Watch the two videos side by side. It is the same young woman, in the same place, with the same llama. Look at the eyes, the nose, the shape of the face and the hair.
In the second take she says something else: «Me trying to look pretty in photos. And this llama sticking its tongue out at me! How cheeky!». The words change, and the face stays.
Why a face drifts
Every time you ask the AI for an image, it draws it again. If you only describe the person in words, each photo comes out a little different: a mole that moves to the other cheek, a thinner face, freckles that were not there before.
A reference photo, which the AI looks at every time it works, keeps the face. The people who teach how to make AI influencers say the same: it all depends on one well-made reference image.
First the face is locked. Then comes everything else: the clothes, the place, the product and the video.
The order, step by step
- Close-up photos of the face, from several angles.
- A sheet with the face and the body, which the AI uses as a guide.
- Every new photo and video starts from those same images.
In UGC Viral Studio, your influencer's photos and videos open once those face images exist. That way, nothing is generated without its reference.
What a locked face gets you
People who follow you recognize your AI influencer in the first second, before they hear what it says. Every new video adds to the same face.
And you are spared checking every photo with a magnifying glass to see whether the nose changed or a mole moved.
The face is half of it. The other half is the voice, and that is the next lesson.