Peiye Zhuang

I met the same issue using normal CelebA and seemingly it doesn't work at all... 

> Hi, I'm very interested in your work. And I'd like to try editing outdoor images, could provide the classifier weights pretrained on Transient dataset? Thank you. Hi, the regressor...

Hi, thanks for your interest and sorry for the late reply. I uploaded a file named "train_multi_attr.py" for training multiple attributes at a time. Basically, you can control the target...

Sure, added the real image backpropagation code in https://github.com/KelestZ/Latent2im/blob/main/BP.py.

> > Sure, added the real image backpropagation code in https://github.com/KelestZ/Latent2im/blob/main/BP.py. > >  > hello, could you privide this module. thank you very much It is a naive VGG16....

> Thank you for publishing your code, your work is great. Do you have the evaluation script that created Table 1 (Quantitative evaluation of numerical changes on the other semantically...

> Isn't there a way to use pre-trained weights and do inference? Thanks for your interest in our work! Could you please mention a bit more details? I'm assuming it's...

> I would also really appreciate that! Thanks for the interest and got it. I'll start on it!

> Thanks for your reply! I reimplement your paper on pre-trained StyleGAN2 with attribute regressor pre-trained on CelebA dataset. However, the regression loss fails to drop smoothly. I have some...

> Thanks for your reply! I reimplement your paper on pre-trained StyleGAN2 with attribute regressor pre-trained on CelebA dataset. However, the regression loss fails to drop smoothly. I have some...