FaceFusion is an excellent tool for swapping faces, colorizing, or enhancing the quality of photos and videos. It's completely free and can be run on your own machine (requires sufficient computing power).

<aside> ☝

The settings and notes in this guide are based on FaceFusion v3.4.1.

</aside>

References

• FaceFusion documentation:
  • The official GitHub repository.
  • The official documentation.
  • The error codes reference. Note that when errors occur, nothing appears in the terminal. To see the error code, run echo "Exit code: $?" immediately after.
• Vast.ai documentation:
  • Volumes - Vast.ai Documentation
• My custom versions: facefusion-3.4.1 and facefusion-3.3.2. These versions disable filters and add several new features.

Understanding Key Parameters

• --execution-thread-count
• Face detection parameters
  • --face-detector-model
  • --face-detector-score
• Face Swapper Parameters
  • --face-swapper-model
  • --face-swapper-weight
  • --face-swapper-pixel-boost
• Face enhancer parameters
  • --face-enhancer-model
  • --face-enhancer-weight
• Face masker parameters
  • --face-mask-types
  • --face-mask-blur
  • --face-mask-padding
• --reference-face-distance

Tips for Using FaceFusion

• Avoid using reference photos with glasses when swapping faces.
• Use 4-6 reference photos with different angles and expressions for best results.
• Instead of splitting videos for testing, use the "Trim Frame" option (processes only the selected segment) and "Preview Frame" (displays the generated result at each frame).
• Models are stored in .assets/models/
• Temporary files (uploaded references and targets) are stored in /tmp/facefusion/ on Linux systems.
• Output fps of 35 is sufficient—higher values aren't necessary.