faceswap VS DeepFaceLab

Head over to deepfakes/faceswap and install all the stuff that it asks you to do and then open the terminal with in faceswap env from anaconda.

But there's a nice survey on Arxiv here of various deepfake / face swap methods. Some of methods listed in the table on page 4, such as Faceswap and Faceswap-GAN, apparently use encoder-decoder models. I think Faceswap-GAN was the one that I was thinking of in particular; apparently it adds a perceptual loss and an adversarial loss to an autoencoder.

My current challenge is getting those masked wheel images to be able to swap between images, or to apply a new wheel on a car image. To get a decent result that doesn't look fake, it would have to do some minor warping and resizing. To me, this seems like exactly what the Deep Fakes repo does. https://github.com/deepfakes/faceswap

It keeps getting stuck at either"fatal: unable to access 'https://github.com/deepfakes/faceswap.git/':" or "Please run this script with Python version 3.7 or 3.8 64bit and try again."

https://github.com/deepfakes/faceswap bu uygulamadan yaparsın

Here's the faceswap github

Deepfaking as it is right now, is mostly two types of software, and two very different methods. The first is making a video with a "face swap". Those use software called Deepfacelab or Faceswap. I personally like Deepfacelab because it's more "configurable" at least how I see it. This is what you see on youtube or what makes headlines when the media decides to show clips of the dangerous political deepfakes.

AFAIK the only app which supports whole head swapping is DeepfaceLab but it needs to be trained with lots of images and lacks a proper GUI.

I think DeepFaceLab is still the leading open source option.

Click to Learn more...

Here it is for free: https://github.com/iperov/DeepFaceLab I guess the repo isn't available in China.