Option 1: a face swap app on the original clip
Several apps let you upload the viral clip and a selfie and paste your face onto the performer. It is quick, but the result is still the original footage with one change: the venue, the crowd, the camera work and the body all belong to the original video. Faces often slide when the performer turns, and you need a copy of the clip to start with.
Option 2: prompt a video model yourself
Reference-to-video models can take a photo plus a short reference clip and generate a new performance. You write a prompt that names the photo as the lead and the clip as the moves, choose the length and ratio, and wait. Expect a few paid attempts: the lead can drift from the photo, the gestures can lose the timing, and recognised music in the reference can be refused.
Option 3: one photo on Dame Hombro AI
This site does the reference, the prompt and the sound for you.
- Upload one photo One adult, face uncovered, head and upper body visible. The outfit in the photo is the outfit in the video.
- Confirm and generate Confirm the photo rules and start the video. It costs 40 credits.
- Download A few minutes later, save the 10-second vertical MP4 with sound, or find it later in your account.
Common problems and fixes
Most failures come from the photo, not the scene.
- The lead looks different: use a sharper, front-facing photo with the face clearly visible and nothing covering it.
- The outfit changed: show the upper body in the photo; a tight head shot leaves the clothes to the model.
- The photo was refused: our checks reject celebrities, public figures, children and photos that are not of one adult.
- Two people in the result: upload a photo with exactly one person in it.