stylegan2-pytorch VS stylegan2-ada

Another great place to see AI generated faces: https://thispersondoesnotexist.com/

Check out https://thispersondoesnotexist.com/

Similarly check out This Person Does not Exist . com

There is this website called This Person Does Not Exist that shows you a random AI generated photo each time you refresh.

Jesus Christ you are all so gullible, it's a picture made by this

Yeah same, I'm usually able to spot stuff made by AI, But I actually thought this was real. I went to the site op used and some of the results I got were genuinely convincing.

Do they already know what you look like? If not, you can try this, just keep refreshing until you find someone you like. https://thispersondoesnotexist.com/

If you smudge just a few locations I doubt it would fool a simple discriminator. You could also train a discriminator that is robust to post-processing by using augmentations. This was popular with StyleGAN models: https://github.com/NVlabs/stylegan2-ada

https://github.com/NVlabs/stylegan2-ada + clip guided diffusion

I understand the concept you're describing, but I'm unsure how to implement it in practice. The documentation on Stylegan2's GitHub page doesn't mention anything on it I can see. https://github.com/NVlabs/stylegan2-ada

You have to modify the code, it's line 60 in https://github.com/NVlabs/stylegan2-ada/blob/main/generate.py

You should check stylegan-2 ada, it works on colab and can be trained less than 12 hours tensorflow implementation

with a process called ADA https://github.com/NVlabs/stylegan2-ada

Training generative adversarial networks with limited data [NeurIPS 2020] Tero Karras, Miika Aittala, Janne Hellsten, Samuli Laine, Jaakko Lehtinen, and Timo Aila. arxiv: https://arxiv.org/abs/2006.06676v2 git: https://github.com/NVlabs/stylegan2-ada

As official as it can be.

I direct you to the issue on the subject in the NVLabs/stylegan2-ada repository.