The Changeling logo
Changeling

Imago Realtime

Your expressions, someone else's face, live

You become the face in one picture, with your own smile, your blinks and your mouth as you talk, and your call sees it as an ordinary camera: Zoom, Teams, Discord or Meet. The personas are AI-generated people who don't exist.

InsightFace inswapperXSeg occlusiononnxruntime-gpu OBS Virtual CameraPython

What it looks like

Each row: a face as the camera sees it, the same frame through Changeling Imago Realtime, and the persona it took the face from. The pose, the expression, the hair and the clothes stay the camera's; the face is the persona's. Click to enlarge.

Three faces, each as the camera sees it, then swapped to the persona, then the persona picture

Use it

  1. Put the persona's face in personas/input.jpeg (the repo comes with one).
  2. Start run.bat (or run.ps1, run.sh). The first time it sets itself up: Python 3.11, PyTorch's CUDA libraries, InsightFace and the GPU ONNX runtime (about 3 GB), the swap model (about 550 MB) and the occlusion model.
  3. A dialog asks: use the default input? No opens a file picker in personas/.
  4. In your call, pick the camera OBS Virtual Camera. OBS Studio must be installed for its virtual camera; it doesn't have to be running.

Keys

Sswap on / off
Oocclusion on / off: hands, hair, a cup stay in front
Vvirtual camera on / off
Ppick another persona, no restart
Rread the persona's picture again
Q Escquit

Sliders

smooth %steadies where the face sits from frame to frame
strength %how much of the swapped face shows

The personas

Generate them rather than using photos of real people: becoming a real person on a call isn't fair on them. What works best: looking straight at the camera, the whole face visible, even light, sharp, a neutral face or a slight smile, one face per picture.

How it works

1

Find you

InsightFace's SCRFD finds your face in every frame; its five key points are steadied against the last frame's, so the swap doesn't jitter.

2

Swap

inswapper_128 draws the persona's identity, read once from her picture, onto your face, on the GPU.

3

What's in front

FaceFusion's XSeg occluder marks hands, hair and cups in front of your face, and the swap is left out there.

4

You, still

Your own eyes show through when you blink, and the result goes back into the frame, to the preview and OBS's virtual camera.

InsightFace's models (buffalo_l, inswapper_128) are for non-commercial use only.