# How to Transfer a Face From One Image to Another

You can mix the face from one image with body from another one by following these steps:

**1.** Generate the image that has the face you would like to use. Save it to the gallery

<div align="left"><figure><img src="https://3618323355-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjGm8Y44m4zCkKjnnr9Na%2Fuploads%2F4R0JNM9i6NORPiVssgGr%2Fimage.png?alt=media&#x26;token=7dbea830-afac-4ffe-a625-a88b76bc2e9a" alt="" width="375"><figcaption></figcaption></figure></div>

**2.** Generate the image that has the body you would like to use. Save it to the gallery

<div align="left"><figure><img src="https://3618323355-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjGm8Y44m4zCkKjnnr9Na%2Fuploads%2FMihApFpfXFJKNZLk9rXn%2Fimage.png?alt=media&#x26;token=6f85bb58-4035-41ae-9155-244d8739abea" alt="" width="375"><figcaption></figcaption></figure></div>

**3.** Select the Stable Diffusion model you want to use for the final image. It should be one that supports ControlNet.

**4.** Create a mask image that only has the face with transparent background. You can use the ControlNet preprocessor to do this:

1. Drag the “face” image from the gallery to the controlNet module.&#x20;
2. Select the Segmentation preprocessor (second one from the bottom on the left side of the CN image)&#x20;
3. Tap the subject mask button and make sure you have transparent background selected. This should create the mask you need with only the face. Note: If your face image also has a body or other things you need to remove from the mask, click the “Send to Main Window” button to the right of the CN image, then activate the "Lasso Eraser / Mask Tool".

**5.** Select the InPaint CN Model.

**6.** Restore the “body” image from the gallery with all of its settings, and turn seed shuffle OFF.

**7.** Drag and drop the “body” image from the gallery to the Guide Image module. Set strength between 50-65%.

**8.** Zoom and pan the CN face mask image so it is in the correct position relative to the body image.

**9.** Generate the image

**10.** With real-time rendering ON, adjust the position of the face mask, steps, guidance, etc. until the image is right.

<figure><img src="https://3618323355-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FjGm8Y44m4zCkKjnnr9Na%2Fuploads%2F7N2ybnUIADDkunUB1WBc%2Fimage.png?alt=media&#x26;token=f4708bd4-c482-439b-bdbb-6ff4be80101a" alt=""><figcaption></figcaption></figure>
