After a long period of AI technology development, the current AI face swapping technology has become increasingly mature. In addition to being able to swap faces in photos, it can also swap faces in videos.
FaceFusion is a free face swapping tool that also offers facial repair and enhancement features. It can be self-deployed or used for free with the Google Colab version.
The FaceFusion GitHub project is licensed under MIT:
https://github.com/facefusion/facefusion
FaceFusion provides a simple main interface with easy operation:
Local deployment
If you want to deploy it yourself, you can refer to the following installation guide.
https://docs.facefusion.io/installation
Google Colab (requires magic)
If you don't want to bother or don't have a GPU to avoid slow running, you can directly use the Colab version for a quick experience.
First time using Google Colab? Take a look at this article: Free use of Google Colab to deploy Stable Diffusion
-
Open Colab
https://colab.research.google.com/github/facefusion/facefusion-colab/blob/master/facefusion.ipynb -
Launch the project
Click the play button inside the red box in order from top to bottom, and then wait for the project to download, install, and launch.
3. Access FaceFusion
After about 10 minutes, if you see the following content, congratulations on a successful launch.
Click the link inside the red box to enter the FaceFusion main interface.
4. Start face swapping
While writing this article, I thought of the battle of Ma and Zha. So I will find a photo of each and use Musk's body shape and attire to replace Ma's face with Zuckerberg's, without even changing the hairstyle.
Upload Zuckerberg's photo in the SOURCE panel:
Upload Ma's photo in the TARGET panel:
Then in the PREVIEW panel in the upper right corner, the "Ma body with Zha face" after the face swap will be automatically displayed.
Looks like Ma and Zha, haha!!! Just a little blurry!