Player not loading? Watch on YouTube
This tutorial covers a Windows installation of FaceFusion 3.8.2 with OpenVINO for CPU-based face swapping. The speaker says OpenVINO can improve CPU processing without a dedicated graphics card, but gives no performance measurements. They also describe face enhancement, lip sync, age changing and background removal; the usage example focuses on swapping faces between pictures.
The listed prerequisites are Python 3.11.0, FFmpeg, Git for Windows and Visual Studio C++ tooling. The walkthrough clones the FaceFusion repository, creates and activates a virtual environment, and installs Torch, Torch Audio and Torch Vision. It then edits the requirements file before installing dependencies. The narration says to remove ONNX Runtime and change a dependency it calls "sippy", but does not state the replacement version.
Next come OpenVINO and ONNX Runtime OpenVINO. The first launch downloads models and displays a local URL for opening the interface in a browser. This local AI setup therefore needs model downloads before use. The speaker also shows how to save a run.bat launcher for subsequent sessions.
For the basic swap, the user uploads a face picture and a target picture, waits for the preview, then clicks Start. The description asks users to work with content they have permission to use and avoid impersonation or deception.
