the downloaded models, but I'm less clear how to do the inverse mapping x => z through the projection model and optimization objective.

junyanz commented 7 years ago

Hi Tom, you are right. It's not included right now. I will release the script for predicting latent vector z from given image x, as well as other training scripts soon (hopefully by Sept 25, Sunday).

junyanz commented 7 years ago

I just uploaded a script `iGAN_predict.py' for this. Let me know if it works for you.