-
thank for your work .
i am confused with the stage in lag ,
in the code ,the channels of weight are smaller and amaller along the bigger of stage .
i do not unnderstand the meaning of this op…
-
I want to ask some questions about the model:
1. Why to resize the image into 8/16/32/64/128/256..., if I want to generate a 256 resolution image, I just need to resize the image to 256 * 256 for t…
-
It's mentioned briefly in stylegan2 paper's appendix b, but it is described in [Progressive Growing of GANs](https://arxiv.org/abs/1710.10196) in detail in section 4.1.
It's done for both conv and …
-
May be the resolution of the image in the sample looks like 28*28
-
https://github.com/facebookresearch/pytorch_GAN_zoo/blob/7275ecbf53a9db7e4bc38c4c5136c10c4950724b/models/networks/custom_layers.py#L72-L74
The above implementation applies the weight scaling to the…
-
# Abstract
Generative Adversarial Networks(GAN)은 데이터 생성에서 뛰어난 모습을 보이고 있다. 많은 영역에서 쓰이고 있지만 여전히 안정적인 학습에는 어려움이 따른다. 문제점으로는 Nash-equilibrium, internal covariate shift, mode collapse, vanishing gradient,…
-
We have been exploring SimCLR as an auxiliary loss, but one could also just train SimCLR *directly*, skipping GANs entirely. One could probably use the official Google codebase, which already has TPU …
gwern updated
10 months ago
-
-
Hi, I'm reading your paper recently and have the following questions.
1. In your paper, you use StyleGAN and PGGAN in the experiments, and I wonder if InterfaceGAN can manipulate any pre-trained GAN…
-
Can you share the trained model and relevant dataset, if it is ok.
Thanks in advance.